2014 CCM Review Notes

CRITICAL
CARE
MEDICINE
REVIEW
NOTES
Jon-Emile S. Kenny M.D.
2014
Critical Care Board Study

Notes 2014
Esteemed
reader,
I compiled these notes while studying for the ABIM

Critical Care Medicine Board examination. These
notes were made from a number of fairly recent
resources including SCCM lectures, SEEK questions,
ACCP study guides and, of course, pulmccm.org. I
cannot guarantee their correctness in content,
grammar, and spelling; nor are these notes peerreviewed. They absolutely should not be used as a
resource for patient care. They should be used for
board exam preparation; that is all I will use them for
as well. This is not my heart-lung physiology text.
Please share these review notes freely. #FOAMcc
Happy
Studies,
Jon-Emile
lung.org
JE
K
Critical Care Medicine

Review Notes
03
23
48
64
74
89
101
110
124

Notes 2014
h
3
The quality of CPR is associated with

survival to discharge for in-house
arrest. It is not the application of an
AED, which has been studied. 40% of
chest compressions in house are of
insufficient depth, there is a long period
of time during codes when health care
providers are not actually on the chest.
Compressions should be hard and fast
[to Bee Gees stayn alive] 100 per
minute, 2 inch depth. There should be
minimization of time off the chest even
for pulse checks and defibrillation. It is
CPR that matters, 2 minutes of CPR
interspersed with defibrillation, 1 mg
epinephrine every 3-5 minutes or 40 U
of vasopressin. Atropine is no longer a
part of PEA.
Avoid hyperventilation as this can
reduce cerebral perfusion via alkalemia
and more importantly cause dynamic
hyperinflation.
Previously, the brain injury [apoptosis,
excitability, edema, etc.] induced by
ischemia reperfusion was thought to
be treatable. The Feb. 2002 NEJM
article with 137 patients in each arm,
enrolled witnessed, shockable cardiac
arrest patient s with ROSC but still not
following commands. Patients were
randomized to therapeutic
hypothermia target 32 to 34 celsius
over 8 hours, held at that temp for 24
hours and then passively re-warmed
over 8 hours. A meta- analysis of the
three largest trials [380 patients]
showed an odds ratio of 1.7 for
favourable neurological outcome. The
largest of the three trials selected
patients who were resuscitated within
60 minutes with a shockable rhythm,
comatose, and not in shock. Cooling was

done within 2 hours of the event.
Therapeutic hypothermia [TH] is commonly
associated with coagulopathy,
hyperglycemia, bradycardia and
hypovolemia the latter on
1. CRITICAL CARE
CARDIOLOGY
Notes 2014 RESUSCITATION
CARDIAC
account of the cold diuresis. There
is an increased risk of infection
following TH and altered drug
metabolism.
However, skeptics noted the lack
of blinding to treatment allocation
in the above studies.
Further, in the biggest of those
hypothermia studies, a large
number of patients in the usual
care group developed fever which
is associated with worse outcomes
after cardiac arrest. So it was
thought that perhaps TH simply
had to accomplish fever reduction
to improve outcome.
So the big one came out in NEJM
November 17, 2013 [950 patients
in 3 years, 80% vfib, 12% asystole,
8% PEA, and randomized the
patients to celsius 33 or 36 ASAP
for 28 hours and then fever
reduction for 72 hours]. After 72
hours, a neurologist blinded to
initial treatment allocation
recommended withdrawal or
In a related vein, in 2013 JAMA, 100 patients
with bacterial meningitis were randomized
to 32-34 degrees or standard care and this
trial was stopped early for a 20% absolute
risk increase in mortality!
What about the prognosis following cardiac
arrest? What is the false positive rate for
diagnosis of poor neurological outcome? We
should strive for zero. Pupillary reaction to
light has a false positive rate of 0-31% at
day one. At day three, no pupillary reaction
has a false positive rate for poor outcome of
zero. So absent pupillary response at day
three is important information. The absence
of corneal reflex is similar at 72 hours.
What about posturing?
Similar at 72 hours. All of this data was
generated during an era of no therapeutic
hypothermia and the reason why
h
continued care based on standardized
criteria, with withdrawal recommended
only for known predictors of a terrible
outcome [e.g., refractory status
epilepticus; Glasgow motor score 1-2
with bilateral absence of N20 peak on
median nerve SSEP]. CPC more than 2
or Rankin more than 3 was defined as
severe disability. There was no
difference in death or disability
between the two groups [i.e. 33 versus
36 degrees] even in the 80% shockable
group.
Temperature should be maintained at

36 C or below after out-of-hospital
cardiac arrest.
Despite its physiologic rationale and
evidence of benefit in prior smaller
studies, targeted temperature
management below 36 probably does
not improve outcomes after out-ofhospital cardiac arrest of any type.
Because average human core
temperature is 37 C, maintaining
temperature continuously at or below
36 C still would require cooling in
almost all patients.
neurological assessment for continued care in the
aforementioned trial took place at 72 hours.
So what has a perfect specificity for determining poor
neurological outcome? There is a review in the Lancet
day 3 absent motor, absent pupillary response and
abnormalities in somatosensory evoked potentials. SSEP
occurs when the median nerve is stimulated. Bilateral
absence is poor neurological prognosis from NEJM article
in 2009. Day 1-3 there is nothing great to help. You must
wait at least 72 hours and you must wait for analgesia to
wear off. This is important with TH because there is
alteration of sedation and analgesia metabolism with
hypothermia it will stay longer. There is no clear answer
as to when the patient may wake up, but anecdotally up
to one week as the effects of hypothermia wear off and
sedation is cleared, though this will be less important as
TH is used less with the results of the Nov. 2013 NEJM
trial.

Notes 2014
Prognosis after a first cardiac arrest
occurring in the ICU is poor. An
observational study of almost 50,000 such
arrests showed an overall survival to
hospital discharge of 16%. Requiring
vasopressors
prior to the arrest was a major discriminator

in outcome:
Only 10% of patients on pressors prior to
arrest survived to discharge, and only 4%
overall were discharged home [the others
went to rehab or long-term acute care].
Among those with PEA/asystole despite
pressors, only 1.7% were able to perform
their own activities of daily living at the time
of discharge.
People with ventricular fibrillation or
tachycardia not requiring pressors prior to
arrest did much better: 40% survived, 20%
went home, and
17% had good neuro outcomes.
CONGESTIVE HEART FAILURE
About 50% of the mortality with heart failure

is sudden, the other half is slow progression.
There are stages of heart failure described in
2001 [ACC/AHA] where stage A is those
patients at risk, B [structural disease without
symptoms] is NYHA I, stage C [current or
prior symptoms] corresponds to NYHA II and
III and stage D is NYHA IV.
NYHA Class IV heart failure has a 60%
mortality at one year. In acute, severe heart
failure a poor prognosticator is: hypotension.
Hypernatremia and polycythemia are not.
Hypotension is defined as a systolic below
115 mmHg. Other bad prognoses:
hyponatremia, CAD, high BUN or creatinine,
low EF, elevated BNP or troponin, anemia,
diabetes.
Precipitants of ADHF? 25% excess salt and
fluid intake, 25% noncompliance with
medications, and 16% from adverse
medication effects [e.g. new CCB, NSAID,
steroids, glitazones and ethanol, new antiarrhythmics]. The rest are acute medical
causes such as ischemia, PE, HTN, valve
dysfunction, arrhythmia. There are extracardiac causes as well sepsis, infection,
renal failure, thyroid, new anemia.
Critical Care Board Study Notes 2014

FLAVORS OF MYOCARDIAL DYSFUNCTION
There are different kinds of myocardial
dysfunction to consider hibernating,
stunned myocardium, sepsis,
myocarditis, etc. Stunned myocardium
occurs when there is ischemia that
relates to a wall motion abnormality,
but when blood flow resumes, there is
a persistent WMA that lasts from hours
to days. The patient must be supported
during these time consider stress
cardiomyopathy as an example of
stunned myocardium. The patient may
present with all the signs and
symptoms of ischemia and heart
failure with stress cardiomyopathy.
There must be no coronary stenosis to
explain the disease and WMA do not
correspond to single coronary
distribution. There may be STE and big
T wave inversions.
bypass surgery [versus medical therapy] and there

was improved death from cardiac causes,
hospitalization [not overall death] with CABG.
Based on CAST and STITCH, improving hibernating
myocardium by revascularization can improve
NYHA class,
Recognize Takotsubo cardiomyopathy

the major complications of this are
mechanical in nature, heart failure,
shock and arrhythmia. It presents
exactly like a STEMI, or at least it can,
but angiogram is normal and there is
apical hypokinesis on the LV gram, this
invariably resolves by 2 months. Some
argue against the use of inotropes or
catecholamines in these patients as
they catechols be the cause; they
argue that IABPs should be used. This
disease usually happens in postmenopausal women and has a
mortality rate of 0-8% in house.
Hibernating myocardium is due to
chronic ischemia without infarction.
There will be a WMA that lowers
ejection fraction. If the occlusion is
reversed, the ejection fraction will
return to normal. What kind of
revascularization? In the CAST trial,
those with a low ejection fraction who
got CABG did better. The STITCH April
2011 NEJM looked at chronic heart
failure patients [EF less than 35%] with
www.heartJE
lung.org
K

Review Notes
of symptoms high dose is double their

outpatient dose. There was no
difference in creatinine, readmission, no
cardiac performance change, there is
no survival benefit shown.
exercise tolerance, cardiac performance
What about vasodilators? V-HEFT

trial was original hydralazine with
isosorbide versus prazosin versus
placebo. The hydral-nitrate group
had a mortality benefit.
and
maybe survival.
Nesiritide is a formulation of BNP which

was initially derived from pig brain. It
antagonizes the renin-angiotensin axis
and comes from the ventricles of the
heart it is released in response to high
ventricular volume and pressure. It
causes diuresis & vasodilation. It causes
a drop in filling pressure, blood pressure
and increases cardiac output. It lowers
filling pressure compared to
nitroglycerine and diuretics. But
nesiritide doesnt significantly improve
clinical end-points in Oconnor et al.
NEJM 2011 it might improve dyspnea.
There is no effect on death [does not
increase mortality] and rehospitalization.
With myocarditis, patients

typically get better
themselves, not good data
but prednisone likely
improves EF by 5% but only
if there is cellular infiltrate
on biopsy. There is
improved exercise tolerance
and cardiac performance.
TREATMENT MODALITIES FOR HEART FAILURE
Treatment should improve
quality & quantity of life;
some meds do one or the
other or both.
What about diuretics? There
is no difference in
continuous infusion versus
bolus - Felker et al in NEJM
2011 [DOSE trial]. However
high dose seemed to
improve global assessment
lung.org
ACE inhibitors do everything!

Consensus in NEJM 1987 was quite
impressive in terms of mortality benefit
for severe, symptomatic systolic heart
failure. Angiotensin II blockers are
likely as good.
JE
K

Review Notes

Notes 2014
The addition of ARB to ACEI may
improve in some
patients [CHARM trial].
NEJM - RALES in 1999 improved
mortality in severe heart failure with
spironolactone, but overall few of the
patients in RALES were on chronic
beta-blockers as it was not an
entirely accepted therapy at the time
of enrollment for that trial. Juurlink
NEJM -2004 ICES database review
showed excess hyperkalemia
hospitalization and deaths after
RALES. Eplerinone [less
gynecomastia] can reduce mortality
following MI with LV dysfunction, in
systolic heart failure with mild
symptoms EMPHESIS-HF also reduced
mortality [NEJM 2003], and there is
an increase in potassium as well. It
improves cardiac performance.
What about inotropes? In the 1980s,
patients who had the highest
norepinephrine levels did the worst.
Chronic use of inotropes will worsen
mortality dose-dependent decrease in
survival. They should only be used
when nothing else works.
Both US carvedilol [NEJM 1996] and the
MERIT- HF [lancet 1999] trial showed
survival benefit with beta-blockers in
low EF [carvedilol and XL- metoprolol,
respectively] and Copernicus in 2001
[very low EF patients getting
carvedilol].
Digoxin reduces hospitalizations,
reduces
symptoms in the Digitalis investigation
group [DIG
- NEJM 1997].
What about anti-arrhythmic therapy?
The MUST trial NEJM 1999 showed
benefit with defibrillator
addition of medical anti-arrhythmia
worsens mortality [similar to afib trials
RACE and AFFIRM]. SCD-HEFT showed
h
8
that amiodarone does not increase mortality
compared to placebo, but ICD improves survival for
sure especially those with LVEF less than 35%
[also MADDIT 1 &2 trials].
If there is LBBB and QRS more than 120 ms, and
depressed LVEF with symptoms, the patient can
begin CRT therapy BiV pacing to try to improve

Notes 2014
electromechanical
association. The
COMPANION
trial in NEJM showed that
CRT in patients with NYHA
III-IV HF with an EF of less
than 35% and a wide QRS
improved a combined endpoint of mortality and
hospitalization compared
to standard care. All-cause
mortality by itself was
only lowered in the CRT
plus AICD group. CRT does
not increase myocardial
oxygen demand and is
associated with reverse
remodeling of the
myocardium with an
improved LVEF. CRT also:
improves stroke volume,
dyspnea [functional class],
mitral regurgitation, and
lowers Ppao.
How about VADs? Eric Rose
et al. NEJM 2001 [REMATCH
trial] in late class IV heart
failure showed improved
survival [ARR 25% in death]
using pulsatile flow or
pusher plate. Continuous
shoveling snow, earthquakes, bad
traffic,
watching stressful sporting events [e.g.
the Leafs or the Canucks], respiratory
tract infections, waking up in the
morning. There is evidence to suggest
that the influenza vaccine reduces MI
risk! So vaccinate patients with
coronary disease.
ECG AND STEMI
There may be profound, dynamic ECG
changes in a matter of minutes from
t wave inversions to STE. With
confirmed MI [NSTEMI or STEMI], up to
one third did not have chest pain.
flow VAD, however, is better per

Slaugther 2009 NEJM.
Heart transplant? Hearts are limited

resources. The one year survival is 8590%, 10 year survival is about 50% for
heart transplant. Many get to NYHA I
and II with transplant.
The ESCAPE study looked at the use of
pulmonary artery catheters in heart
failure, there is no mortality benefit, but
perhaps better hemodynamics.
ST-ELEVATION MYOCARDIAL INFARCTI
With STEMI, about 50% of people just

drop dead, without crushing chest pain
preceding the event. It is common for
there to be minimal or no coronary
stenosis seen on an angiogram prior to
there being ACS. Yet, there is abnormal
flow on angiography in more than 90%
of patients with STEMI [60% with TIMI
zero or total occlusion, 13% with TIMI 1
or penetration but no distal perfusion,
and 19% with TIMI 2 flow or complete
penetration, but delayed perfusion].
What are the clinical events associated

with plaque rupture? Clearly many, but
documented:
When there is no chest pain, patients are treated
differently and this may triple their mortality.
A 12 lead ECG must be performed immediately in
the ED that is within 10 mins, this is class IB
because the artery must be open within 90 mins.
[STREAM NEJM 2013 showed that a strategy
involving early fibrinolysis with bonus tenecteplase
and contemporary anti-thrombotic therapy offers
similar efficacy as primary PCI in patients with
STEMI who present within 3 hours of symptom
onset and who could not undergo primary PCI
within 1 hour of first medical contact].
If the first ECG is non-diagnostic and suspicion is
high, repeat ECG should be performed within 1530 minutes and this too is class IB evidence.

Notes 2014
STE has a differential diagnosis; LVH,
hyperK+, Brugada syndrome can all
present with STE. Q waves are present
within 6 hours in about 50% of STEMI
patients, and this is a poor prognosis.
29% of trans-mural MIs did not have Q
wave, 28% of sub-endocardial MIs [nontransmural] MIs did have Q waves, so it
is neither sensitive nor specific for
trans-murality [based on MRI study in
2004], but the probability of Q waves
does increase as the MI size and
number of trans-mural segments
increased.
Recognize that characteristic
symptoms for coronary occlusion plus a
new LBBB pattern means that a patient
should receive lytics or go to
10
the cath lab emergently. The LBBB is known to
obscure the normal patterns of
ischemia on the ECG and LBBB patients
tend to have a worse prognosis during
coronary ischemia, so there is more to
be gained by revascularization.
Wenkebach is not a classic finding in
coronary ischemia, but rather
pathological changes in the AV node
[though it can accompany right
coronary ischemia, this should also
present with ST changes in the inferior
leads]. Complete heart block is less
commonly a result of coronary
ischemia. Long runs of Vtach will
certainly obscure STEMI, but if the runs
are short, then the interceding ECG
should still detect current of injury.
Recognize cerebral T waves on an
ECG. Consider a patient with a
hypertensive internal capsule
hemorrhage with an ECG with deep,
inverted T waves in the precordial
leads. These are known to occur
following head bleeds, and sometimes
ischemic strokes. They are thought to
be the result of massive sympathetic
discharge.
Classically, Wellens sign is biphasic T
waves in the precordium, though it can
present symmetrically.
Note that left circumflex occlusion is
commonly silent on a 12 lead ECG
[distal circumflex marginal artery]. This
may lead to rupture of the anterolateral
papillary muscle in days after the
infarction. The presence of a normal
ECG with pulmonary edema should
prompt evaluation for lateral coronary
disease. If a PAC were placed, one
would see tall V-waves. Importantly, on
the ddx of tall v waves is ventricular
septal rupture, but this would give a 5
mmHg step up of oxygen saturation on
the PAC [see below].
TREATMENT MODALITIES FOR STEMI

Notes 2014
Three doses of SL NTG of
0.4 mg may be given every
In the patient with STEMI, there should
be dual
anti-platelet therapy and heparin
begun. If thrombolysis is given,
standard of care is not to give GPIIBIIIA.
For STEMI, morphine is still IC for
analgesia, all NSAIDs should be stopped
at presentation IC. IIIC evidence says
that NSAIDs should not even be started
during hospitalization.
ISIS-2 trial 160 mg of ASA had a 23%
reduction in mortality. In patients with
STEMI, clopidogrel does reduce
mortality regardless of whether
reperfusion is received it should be
continued for 14 days. Unless there is
CABG planned, then clopidogrel should
be stopped 5 days prior to OR.
Anticoagulation should be used for at
least 48 hours, but up to 8 days even
with reperfusion therapy. The
guidelines favor LMWH compared to
UFH. In STEMI, for every 100 patients
treated with LMWH there are 15 less
non-fatal MI, 7 less urgent
revascularizations, 6 fewer deaths but
4 more non-fatal bleeds [no increase in
ICH].
Pooled analysis of ECG and fibrinolytic
therapy showed that per 1000 patients,
49 lives were saved if baseline ECG
showed LBBB, 37 saved with anterior
STE, 8 saved if inferior STE and 14
harmed/died if ST depression on
baseline ECG. Indication for fibrinolysis
is symptoms within 12 to 24 hours and
STE in 2 contiguous leads or new or
presumed new LBBB. Prior ICH, known
cerebral vascular lesions, ischemic
stroke within 3 months, suspected
aortic dissection, active bleeding
excluding menses, significant head or
facial trauma or neurosurgery within 3
months are the absolute
contraindications to fibrinolysis. The
h
11
5 minutes for a total of 3 doses after
which IV NTG should be considered.
relative contraindications are: systolic more than
180, diastolic 110, prolonged [10 minutes] CPR,
major surgery [including eye surgery] or internal
bleeding such as GI/GU bleed within one month,
non-compressible vascular punctures [e.g. recent
trans-venous pacemaker in subclavian vein],
pregnancy, and current use of anti-coagulants. If
you cant re-perfuse an artery within 90 minutes
with STEMI, you should give a fibrinolytic, there is a
direct correlation between

Notes 2014
the number of leads with
elevation and the
benefit of fibrinolytic, in
fact, and there is little
benefit if only 2-3 leads
have STE, but it is still
recommended.
What about GPIIBIIIA in
STEMI? It is a 2B
recommendation, the
benefit is uncertain based
on 2009 guidelines. It is
reasonable to start these
drugs at the time of PCI, so
this is really the call of the
interventionalist. Abciximab
is a long-acting GPIIbIIIA
inhibitor, so it should never
be given prior to potential
bypass surgery. Abiciximab
is not cleared by the
kidneys, but eptifibitide and
tirofiban are.
The mortality benefit of
primary PCI compared with
onsite fibrinolysis was
nullified when the time
delay to primary PCI was
more than 2 hours.
Fibrinolytic therapy in
STEMI should be infused
when an anticipated delay
NYHA 1 or 2 at one year. The minority
of patients
in the SHOCK trial presented with
hypotension to the hospital, but go on
to do so within 24 hours. Further, 25%
of patients from the SHOCK registry
lacked pulmonary congestion on
examination [recall, all of these
patients were in cardiogenic shock].
No trials support the use of IABP in
acute myocardial infarction with shock
[in fact, the IABP-SHOCK2 trial in NEJM
2012 refutes its value], though it is
often done. In the original SHOCK trial,
to performing primary PCI is more than

120 minutes of first medical contact
[FMC]; FMC defined as the time at
which the EMS provider arrives to the
patient.
12
If someone gets fibrinolysis, what is the

benefit for rescue PCI? This is the
TRANSFER-AMI trial and there seems to
be clinical benefit for transfer
immediately to cath lab following lysis
[up to 6 hours post lysis]. Those going
to a non-PCI facility should receive lysis
or immediate transfer for PCI, the
decision depends upon the mortality
risk, lysis risk, duration of symptoms on
presentation, time required for
transport. Those patients best suited
for transfer are: more than 4 hours after
symptoms, high risk lesions or features
[e.g. shock, see below], high bleeding
risk. For patients going for PCI, AC
should be initiated, do not use
fondaparenux as it can cause catheter
clotting.
Know the SHOCK trial [NEJM 1999]. In
patients presenting with STEMI and
cardiogenic shock, emergent
revascularization [either PCI or
emergent CABG] provided a mortality
benefit compared to thrombolysis. Of
those who did survive, the vast majority
[more than 80%] were
the patients in the medical arm could not be revascularized until 54 hours after the event. The 30
day mortality was no different between the two
arms [47 versus 56% mortality]! However, the 6
month mortality was significant [50 versus 63%]!
While there was mortality non- significant mortality
reduction at 30 days, in patients under 75 years, it
was significant at 30 days.
Based on the SHOCK trial, cardiogenic shock is an
AHA class I indication for emergent
revascularization. Nitrates, beta-blockers and ace
inhibitors are generally avoided in cardiogenic
shock because they can exacerbate hypotension.

Notes 2014
Oral BB should be started in the first
24 hours, unless there are concern for
shock, pulmonary edema, elderly,
more than 12 hours of symptoms,
bradycardia or heart block, reactive
airway disease. ACEI should be started
if LVEF less than 40% or with DM, CKD,
HTN.
RIGHT VENTRICULAR INFARCTION
Hypotension with acute inferior MI
should prompt right sided precordial
leads. RV infarction may show shock
with clear lungs, elevated venous
pressures, Kussmauls sign, and square
root sign on PAC [see below].
Look for tall R waves in the anterior precordial leads. The patient will present
with hypotension, bradycardia and an
absence of pulmonary congestion.
There will be STE in right sided leads,
depressed RV function on TTE. RV

infarction is
seen in 30% of patients with inferior MI
and is clinically significant in about
10% of those with inferior MI.
The treatment is maintaining RV
preload, inotropic support, reperfusion.
Diuretics should not be used. IABP is
an effective option in RV infarction.
Sometimes pacing to maintain AV
synchrony is needed, cardioversion for
SVTs.
Reperfusion typically results in RV
improvement in days. Nitroglycerin
should not be given to hypotensive
patients, or suspected RV infarction,
nor patients who have received PDEI in
24-48 hours. With administration of
lytics, the blood pressure remains low,
but the heart rate increases. Volume
infusion should be considered to
improve preload. Streptokinase itself
can cause some hypotension, but this
is much less likely with tPA.
MECHANICAL COMPLICATIONS OF STEMI
Recognize mechanical complications
post infarction. Pra, RV diastolic, PA
diastolic and pulmonary artery
occlusion pressure [Ppao] all elevate
and are within 5 mmHg of each other
with tamponade secondary to cardiac
rupture. Post-infarction rupture is
different from hemorrhagic pericarditis,
which may occur in the setting of
thrombolysis. Hemorrhagic pericarditis
is not the result of a ruptured muscular
wall. It can be treated with needle
decompression, but this often clogs,
and sometimes a surgical window is
required.
The most difficult ddx in the setting of
equalized cardiac pressures is that of
an RV infarction secondary to right
coronary occlusion which can also
raise Pra, Prv, and Ppao [via ventricular
13

Notes 2014
interdependence] but in
the setting of a known LAD
lesion, rupture would be
more likely.
Other considerations are: VSD which

would display an oxygen step up of 5
mmHg from the RA to the RV.
14

Notes 2014
Another option would be papillary
muscle
rupture of the posterolateral leaflet,
but this would be presented with a
murmur, crackles and tall V waves.
Consider this in a patient 48 hours post
inferior MI with sudden onset, bilateral
pulmonary edema with low output.
There may be no audible murmur. An
inferior infarction can cause rupture or
the posterolateral papillary muscle as it
has a single blood supply and is
susceptible to ischemia the most. The
murmur is often absent because of
equalization of pressures.
15
medications: LMWH and eptifibitide this increased
the risk of death.
What about anti-platelets? Those at medium to
high risk and in whom an initial invasive strategy is
chosen, the patient should receive ASA
In each of the mechanical

complications of MI, the mortality is
very high. Surgical mortality is about
50% for VSD [worse with inferior
ischemia] and papillary muscle rupture
and there are case reports of surgical
survival in patients with free wall
rupture. Medical mortality approaches
90% in all forms.
NON-ST SEGMENT ELEVATION MYOCARDIAL INFARCTION
Ischemic chest pain may involve an

occlusive thrombus or non-occlusive
thrombus and these clinical scenarios
can be detected by the ECG as STEMI
and NSTEMI, respectively. Unstable
angina is the term for ischemic chest
pain without ECG change and no
biomarker elevation.
Consider an NSTEMI patient on dialysis

with recent history of TIA/stroke all of
the following are contraindicated:
prasugrel, LMWH, eptifibitide and
fibrinolysis. Recall that with ST segment
depression there is harm with the
administration of fibrinolysis. There is
an increased risk of ICH absolute of
2.3% in those with a history of stroke or
TIA. Too many patients in the US on
dialysis received the contraindicated

Notes 2014
immediately. For ASA, the
dose doesnt seem to
matter, but less is more in
terms of bleeding. Plavix
reduces the risk of
composite end point by
2.1% in the CURE trial
NEJM 2001. Whether or not
patients will not receive a
cath, Plavix should be
started ASAP [600 mg per
CURRENT OASIS 7 NEJM].
Recall that some PPIs are
metabolized by CYP2C19,
omeprazole and Plavix
moves through that CYP
the data about this is
totally mixed. It is probably
OK to use clopidogrel and
PPIs concomitantly, but
often patients dont really
need the PPI.
The choice of the second
anti-platelet agent to be
added to ASA includes: as
above, clopidogreal before
or at time of PCI, an IV
GPIIBIIIA before or at time
of PCI, or prasugrel at the
time of PCI. Fondaparinux
should not be used if going
to the cath lab because of
catheter thrombosis.
strongly considered for cardiac
catheterization.
This is essentially a risk stratification; an
LV gram should be considered in all
NSTEMI patients.
What is best in NSTEMI, conservative or
invasive management? There is
evidence to suggest improvement with
an invasive approach if there is baseline
troponin elevation or ST segment
change. A meta-analysis showed that
there was no difference between
invasive and conservative strategies in
low risk patients. What about early
versus late invasive intervention in ACS
h
16
Anti-coagulation should be added ASAP
with presentation. When invasive
strategy is chosen, the patient should
get either LMWH, UFH or bivalirudin. In
a patient with renal failure, or CrCl less
than 30 ml/min, fondaparinux is contraindicated, LMWH can be used, but it
must be dose-adjusted. UFH may be
used. The general dose of UFH for ACS
is a loading dose of 60 U/kg bolus and
maintenance of 12/Kg/hr for PTT of 5070. Fondaparinux is 2.5 mg SQ once,
but cannot be used if CrCl is less than
30. It should not be given if going to the
cath lab. Bivalirudin is OK with a history
of HIT.
What about CCB? There is old evidence
from the late 1980s, in patients with
pulmonary edema by CXR or exam,
diltiazem can worsen outcome.
But, a CCB can be used over a BB if the
patient has a normal LVEF and is not in
failure as a means to minimize
myocardial oxygen demand.
An ACEI should be given within 24
hours in NSTEMI in patients with
pulmonary congestion or depressed
LVEF unless there is hypotension or
contraindication to that class of
medication.
Patients with NSTEMI and a low EF
should be
the TIMI risk score and GRACE risk score were used
and again, those at high risk benefitted from early
intervention, not low to moderate. This is Mehta
NEJM 2009. If there is recurrent angina, elevated
troponins, new ST depressions, HF, instability, prior
CABG, PCI within 6 months, high TIMI or GRACE
score, reduced LVEF the patient should receive an
early invasive approach.
Recognize and treat Prinzmetal or variant angina.
The classic picture of this is abrupt onset [at rest]
chest pain with inferior ST elevations that completely
resolve with resolution of chest pain. It is typically in
leads II, III and avF and PVCs are common. Variant
angina may produce all of the complications of

Notes 2014
plaque rupture. The gold standard for
diagnosis, really, is coronary angiogram,
but even patients with variant angina
may have pre-morbid, severe coronary
obstruction. The treatment of this
disease involves selective coronary
vasodilators [nitrates and calcium
channel blockers]. It is argued that
patients should not get beta-receptor
selective beta-blockers because this
may lead to unopposed alpha
stimulation which might make things
worse. Variant angina may cause
bradycardia if the right coronary is
involved [from AV node ischemia].
Again, the clue is the ECG with transient
ST elevation without infarction.
On angiography, ergonivine [which
stimulates both alpha receptors and
serotonin receptors] may be infused at
very low dose, but some clinicians think
that this is dangerous.
BRADYCARDIAS & PACEMAKERS

17
The sinus node is spontaneous. The AV
node introduces AV delay to allow
mechanical coupling of atrial systole and
ventricular systole this is the PR
interval. The AV node is also protective,
if atrial fibrillation develops, it prevents
ventricular fibrillation.
The His-Purkinje, by contrast, is very,
very rapid this gives rise to
synchronous ventricular contraction.
There are subsidiary pacemakers down
the conduction system which can give
rise to escape beats which can be
narrow complex if junctional or above AV
node, but become progressively slower
down the pathway.
Junctional escape must be clarified as
there are different implications e.g.
junctional escape with sinus arrest
versus junctional escape with complete
heart block.
Describe first what the underlying
bradycardia is. Sinus bradycardia is
common and usually benign
at night, during sleep there can be
very long pauses in young healthy
athletes.
It is common to see comorbid atrial
fibrillation and sinus node disease
[tachy-brady] as the pathophysiology for
the two is essentially the same, also
medications used can do this.
Following conversion from atrial
fibrillation there is quite commonly a
long pause/asystole before the sinus
node recovers. The pause may be
alarmingly long, but these patients often
do well without a pacemaker.
SECOND DEGREE HEART BLOCK
A prolonged PR interval is much less
pathological compared to a patient with
pathological AV block. Type I AV block
tends to occur in the AV node; there is
intermittent failure of the AV conduction

Notes 2014
system. There is
progressive PR
prolongation prior to
progressive, usually does not need a
PPM and
common causes include drugs, vagal
tone, inferior infarction; it resolves with
atropine and exercise.
Type II AV block is due to block in the
His-Purkinje system and is characterized
by a stable PR prolongation prior to AV
block; this is typically the result of block
below the AV node, tends to be
progressive and frequently needs or
should require a PPM. It is often caused
by conduction disease, anterior
infarction and worsens with atropine and
exercise.
Recognize Mobitz II heart block
following an anterior AMI. A patient will
have bradycardia, normal blood
pressure but chest pain. There is
grouped beating. This is Mobitz II and
requires a PPM. Adams-Stokes attacks
can commonly occur with Mobitz II
block as it frequently progresses to high
degree AV block with syncope. By
contrast, Mobitz I may complicate
inferior MI, but typically does not
progress to high degree block and
resolves with atropine. Type II usually
occurs in the setting of an anterior
infarction and is commonly associated
with permanent pacemaker
dependence and pump failure.
Anatomically, type I [mobitz] is an AV
node phenomenon whereas type II is
the result of destruction to the HisPurkinje system. This is also why
atropine is not effective in speeding up
the heart rate.
Always assess the sinus rate if slow
at the time of AV block, it suggests that
there may be high vagal tone. There
will be a slowing of the sinus node and
then block. That may be caused by
pain or suctioning.
h
18
complete block. When there is grouped
beating of the RR intervals on the ECG,
think of Wenckebach. It is usually not
THIRD DEGREE AV BLOCK
Complete heart block is more severe His-Purkinje
disease and a result of structural heart disease.
Remember that AV dissociation is a generic term
it occurs with ventricular tachycardia so
describe the bradycardia first complete heart

Notes 2014
block with AV dissociation.
In complete heart
block there is no association
between atrial and
ventricular activity.
Consider a patient with first
degree AV block, LAFB, and
RBBB who then gets
complete heart block. There
is no slowing of the sinus
rate prior to block. This is a
serious situation.
But heart block, in reality, is
more muddy than the
aforementioned, that is, 2:1
AV block or high- grade AV
block. The level of block is
important.
2:1 AV block is a failure of AV
conduction with every other
beat, so PR prolongation is
difficult to detect on the ECG.
In this situation, clues for
Mobitz I versus II: If the QRS
is very wide, it is more likely
Mobitz II, check the rhythm
strip for times when
conduction is better to
determine if PR prolongation
occurs prior to block.
however, it is not associated with
oligoarticular
arthritis, nor cephalalgia to the extent
that Lyme does. Bullous myringitis and
erythema migrans can be seen with
mycoplasma perimyocarditis. Stage III
lyme disease [recurrent oligoarthritis]
is more difficult to eradicate than stage
II [even with meningeal involvement].
TREATMENT OF BRADYCARDIA AND HEART
BLOCK
ACLS if bradycardia, look for cause,
get IV access, determine symptoms.
First give 0.5 mg atropine, then consider
IV infusion of epinephrine or dopamine.
19
High grade AV block may also be called
intermittent third degree and occurs
when more than one P wave blocks in a
row. The level of block can be
determined as above. There will be
intermittent AV conduction in high
grade block.
The difference between high degree AV
block and complete heart block is that
the latter reveals a complete
dissociation between atrium and
ventricles, i.e. both p waves and QRS
complexes should march out perfectly
but in a mutually exclusive fashion. If
there is any irregularity to the
ventricular rhythm, then there is some
conduction, unless there are PVCs
during the complete block.
Recognize that Lyme myocarditis may
present with complete heart block. The
patient will also present with dyspnea,
Bells palsy and bad headaches. There
may not be a history of stage I Lyme
[erythema migrans], but there likely will
be arthritis. 4-8% of patients with Lyme
disease have some element of
myocardial involvement. The Ddx here
includes: mycoplasma perimyocarditis
which could certainly present like a
stage II Lyme disease involving the
heart,
tracking the atrial rhythm. Old pacers will revert
to VVI after a long period for example 8
years or so. This may precipitate heart
failure.
A pacing stimulus on a T wave can
rarely lead to ventricular fibrillation.
Making the pacer more sensitive [to
detect a smaller R wave] can prevent
this [so the PPM can recognize the R
SUPRAVENTRICULAR
TACHYCARDIA
wave].
Trans-cutaneous [TC] pacing is useful, but often not
tolerated. Must confirm capture mechanically [feel
the pulse, observe on art line/plethysmography]. TC
rate may be set at 60/min initially and dialed up if

Notes 2014
perfusion remains inadequate. The
usual initial dose is 2 mA above the
dose that induces "capture."
PPMs the first letter is the chamber
that is paced, the second letter is the
chamber sensed and the third is the
response to sensed event. The sinus
rate can be an indicator of the degree
of stress upon the patient, so if there is
block with high sinus rate, this is
concerning as the patient is having an
adrenergic response to this conduction
disease.
Placing a magnet over a dual chamber
PPM will cause it to become DOO, that
is, pace both chambers in sequence at
70 bpm, if its a single chamber pacer, a
magnet will make it VOO.
On the boards, after EP lab intervention
[e.g. PPM placement], especially with
hypotension & chest pain, its usually
perforation and tamponade until proven
otherwise.
As an aside, left atrial-esophageal
fistula can occur following an afib
ablation. There can be stroke-like
symptoms given food emboli and
bacteremia - risk is less than 1 per
1000. Fever post device could be a
sponge in the pocket. If someone has
a dual chamber pacer, modern pacers
can detect atrial fibrillation and stop
SVT is usually 100-260 bpm. It can be

narrow or
wide depending on underlying bundle
disease, rate-dependent aberrations
or pre-excitation [e.g. WPW].
PSVT means sudden onset, sudden
offset. The primary re-entrant
mechanisms of PSVT are AVNRT [65%],
PATHOPHYSIOLOGY
20
There are 3 mechanisms of tachyarrhythmias for both the atria and

ventricles. 1. Re-entry [most common]
for example scar-based monomorphic
VT post-MI, or atrial flutter. 2. Abnormal
automaticity an example would be
sinus tachycardia or MAT and 3.
Triggered activity which is frequently a
drug-effect for example, torsades de
pointe, digoxin toxicity. Triggered
activity occurs from a DADs or EADs
[delayed or early after-depolarizations].
The most common is re-entry, either
macro re- entry or micro-re-entry.
Monomorphic Vtach is a macro re-entry
typically around a scar. What is reentry? Most of the heart is
homogeneous, with similar conduction
velocities and refractory periods. But
the introduction of a scar can increase
heterogeneity or dispersion of
depolarization and repolarization within
the myocardium - producing areas of
rapid and slow conduction. If a premature beat blocks in one limb and reenters through the slow pathway that is
now repolarized then re-entry has
begun.
Atrial flutter is a macro-re-entry about
the tricuspid valve. AV re-entry can
occur about the AV node. Vtach is about
a scar. The drugs given tend to change
the refractory periods. The shock
applied during a defibrillation causes a
synchronous depolarization at once to
eliminate the re-entrant circuit.
SUPRAVENTRICULAR TACHYCARDIA
AVRT [30% - about half of these have pre-excitation
like WPW, the other half have a concealed pathway
found on EP study] and lastly about 5% have AT.
Thus, the AV node is a critical part of the re-entry in
AVNRT and AVRT, but not AT. Blocking the AV node
[e.g. adenosine] with AT will not terminate the
rhythm.

Notes 2014
In AVNRT, the p wave is usually within
the QRS or slightly thereafter. In AVRT,
the p wave typically comes after the
QRS as they are depolarized in series.
WPW describes a patient with both preexcitation on ECG and symptoms. This is
the difference between pattern and
syndrome.
There is a short PR with delta wave
which is the slurred upstroke of the QRS
[the pattern].
The acute treatment of these rhythms
depends on symptoms. Adenosine is
great for AVNRT, AVRT, 6-18 mg bolus is
given rapidly. If a patient is stable with
termination of PSVT with 6 of adenosine,
but then back in one hour, what is the
next step? A higher dose of adenosine
will not lead to prevention of recurrence,
the best next step would be to use a
drug that blocks the AV node over a
longer period of time such as a CCB or
BB.
Recognize that in a patient on a methylxanthine [e.g. aminophylline]
adenosine will not be effective because
of antagonism effects. For example, a
patient admitted for an asthma
exacerbation is treated with
aminophylline and develops what
appears to be an orthodromic SVT. The
correct answer is to give verapamil [not
adenosine] because the patient is on a
methylxanthine. Beta-blockers are not
as effective at treating SVT as CCB and
adenosine are. BB should generally be
avoided in acute
exacerbations of OVD. Lower doses of

adenosine
should be considered for patients
who: are on dipyridamole, have a
heart transplant or who receive the
drug through a central line
21
MULTIFOCAL ATRIAL TACHYCARDIA

Treat multifocal atrial tachycardia.
Consider a patient with pneumonia, a
COPD exacerbation [treated with a
methylxanthine], and rapid MAT.
Treatment is with amiodarone. The
patient has a number of Ashmanns
beats in his ECG. MAT is associated with
theophylline use, but also more
prominently, pulmonary disease. The
treatment of MAT is to treat the cause
[stop theophylline, treat respiratory
disease]; if treating the PNA and COPD
does not slow the rate, you should
medically treat the MAT.
Amiodarone is a type III anti-arrhythmic
with modest beta-blocking properties. It
has been shown to slow and convert
MAT. DC cardioversion, by contrast, is
most effective for re-entry type
arrhythmias such as atrial fibrillation
and atrial tachycardia and most types
of ventricular tachycardia. MAT is an
automatic arrhythmia, so DC
cardioversion will not be as effective.
Similarly, MAT will not respond to
digitalis as it is an automatic
arrhythmia.
Lidocaine tends to suppress only
ventricular arrhythmias, so it would not
be as effective. Calcium channel
blockers could be considered, but they
will likely cause hypotension more than
amiodarone.
ATRIAL FLUTTER AND FIBRILLATION
What about atrial flutter? It is almost
always a right atrial rhythm. Typical
flutter has a constant flutter wave
morphology in the inferior leads.

Notes 2014
The mechanism is a single
re-entrant mechanism.
Right atrial stretch/disease
can lead to flutter. It is easy
to ablate. It passes through
left atrial flutters. The crista terminalis
is a
natural barrier between smooth and
trabeculated portions of the right
atrium. Upright in V1, down in lead II is
typical for classic flutter. In V1, large
amplitude fibrillation waves can be
confused with flutter.
Atrial fibrillation or a disorganized atrial
activity is the most common arrhythmia
is the world.
Paroxysmal = spontaneous conversions,
persistent means that there is some
attempt to convert the patient and
permanent means that there have been
failed attempts at conversion. There can
be asystole from adenosine [rare] there
can be pause-dependent polymorphic
VT from adenosine [rare], so try to avoid
adenosine if you are sure that the
rhythm is atrial fibrillation. Atrial
fibrillation and flutter themselves
uncommonly cause shock, it is usually
some other co-morbid disease that is
the reason for the shock such as MI,
tamponade or sepsis.
The focus in the ICU should be
immediate rate control as well as AC!
Digoxin is less effective in states of high
sympathetic tone, but when given in
high doses, digoxin can be helpful.
Cardioversion must be synchronized. A
shock on the upslope of the T wave is
the most sensitive time of the cardiac
cycle to cause a ventricular arrhythmia.
Synchronization times the shock with
the QRS which is the portion of the
cardiac cycle most likely to correct the
rhythm. Sometimes the defibrillator
auto-corrects to unsynchronized
cardioversion following an initial shock.
h
22
an area near the IVC and tricuspid valve
or isthmus-dependent atrial flutter, type
I flutter. But there is also scar- based
atrial flutter, valve surgeries can result
in
Medical conversion is best carried out by ibutilide
[60% effective] but when a patient is unstable, the
correct answer is electricity. IV ibutilide can be
effective [1 mg over 10 mins, followed by 1 mg
thereafter it can cause QT prolongation].
Ibutilide is also helpful immediately after an
electrical cardioversion, or if too much electricity is
required to cardiovert, but this is less problematic
with biphasic cardioverters.

Notes 2014
Paroxysmal atrial
tachycardia with block is
classic
for digitalis toxicity.
Dronedarone is known to
cause liver injury [Multaq is
dronedarone], transplant
has been required in this
situation, but it is very rare.
Get baseline and repeat
liver enzymes. Flecainide is
usually associated with HA,
neuro, ventricular proarrhythmia, propafenone
can cause metallic taste,
both are IC, sodium channel
blockers. Defetilide is class
III, can prolong QT. Sotalol
is partially a beta- blocker.
ANTICOAGULATION
23
When using a drug or shock to

cardiovert a patient, there is a high risk
of stroke; it is not caused by the shock,
it is caused by the conversion to sinus
rhythm. If the AF is less than 48 hours in
duration, cardioversion is OK. After a
couple of days, then you need to think
about anticoagulation. If longer than 48
hours, then 3 weeks OAC pre
cardioversion, then 4 weeks post. If the
patient has a TEE and there is no clot,
the patient still needs 4 weeks of OAC
post cardioversion as it is the stunned
myocardium that increases the risk of
thromboembolism. If the patient cannot
receive AC, there is no point to getting
TEE.
AF with WPW is badness,

the risk of death chronically
is 0.1% per year. AF with
WPW, if hemodynamically
unstable is treated with
synchronized shock, IV
procainamide is the sodium
channel blocker of choice.
Could consider Ibutilide or
amiodarone, but there is no
data there. Never give CCB
or BB.
ventricles and the dysrhythmia may be
monomorphic or polymorphic.
In a low CVA risk patient with AF, it is

reasonable to hold AC for one week
without heparin.
Monomorphic is commonly scar-based

VT, RV outflow tract [RVOT-VT] can occur
idiopathically.
Wide complex QRS tachycardia has a differential.

Obviously it can be VT, but also SVT with aberrancy
[rate-dependent aberration, or pre- existing bundle],
it could be antidromic AV re- entry, ventricular
pacing [if the pacing stimulus is small and not seen
on the monitor], drug or electrolytes, artifact, or
even ST elevation if seen in only one lead. Patient
movement, infusion pumps, loose electrodes, etc.
can all look like Vtach. Following Vtach, there tends
to be a pause, if there is no pause, it may be
artifact.
Usually with artifact, there are QRS complexes
marching through the artifactual waves.
Polymorphic VT can look like Vfib, but

polymorphic VT displays gradual change
in the size of the QRS. If it turns about a
point and, technically, if the patient has
a long QT at baseline, this is known as
Torsades de Pointe, as this is a clinical
syndrome.
If a patient has VT more than 30
seconds it is sustained or, if there are
VENTRICULAR TACHYCARDIA
Ventricular tachycardia is at a rate of

100-120 bpm with three of more
consecutive beats at that rate. There is
a slow spread through the
symptoms despite being less than 30 seconds it is

also sustained in nature.

Notes 2014
By contrast, a long pause, followed by a
PAC, followed by a wide-complex
tachycardia is most likely Ashman
phenomenon. Recall that the right
bundle has a slightly longer refractory
period than the left. Further, the
refractory period of the bundle is
directly related to the length of the
preceding beat. So if there is a long
depolarization, there will be a longer
refractory period. If, during this longer
refractory period, premature
supraventricular activity occurs [e.g. a
PAC or an afib beat], then it will hit the
right bundle during its refractory period
and have right bundle configuration
[wide-complex morphology]. If this
occurs during rapid afib, this can look a
lot like V-tach. So if the patient has just
had a long pause, followed by a narrow
complex beat and then wide-complex

tachycardia
in a RBBB pattern, this is an excellent
way to distinguish V-tach from SVT
with aberrancy.
24
Differentiating VT from SVT with

aberrancy is also aided by considering
the clinical context, like if the patient
came in with prior MI, heart failure, has
an AICD, etc. worry more about VT. The
rate and hemodynamic stability can be
misleading; always treat as VT if
uncertain.
Ask yourself if it looks like RBBB or
LBBB? The best lead is V1. If looks like
typical RBB or LBB, its probably
aberration. A sharp R wave just prior to
the wide S wave is strongly suggestive
of LBBB pattern. If there is a notching
within the wide S wave, it is more
suggestive of VT.
The Brugada criteria first looks for
precordial concordance; if all the R
waves are in the same directions in V1V6, it strongly suggests a ventricular
source [VT] but if they are up in V1 and
down in V6, this suggests RBBB, or if
down in V1 and up in V6 - LBBB. If the
beginning of the R wave to the nadir of
the S wave is short, then likely PSVT, if
long, likely VT. AV dissociation is also
VT. But these rules are not perfect,
there can be pre-cordial discordance and
there still may be VT, especially if there
is AV dissociation or the presence of
fusion or capture beats.
In a wide complex tachycardia of
uncertain etiology, the use of
adenosine is reasonable to differentiate
the two [VT versus SVT with
aberrancy], if the patient is stable. Be
very wary of WPW with pre-excitation,
though.
As in the preceding section, the
treatment of WPW-associated atrial
fibrillation is procainamide. This will

Notes 2014
present as a young patient
with a heart rate above
200. The ECG may reveal
a wide complex with a
RBBB pattern with
precordial concordance. It
the refractory period of the accessory
pathway,
and therefore increase the ventricular
rate. Lidocaine has also been reported
to increase the ventricular rate in WPW.
Substitutes for procainamide here are
amiodarone, ibutilide and sotolol.
What about VF? The etiology is likely
multiple re- entrant wavelets within the
HYPERTENSIVE
EMERGENCY
ventricles, it is lifethreatening. The use
of drugs is for Vfib is amiodarone 300
mg or lidocaine 1 mg/kg, vaso 40 U, epi
1 mg. Lidocaine blocks both activated
and inactivated sodium channels, it
cannot be given orally, but mexilitine
can be. Lidocaine may be more
effective during active ischemia.
Amiodarone is a class III agent, it blocks
potassium and inactivated Na channels,
as well as alpha, beta and calcium
channels. It has iodine attached to it, so
can affect the thyroid, with prolonged
half-life. 200 mg per day is usually afib,
400 mg is for VT. Liver, thyroid, lung,
ophtho side-effects. Even though it
prolongs the QT, it rarely leads to
Torsades, probably because of its transmural stabilization effects.
Remember that lots of drugs prolong
QT and predispose to Torsades.
Methadone is one, many antibiotics,
antifungals, anti-psychotics. The
treatment of torsades is magnesium
and speed up the heart rate.
The treatment of long QT syndrome
either congenital [channelopathies,
usually potassium channels] or drug
effect is treated with IV magnesium.
Standard treatment with long QT is
beta-blocker potentially with AICD.
h
25
is unusually irregular, however. The
accessory pathway of WPW does not
have decremental conduction.
The treatment is to slow this pathway
with procainamide. Digoxin is
reported to decrease
The shocking lead of an AICD can be seen in the RV.
If a patient is having afib and getting multiple
shocks as a result. A magnet will suspend all
tachycardia therapies thus it will also prevent
shocking Vfib. If a magnet is placed over a
pacemaker it will change to DOO or VOO, that is, an
asynchronous pacing mode.

Notes 2014
Severe hypertension with
new or progressive
end-organ
dysfunction.
This requires immediate
reduction
in
blood
pressure. There is no
specific blood pressure
which
defines
an
emergency.
The brain, heart and kidney
should be evaluated
immediately check for
mental status, neuro
deficits, retinal changes,
consider CT scan and drug
screen.
Consider chest pain,
dyspnea, rales, S3 or pulse
deficits, a new murmur
[e.g. aortic dissection],
ECG, CXR, biomarkers and
TTE can all help triage.
For the kidneys evaluate
urine output, hematuria,
BUN/Cr, lytes, UA, CBC [e.g.
uncommon, but MAHA].
Consider drugs, pheo,
MAOI, clonidine withdrawal,
pre-clampsia.
vasodilator with minimal negative
inotropic
effects. It is hepatically cleared, so
watch for liver and older people.
Clevidipine is a very short- acting
arteriolar dilator. There is little change in
heart rate or cardiac output. It is
metabolized by tissue-esterases which
is nice though it is fairly expensive.
Fenoldopam is a dopamine 1 receptor
agonist which increases renal blood
flow, with no toxic metabolites; it can
lower potassium from increased renal
blood flow.
Esmolol is a cardioselective betablocker without vasodilating effects.
26
The treatment of hypertensive
emergency is to lower the MAP by 15[20]-25%. The systolic pressure can
vary quite a bit so dont rely on it that,
nor diastolic. The goal is to stop and
reverse end-organ injury while
simultaneously preventing iatrogenic
complications by correcting too rapidly.
Cerebral blood flow tends to remain
constant between a MAP of 50 to 150,
but in chronic hypertension, the curve is
shifted rightwards, so lowering MAP too
much can impair organ perfusion.
AGENTS
Which medication to use? Must consider
pre- existing conditions. Titratable,
potent and safe drugs that are
parenteral are ideal initially, then start
oral agents within 12-24 hours.
Nipride is an arterial and venous
vasodilator. It requires the ICU for
very close BP monitoring, 0.3-1.0
mcg/kg/min. There is cyanide toxicity.
Labetolol can be bolus or infusion. It is
mostly a beta-blocker, but some alpha
[6-7:1 is the ratio of beta-to-alpha
effects]. Nitroglycerin is a venodilator.
Over 100 mcg is more of an arterial
effect. Nicardipine is a systemic and
coronary
Hydralazine has a long duration and variable effect
there may be precipitous drops in BP there is also
reflex tachycardia; in many ICUs hydralazine is
falling out of favor for these reasons. Enalaprilat is
low potency with variable response in emergencies.
DISEASES
Hypertensive encephalopathy is treated with
nipride, labetolol, nicardipine, nitroglycerine,
clevidipine, fenoldopam. PRES may be seen in these
cases.
Nitroglycerine is usually the right answer for
hypertensive emergency with acute coronary
syndrome, may also consider labetolol, esmolol,
nicardipine, nipride. What about the treatment of

Notes 2014
acute heart failure, with systolic
dysfunction? NTG, nicardipine, labetolol
[with caution], nipride should be tried,
remember that diuresis may be touchy
is there is a small cavity from LVH, and
in many instances, these patients are
actually volume down in hypertensive
emergencies from a pressure
natriuresis.
What about an acute aortic dissection?
The correct answer is labetolol to a
target of 100/70. Nipride should never
be used alone during a dissection, there
must always be a beta-blocker to lower
the aortic dP/dt. Recognize and treat a
Stanford type-A thoracic aortic
dissection. The patient will present with
severe hypertension and chest pain. The
CT angiogram will demonstrate a
monster aorta with an intimal flap.
Notably 10-
27
20% of CXR with dissecting aortic aneurysms do
not have a widened appearance.
Surgical outcome is better than medical
therapy for Stanford type-A dissection,
whereas medical therapy is better for
Stanford B. Some of the latter patients
may need surgery if the diameter of the
aorta is excessive [more than 5 cm], if
there is ongoing ischemic symptoms or
if BP cannot be controlled.
What about head bleeds? See section
on hemorrhage in neurology section,
but briefly. INTERACT 1 trial lowered to
systolic of 140, and this lowered
hematoma growth [INTERACT 2 did not
find this result though]. The guidelines
are blurred, consider baseline BP, age,
ICP, cause of bleed. If ICP is high,
reduce the MAP, but keep CPP high
certainly above 55. Goal is MAP of 110
versus 130 mmHg if the ICP is high.
The agents often used are labetolol,
nicardipine, nipride, NTG, esmolol,
hydralazine, enalaprilat.
What about SAH? The patient will need
nimodipine and pain treatment. Shoot
for systolic less than 160 with betablockers or nicardipine IV. Ischemic
stroke, the systolic must be less than
185/110 if tPA to be administered. All
others, the BP is 220/120 goal. CHHIPS
study is looking to treat systolic more
than 160 in ischemic stroke.
Perioperative hypertension requires
pain relief, nitroglycerin, nicardipine,
clevidipine, labetolol, nipride.
Excess catecholamine states,
phentolamine is the boards answer. The
choice is always alpha blocker before
beta-blocker.
Hypertensive urgency is severe
hypertension, but no acute end-organ
injury. The blood pressure should be
lowered over 24 hours. Usually just give
them back their home meds PO no
need for ICU level care.
HEMODYNAMIC MONITORING

Notes 2014
Central pressure monitoring involves
central
venous pressure [CVP] and pulmonary
artery occlusion pressure [Ppao].
PRINCIPLES OF MEASUREMENT
The phlebostatic axis is an attempt to
find the left or right atrium the nipple
line at mid chest once zeroed, the
patient should not really be moved. The
system must be opened to atmosphere
at that point. The CVP is measured
following the a-wave which is the first
positive deflection following the ECG p
wave. For the CVP the a-wave usually
follows P wave but precedes the QRS.
For the Ppao, the a-wave follows the
QRS and the v-wave follows the ECG T
wave.
Find the a-wave at end-expiration. One
method is to take the top of the a-wave
and the bottom of the a-wave and
divide by 2. So if the top of the a-wave
is 12, and the bottom is 6, then the
average is 9. End-expiration is most
difficult to find when the patient is
triggering the ventilator. This typically
requires a bedside assessment while
looking at the patient and looking at the
CVP tracing and ventilator waveforms.
There are clear limitations to the use of
pressure as a surrogate for preload and
it really shouldnt be done. What about
echocardiographic measures? The LV
end-diastolic area [LVEDA] has been
used as a surrogate for preload and
fluid responsiveness, but is an equally
poor predictor.
What about continuous cardiac output?
There are pulse-contour analysis
devices such as PICCO, LiDCO, or
Flotrac. The former two use measures of
cardiac output [e.g. thermodilution or
lithium dilution] to calibrate while
flotrac does not they use the area
under the arterial pulse-pressure curve
as a surrogate for stroke volume.
h
28
Esophageal Doppler monitoring may also be used.
Systolic and pulse pressure variation may be used
as markers of fluid responsiveness but with a
plethora of caveats.

Notes 2014
PATTERN RECOGNITION FROM THE PULMONARY
ARTERY
CATHETER
Pulmonary artery occlusion
pressure is an [admittedly
poor] estimate of LV filling.
The pulmonary artery
catheter must be in West
Zone III physiology. Zone II
and I physiology will result
in faulty measurements. It
will measure airway
pressure in these states.
PEEP tends to decrease the
LV trans-mural pressure as a
consequence of decreased
LV filling
this occurs despite PEEP
increasing the absolute
value of the Ppao.
The PAD-PCWP gradient is
important. The normal
gradient should be less than
5 mmHg. If greater, it
suggests that there is a true
increase in pulmonary
vascular resistance. Use of
the PVR should be largely
abandoned [IMHO] in the
care of ICU patients.
29
Overwedging occurs when the balloon
or PAC tip is stuck up against the vessel
wall and the tracing slowly shoots up
over time. Pulmonary artery rupture
may also occur. There will be
hemoptysis and blush on angiogram.
The treatment is embolization. Risk
factors for rupture are long- term
steroids, PAH, older age, balloon
hyperinflation, cardiac manipulation,
hypothermia.
Thermodilution curves - if there is high
CO there is a very quick up and down
thermodilution curve the low temp
gets there faster and goes away faster,
the opposite is true for low CO [there is
an inverse relationship between the
thermodilution AUC and cardiac output].
The thermodilution cannot be reliably
used in patients with tricuspid
regurgitation. Further, in low output
states, the cold injection may be diluted
in the warm cardiac tissue and therefore
give a falsely small area under the curve
[and therefore high cardiac output].
Potential problems with thermodilution
measurement of cardiac output may
also result from technical factors [such
as variation in

Notes 2014
injectate temperature, volume, or rate
of
injection] or from physiologic factors
such as arrhythmias or respiratory
variation. Variability in calculations of
cardiac output by thermodilution is
estimated at approximately 10%; thus,
changes in cardiac output should
generally be on the order of 15% to be
regarded as valid.
May have a board question on a patient
with a recent MI with sudden worsening,
especially after thrombolysis with tall v
waves on the PA tracing. It may be
confused with the dicrotic notch of the
pulmonary artery tracing. Do not
confuse with a dampened waveform.
The cause is acute papillary muscle
rupture, the treatment is surgery with
balloon pump.
The 4th day following an acute MI,
there is hypotension with a loud
systolic murmur with a step-up in the
right ventricle, say from 60 to 70
mmHg. This is a VSD.
The square root sign occurs in the
setting of constrictive pericarditis and is
the result of the right ventricle filling
against a stiff pericardium [the plateau
portion of the square root sign following
the v wave].
Recognize electrical alternans on an
ECG and anticipate the pressure
waveform on a PAC. The PAC data will
have nearly equal Pra, Ppao and PAD.
The patient with tamponade will have a
dramatic x decent because of the
enhanced systolic filling of the right
atrium, and a diminished y decent
because of the lack of a pressure
differential between the RV and RA at
the onset of systole. The patient may
also have pulsus paradoxus.
Differentiate traumatic tricuspid valve
rupture from constrictive pericarditis on
ECG and RAP tracing. Consider a
2
h
patient in a car accident with hemopericardium that

was drained. One month later he presents with
dyspnea and LE edema.
The ECG reveals right atrial enlargement, and the
RAP tracing depicts tall v waves, with prominent x
and y descents; this is traumatic TR.

Notes 2014
Be able to calculate the
VO2 if given cardiac
output, hemoglobin, mixed
venous and arterial oxygen
saturations. O2ER =
VO2/DO2; and DO2 = Hb
[grams per LITRE
Canadian units] x 1.39 x
cardiac output. Thus VO2 =
O2ER [i.e. the difference
between arterial PaO2 and
mixed venous PvO2] x
[1.39Hb x CO]. If the
arterial PaO2 of 0.99, the
mixed venous of 0.50, the
Hb of 120 g/L and CO of 3.
The VO2 becomes 245.
In very simple terms,
generally: severe
hemorrhage results in low
filling pressures,
pneumococcal sepsis
results in low filling and
high index. AMI will result
in high wedge, low index
and low PAD-PCWP. RV
infarction will result in a
low index, normal wedge
and elevated CVP.
IABP WAVEFORMS
Understand IABP waveform
analysis. The ideal timing of
the IABP is to inflate at the onset of

diastole [the dichrotic notch on the
pressure waveform] around the ECG t
wave and then deflate right before the
following systole which, on the pressure
waveform, is right before systolic
upstroke or right before the QRS on the
ECG. Balloon inflation can be too late
[well after the dichrotic notch] which
limits the coronary perfusion and
peripheral perfusion effects, or it may
be too early [before the dichrotic notch
or before the QRS] which increases
afterload.
Conversely, the balloon may deflate too
early [well before the next upstroke in
arterial blood pressure or well before
the QRS] which will impair diastolic
perfusion. The balloon may also deflate
too late which occurs when it deflates
into the next systole [seen when the
balloon deflation occurs well past the
upstroke of the next beat and well past
its QRS]. In the latter situation, when
the balloon does deflate, there is a
diminutive arterial pulse with dichrotic
notch. This essentially afterloads the
heart [deflation too late is
physiologically akin to inflation too
early]. Importantly, to diagnose these
problems, you must change the IABP to
1:2 or 1:3 and analyze

Notes 2014
the native and augmented pressure
waveforms
with the ECG.
SHOCK BASICS
Shock is a profound and widespread

reduction in the effective delivery of
oxygen and other nutrients to tissues
which leads first to reversible and then,
if prolonged, irreversible cellular injury.
In all types of shock there is an
activation of the SIRS response. Some
of this may be a result of poor gut
perfusion and endotoxin release in all
forms of shock. Neutrophils, endothelial
cells, macrophages, etc. are all
activated and there is microvascular,
tissue and organ dysfunction.
Effective tissue perfusion is determined
by: cardiac performance, vascular
performance, arterial pressure and
cellular function. The third- order
arterioles are the principle determinant
of vascular resistance, these arterioles
are 20-35 micrometers in diameter.
The cardiac [Starling] function curve is
shifted not just by contractility. For
example, increasing the afterload [e.g.
giving vasopressin] also shifts the
cardiac function curve down and to the
right. In patients with systolic
dysfunction, increasing the vascular
resistance will diminish the stroke
volume.
Auto-regulation is important. Between
a MAP of 50 and 150, the blood flow to
an organ is essentially unchanged.
Autoregulation physiology may be
disturbed in shocked states.
The use of the pulmonary artery
catheter can be helpful in the
differential diagnosis of shock and can
be tested on the boards. The PAC is in
decline because there is little benefit in
mortality, but no significant increase in
h
32
mortality. There was likely overuse of the PAC in
less-sick patients. Most monitoring methods show
no mortality effect. In the RCTs that studied PAC, the
patients who the physician thought would benefit
from a PAC were NOT randomized [i.e. they were
placed

Notes 2014
into the PAC group]. The
PAC must be measured
correctly, interpreted
correctly and applied
correctly.
VASOACTIVE AGENTS
The hemodynamic
response to dobutamine is
to increase cardiac output,
heart rate, decrease SVR
and lower filling pressures.
It is a synthetic derivative
of isoproterenol. It has
affinity for beta1 and beta2
receptors. It will also
augment coronary blood
flow. As it has a propensity
to lower systemic vascular
resistance, so this may
cause hypotension in sepsis
or hypovolemia.
Levophed is more of a
vasopressor than an
inotrope, but has some
cardiac effects.
Epinephrine has both very
potent alpha [vasopressor]
and beta [heart rate,
contractility, vasodilatory]
effects. Epi may cause
bronchodilation as well.
There was worsening heart failure and
pulmonary
hypertension.
TYPES OF SHOCK
Hypovolemic shock there is
decompensated shock at 25% blood
loss. At 10% loss there can be a totally
normal blood pressure. With less than
20% blood volume lost there can be
cool extremities, increased capillary
refill, diaphoresis, collapsed veins and
anxiety. With moderate blood loss [2040%] of the blood volume there is the
addition of tachycardia, tachypnea,
oliguria and postural changes. Then in
Dopamine increases renal blood flow

and urine output, but when used
routinely [ANZICS trial in Lancet] there
is no effect on renal function.
33
Vasopressin [AVP] directly stimulates

smooth muscle contraction [V1a].
Concentration of AVP is depressed in
septic shock. In certain shock states it
results in an impressive rise in blood
pressure. It decreases cardiac output
and may produce myocardial ischemia.
Patients in the VASST trial [NEJM 2008]
with myocardial ischemia were
excluded. Additionally, AVP will raise
pulmonary artery pressure, filling
pressure. In low dose vasopressin [i.e.
contemporary dosing from the VASST
trial], there is an increase in
splanchnic blood flow and urine
output.
Whereas in higher doses [when used
to treat severe GI hemorrhaging] AVP
lowers splanchnic blood flow, but these
doses were related to an unacceptably
high risk of myocardial ischemia and
digital ischemia.
Nitric oxide antagonists also increase
blood pressure, but decrease survival in
septic shock.
severe blood loss [more than 40% of the blood
volume] there is hemodynamic instability, marked
tachycardia, hypotension and AMS.
There is older data suggesting delayed
resuscitation with fluids with penetrating trauma is
beneficial [Mattox NEJM -1994] see chapter 9 for
details.
Moving to extra-cardiac obstruction such as
tamponade there can be a very slow accumulation
of fluid that has minor hemodynamic effects, but a
small, but rapid accumulation can be fatal. The
causes of pericardial effusion are usually idiopathic
or malignant [about 20% each], iatrogenic causes
are 18%. Infectious cause are less than 10% and
renal failure is less than 3 percent. Becks acute

Notes 2014
cardiac compression triad is 1.
Hypotension 2. A small quiet heart and
3. a rising systemic venous pressure.
Note that this triad is often absent in
medical patients.
Cardiogenic shock is most often due to
acute MI or a mechanical abnormality.
Sometimes you will get an RV infarction
on the boards [see RV infarction above].
The vast majority of cardiogenic shock
is due to LV pump failure.
There is 40% or more loss of the LV
function. Killip class IV is cardiogenic
shock, and class III is overt pulmonary
edema [clinically, radiographically].
Only 25% were in shock at
34
presentation in the SHOCK registry, and
the
remaining 75% developed shock within
24 hours. In the GUSTO trial 11% were
shocked on presentation. So cardiogenic
shock develops in front of you.
What agent to use in cardiogenic shock?
Based on the SOAP II trial,
norepinephrine is a good choice. Balloon
pumping by diastolic augmentation is
frequently tested on the boards in terms
of timing or inflation and deflation. The
SHOCK trial by Hochman et. al showed
that revascularization was better than
thromobolysis not at 30 days, but at 6
months and one year, there was a
difference. Revascularization saves
more lives than lytics, ASA, betablockers and ACEI in myocardial
infarction. The SHOCK trial performed
revascularization within 18 hours of
presentation.

Notes 2014
2. CRITICAL CARE
PULMONOLOGY
If a patient needs a patent airway, the

patient needs to be intubated. This
applies for frequent suctioning, GI
bleeding with frequent emesis [though
intubation neednt be routine in GIB],
aspiration risk, etc. Clearly for work of
breathing and gas exchange, intubation
is required. There are case reports of
patients who cannot be intubated [e.g.
severe, severe head and neck burns]
and such patients have received ECMO.
In patients capable of protecting their
airways with an upper GIB,
retrospective studies have shown that
patients who were prophylactically
intubated had a much higher risk of
aspiration pneumonia following the
procedure. There was a suggestion of
increased mortality in the intubation
group as well. Nevertheless, massive
bleeding and/or an inability to protect
the airway should prompt intubation
prior to endoscopy.
An airway scenario on the board exam
may be hereditary angioedema.
Consider this in a patient who has a
first degree relative with similar
symptoms of tongue swelling, lip
swelling and upper airway edema.
Additionally, these patients often
present with predominant abdominal
symptoms because of peri-colonic
edema. Complement cascade is
activated in these patients for unclear
reasons and the C4 level is typically
abnormally low. There is an autosomal
dominant inheritance pattern of C1
esterase inhibitor deficiency such that
the complement cascade is readily
activated. There should not be
urticarial lesions. There is frequently
some precipitating agent or event such
h
23
as a minor dental procedure, trauma, etc.

Antihistamines, epinephrine and steroids in
the setting of hereditary angioedema have
an unpredictable effect. Attenuated
androgens may help restore C1E1 levels.
The use of FFP may

Notes 2014
AIRWAY
BASICS
replace C1E, but it also replaces
other complement factors which
may prolong the attack. In Europe,
purified C1E inhibitor is available.
Intubation in these patients may
be challenging.
How to verify tube placement
during intubation? The most
important is direct visualization of
the tube passing between the
vocal cords, also: no air during
gastric auscultation, bilateral
breath sounds, reservoir bag
compliance, bag movement,
condensation, carbon dioxide in
exhaled gas. During cardiac
arrest, there will be no carbon
dioxide in exhaled gas. If carbon
dioxide returns, resuscitation is
adequate.
Mallampati oral pharynx score
correlates with airway grade
somewhat. Malampatti I means
full view of oropharynx [OP] and
this correlates with grade I airway
[fully see both cords between epiglottis24

and arytenoids], Mallampatti II can see
OP, but uvula touches back of tongue
and a grade II airway means that a
portion of the cords may not be seen.
Mallampatti III is loss of uvula with
portion of OP seen, Grade III airway is
no cords seen, space seen between
arytenoids and epiglottis. Mallampatti
IV is tongue completely occludes OP, no
cords seen with Grade IV airway.
PREPARING TO INTUBATE
Needed? laryngoscope, always have
oxygen, always have suction,
medications for hypotension and
cardiac arrest. You must be able to bag
valve mask the patient.
Sniffing position is key, a little reverse
Trendelenburg can help. Mask
ventilation must be mastered. There is
evidence to suggest lower lip mask
ventilation improves mask ventilation
essentially nasal versus oral bag
ventilation can help truly ventilate the
patient.
h
There is a mortality of up to 3% during
an emergent airway. The long term
survival of emergent airways is 4555%, induction can cause cardiac
arrest. Pre-oxygenation is less effective
in emergent airways. In healthy adults
you have many minutes of apnea time.
THE INTUBATION
After three attempts, there is a very
high risk of hypoxemia, there is a 25%
chance of a surgical airway, 50% have
gastric aspiration, 2% cardiac arrest
rate. Persistent attempts are correlated
with poor neurological outcome. After 2
attempts, the 3 time must be the
charm.
Gum elastic bougie can help obtain an
airway, as well more muscle relaxants
and narcotics as this likely decreases
complications; an experienced
attending decreases complications.
Cricoid pressure can decrease the lower
esophageal tone! There is no standard
practice for RSI, it is not clear that it
prevents aspiration, it is done for
medical-legal reasons, and it is what
you do for board exams.
Induction agents include: propofol 0.51.5 mg/kg, etomidate 0.1-0.2 mg/kg,
etc. LMA? Can be helpful, but not the
right answer if a patient is vomiting.
Videolaryngoscopy requires practice,
perhaps maybe more than direct
laryngoscopy.
Recognize differences in
pseudocholinesterase levels and
activities. If a patient receives a bolus
of succinylcholine for intubation and
has not moved in two hours. The
patient may have a genetic abnormality
in pseudocholinesterase activity.
Succinylcholine is a molecule very
similar to acetylcholine. It is a
depolarizing paralytic and is rapidly
degraded [within minutes] by

pseudocholinesterase which is present in the
synapse and the serum. If the patient is deficient in
this enzyme, succinylcholine will last for quite
some time. While levels of pseudocholinesterase
may alter succinylcholine metabolism to a mild
degree, the real question is activity level as
genetic mutations can

severely impair the
enzymes ability to degrade
succinylcholine despite
normal levels of the
enzyme.
10% incidence with both etomidate

and midazolam! It is essentially
ubiquitous and you can expect it in a
good portion of all patients.
Always be ready with fluids and a
vasopressor stick.
FOLLOWING THE INTUBATION
Etomidate - one dose can be associated

with adrenal suppression for 48 hours
or longer in trauma patients. There was
no difference between ketamine and
etomidate in terms of intubation
outcome/hypotension. In terms of
adrenal function, ketamine was better
etomidate has more adrenal
dysfunction [Lancet 2009] but there
was no difference in intubating
conditions. Ketamine was better in
terms of mortality, but not statistically
significant.
Ketamine has several active
metabolites, it can increase ICP.
About 4% of emergent
intubations in the ICU are
complicated by right
mainstem insertion.
Orotracheal intubation is
associated with more ETT
movement and can be
tongued much better by
the patient as compared to
nasotracheal intubation; the
latter is more associated
with sinusitis. If there is a
rapid drop in blood pressure
and jump in heart rate, then
think tension
pneumothorax. Mucous
plugging of the
endotracheal tube could
produce similar results, so
suction should always be
tried.
What about post-intubation
hypotension? It had a 6-
BASICS OF OXYGENATION
Generally, mode, tidal volume, rate,

and other settings have modest effects
on PaO2. In the ARDS Network tidal
volume trial, smaller tidal volume was
associated with a decrement in the
ratio of PaO2 to FIO2 [156 vs 178],
despite the improved mortality!
obtained under high flows [above 0.1 L/s] [note,

normal ventilator flows are given in liters per
MINUTE e.g. 60 L/min] the inspiratory PV curve
represents dynamic conditions and therefore the
shape of this curve cannot speak to lung
recruitment as

OXYGENATION & PEEP
FIO2 less than 0.6 is considered
nontoxic. There is some experimental
evidence that injured lung may be
more resistant to oxygen-induced
injury. Try to limit exposure to
concentrations less than
0.6 for less than 24 h; instead use PEEP,
diuresis, positional maneuvers, or
inhaled vasodilators.
Clinical trials addressed the potential
role of higher PEEP. While each failed to
demonstrate enhanced survival, all
showed a trend in that direction. Liberal
PEEP typically uses PEEP above 12, but
limits end-inspiratory plateaus to less
than 30-35 cm H2O. The ARDSNet trial
targeted a PaO2 of 55 to 80 using
different PEEP-FiO2 tables [conservative
versus liberal]. On day 1 the differences
in PEEP were 8-9 versus 14-15, but the
trial was stopped early for futility [same
length on the ventilator, same
mortality]. The Canadian LOVS trial and
European ExPress trial were both larger
[983 and 850 patients, respectively]
which showed similar outcomes.
The stress-index has been advocated to
titrate PEEP. Normally the airway
opening pressure rises linearly during
constant flow, volume- controlled
ventilation because respiratory system
mechanical properties [compliance and
resistance] do not vary much over the
tidal range. If compliance increases late
during tidal inflation [suggesting that
lung is being recruited], the pressuretime display will be convex [rounded]
upward (stress index < 1): more PEEP is
likely to be helpful. If the pressure-time
display is concave [scooped] upward,
compliance is falling as lung is being
over-distended. Perhaps PEEP should be
reduced.
When the inflation pressure-volume
[PV] curve of the respiratory system is
www.heartlung.org
JE
K

Review Notes
unknown if this worsens outcome or if

modifying it improves outcome.
Under dynamic conditions, nothing
reliable may be obtained from the
deflation curve because airway
pressure is determined by expiratory
h
flow and the resistive elements of the
lung and ventilator tubing [and PEEP
valves]. The only way the deflation
curve can provide meaningful
information is if the ventilator tube is
intermittently occluded to allow the
Paw to equilibrate with the alveolar
pressure. Therefore, the difference
between the two curves does not say
anything about the hysteresis
properties of the lungs [under dynamic
conditions].
lung recruitment is a static parameter.
In
25
addition to the static

characteristics of the lung, a
dynamic PV curve also
includes inadvertent PEEP
and the resistive
components of the
intubated respiratory
system. This cannot provide
accurate data as to where
PEEP should be set [e.g.
above the lower inflection
point] because this inflection
point [under dynamic
conditions] represents more
than a simple recruitment
threshold.
MEAN AIRWAY PRESSURE

In addition to FIO2 and PEEP, the
mean airway pressure affects
recruitment and oxygenation. Highfrequency oscillatory ventilation
(HFOV), inverse ratio ventilation
(IRV), and airway pressure release
ventilation (APRV) are various ways
to raise mean airway pressure.
The most valuable piece of

data from a dynamic PV
curve of the respiratory
system is the upper
inflection point which, if it
flattens, suggests that the
lungs stiffen [compliance
decreases] at high lung
volume [similar to the stress
index]. The problem with
this information is that it is
lung.org
BASICS OF VENTILATION
Carbon
dioxide
tension
is
determined
by
the
balance
between its production and alveolar
ventilation. The latter is controlled
by the
JE
K

Review Notes

Notes 2014
ventilator via minute ventilation. Carbon
dioxide
is also eliminated during HFOV through
various incompletely understood
mechanisms [e.g. Taylor Dispersion].
PHYSIOLOGICAL CONSIDERATIONS
29
The pressure-volume curve of the lung is

sigmoidal, over-distension at high volumes,
atelectrauma at low volumes.
PRESSURE VERSUS VOLUME-PRESET

VENTILATION
Preset, here, refers to the variable that
the clinician wants to keep constant
[i.e. pressure versus volume]. As
below, the mode of ventilation refers to
the types of breaths that the ventilator
allows. A breath is composed of a
trigger, a limit [or target] and a cycle
variable.
Volume preset ventilation [e.g. VACV or
assist- control] is flow-limited and
volume-cycled while pressure-preset
ventilation is pressure-limited but either
flow-cycled [pressure support PS] or
time-cycled [PACV or pressure control].
Pressure-preset modes, in theory,
makes lung protective ventilation
simpler by elimination of the need to
repeatedly determine Pplat and
periodically adjust the VT. During use of
pressure- preset modes, the patient
also has greater control over inspiratory
flow rate which may improve comfort.
Several features of pressure - preset
modes have raised concern that lung
protection cannot be assured. Most
importantly, a safe level of maximal
alveolar pressure is not known.
Moreover, unless the patient is fully
passive, the trans-pulmonary pressure
cannot be controlled or known using
pressure-preset modes. A final
limitation is that pressure-preset modes
do not allow ready determination of the
respiratory system mechanical
properties.
CONVENTIONAL MODES OF MECHANICAL VENTILATION

Notes 2014
Total lung capacity is a
trans-pulmonary pressure
of about 30 cm H2O. Is it
the reduction in the plateau
pressure or tidal volume
that benefits the lung?
Remember that tidal
volume best reflects transpulmonary pressure
whereas plateau pressure
does not. People who play
wind- instruments generate
more than 100 cmH2O in
their airways when playing
the instrument, but this
does not cause lung injury,
because their transpulmonary pressure [i.e.
lung volume] is normal as
they play their instrument.
Understand timeconstants. The compliance
of the respiratory system
[in the passive patient]
multiplied by the airflow
resistance is the timeconstant [TC] and
multiples of TC is the time
it takes for the lung to
deflate by 67%, 90%, 95%,
and 99%. Consider a
patient with a respiratory
flow triggers sensing systems working
simultaneously. Target [or limit] and
cycle are the two other variables [in
addition to trigger] that determine
the type of breath.
Volume-preset breaths classically
deliver a constant flow [flow-limited]
and volume [cycled] and airway
pressure varies. If you set volume, you
have total control over minute
ventilation. By contrast, pressurepreset breaths have a set pressure
[pressure-limited] and set inspiratory
time [time-cycled] and flow and volume
vary.
30
system compliance of 20 mL/cm H2O
and a resistance of 10 cm/1000mL/s.
How long would it take the lung to
empty by 99%? The time constant is
0.2 seconds. Thus this lung would
empty 99% of its inspiratory volume in
[0.2 x 4 =
0.8 seconds]. This is important to know
for the development of intrinsic PEEP.
What about cardiac function effects?
ITP tends to retard venous return. ITP
increases from spontaneous to assisted
ventilation to controlled ventilation; in
sick lungs it is assumed that perhaps
0.25 to 0.33 of the airway pressure
makes it to the pleural pressure
[grossly oversimplified]. If intravascular
volume is low, high ITP can impair
venous return, if intravascular volume
is high, LV afterload reduction
predominates.
BASIC BREATHS
Control breaths are triggered by the
ventilator, assist breaths are triggered
by the patient. When patient-mediated,
the trigger can be a pressure trigger or
flow trigger. The flow trigger used to be
more sensitive, but now with fancy
microprocessors, the pressure trigger is
just as good. Many ventilators have
both pressure and
The nice thing about pressure-preset ventilation is
that the flow is variable and the patient may be
more comfortable if the patient is triggering the
breaths.
How do you cycle a breath? There may be a set
volume [volume-preset breaths, i.e. VC and VA],
time [pressure-preset breaths, i.e. PC and PA] or
flow reduction threshold [pressure support].
There are five basic breaths of mechanical
ventilation [VC, VA, PC, PA, PS]. The combination of
these breaths determines the mode of ventilation
[below].
VOLUME ASSIST-CONTROL VENTILATION [VACV]

Notes 2014
VACV is also known as assist control or
volume control. It is the mode used in
the ARMA trial demonstrating reduced
mortality in patients with ALI and ARDS;
respiratory mechanics can be measured
readily.
The clinician sets the minimum
respiratory rate and tidal volume. The
patient may go above this depending
on how often they are triggering
assisted breaths.
When high inspiratory effort continues
during the ventilator-delivered breath,
the patient may trigger a second,
superimposed (stacked) breath (rarely
a third, as well) aka flow- starvation
as noted by negative deflections in the
pressure tracing.
31
Patient effort can be increased [if the goal is to
exercise the patient] by increasing the
magnitude of the trigger or by lowering
VT [which increases the rate of
assisting]. Lowering f at the same VT
generally has no effect on work of
breathing when the patient is initiating
all breaths.
The mode of VACV is composed of VA
and VC breaths.
SYNCHRONIZED INTERMITTENT MANDATORY VENTILATION
In the passive patient, SIMV cannot be
distinguished from controlled ventilation
in the ACV mode.
The difference occurs when the patient
triggers a breath.
If the triggering effort comes in a brief,
defined interval before the next
mandatory breath is due, the ventilator
will deliver the mandatory breath ahead
of schedule to synchronize with the
patients inspiratory effort [no different
from an ACV-assisted breath].
If a breath is initiated outside of the
synchronization window, VT, flow, and
I:E ratio are determined by patient
effort and respiratory system
mechanics, not by ventilator settings.
The spontaneous breaths tend to be of
small volume and are highly variable.
Most ventilators today add pressure
support to these additional breaths.
SIMV has been shown to prolong
weaning in various RCTs.
SIMV can be composed of PA and PC
breaths or VA and VC breaths with or
without additional PS breaths.
PRESSURE ASSIST-CONTROL VENTILATION [PACV]
PACV is also known simply as pressure
control. In the passive patient,
ventilation is determined by f, the
inspiratory pressure increment (Pinsp-

Notes 2014
PEEP), I:E ratio, and the
In patients without severe obstruction
(i.e., time
constant not long) given a sufficiently
long TI, there is equilibration between
the ventilator- determined Pinsp and
alveolar pressure (Palv) so that
inspiratory flow ceases. In this
situation, VT is highly predictable,
based on pressure difference and the
mechanical properties of the
respiratory system (Crs).
In the presence of severe obstruction or
if TI is too short to allow equilibration
between ventilator and alveoli, VT will
fall below that predicted based on Pinsp
and Crs.
It is typically the case during PACV that
alveolar and ventilator pressures do not
equilibrate either at end-inspiration or
at end-expiration. Thus the maximal
inspiratory alveolar pressure is
generally less than the set inspiratory
pressure on the ventilator and the endexpiratory pressure exceeds the set
expiratory pressure (i.e. there is autoPEEP).
The active patient can trigger
additional breaths by reducing the
airway opening pressure (Pao) below
the triggering threshold, raising the I:E
ratio. The inspiratory reduction in
pleural pressure combines with the
ventilator Pinsp to augment the transpulmonary pressure and the VT. This
point leads many intensivists to be
skeptical regarding the ability of PACV
to ensure lung-protective tidal volumes
in patients with ALI and ARDS!
PRESSURE SUPPORT VENTILATION
Non-invasively, pressure support is
known as BiPAP. Once a breath is
triggered, the ventilator attempts to
maintain Pao [airway pressure or airway
occlusion pressure] at the physician-
h
32
time constant of the patients
respiratory system.
determined Pinsp, using whatever flow is necessary
to achieve this. Eventually flow begins to fall as a
result of either cessation of the patients
inspiratory effort or increasing elastic recoil of the
respiratory system as VT rises. The ventilator will
maintain a constant Pinsp until

Notes 2014
inspiratory flow falls an
arbitrary amount (eg, to
25% of initial flow) or below
an absolute flow rate.
Because patients respiratory
system timeconstants vary
widely [so that the time for
flow to fall to 25% varies
widely], many patients have
to actively work to turn off
the inspiratory pressure,
raising the work of breathing.
Especially in patients with
exacerbations of COPD, a
threshold well above 50% is
often necessary to minimize
this unintended expiratory
work.
Respiratory system
mechanical parameters
cannot be determined
readily on pressure support
because the ventilator and
patient contributions to VT
and flow are not represented
by Pao. That is, because PS
is flow-cycled [or, similarly in
PC, time-cycled], inspiration
ends before proximal airway
pressure equalizes with
breaths [causes more frequent cycling
between
Phigh and Plow]. Plow should almost
always be zero [i.e. dont add traditional
PEEP]. The rapid mahcine expiration
inspiration cycling of APRV creates autoPEEP.
Understand the physiology of inverse
ratio ventilation. The mechanism of
improved oxygenation is most likely the
development of intrinsic PEEP which
occurs as a result of the very short
expiratory time [as above in APRV]. This
tends to increase mean airway pressure,
but reduce peak airway pressure. As
long as IRV and extrinsic PEEP lead to
alveolar pressure; thus measured

airway pressure will not correspond to
alveolar pressure.
Accordingly,
these
important
measurements of Pplat, Ppeak-Pplat,
and autoPEEP are measured during a
brief, daily switch from PSV to volumepreset ventilation.
33
A potential advantage of PSV is

improved patient comfort and, for
patients with very high drive, reduced
work of breathing compared with
volume-preset modes.
LESS CONVENTIONAL MODES OF VENT
How does APRV fit into this scheme? It is

actually a pressure-limited reverse ratio
IMV. It provides a prolonged breath to a
set pressure [P-high]; it is time-cycled
and allows spontaneous breaths to
occur [IMV]; with a long I time and short
E time [reverse ratio] the spontaneous
breaths occur during the inspiratory
phase [i.e. during P-high]. The most
effective way to increase minute
ventilation in APRV is to decrease the
time at P- high or T[high], which
increases the frequency of
the same end-expiratory lung volume [i.e. the same

end-expiratory transpulmonary pressure], the
effects on oxygenation should be the same. There
may be decreased dead space ventilation because
of the prolonged inspiratory time [to allow for gas
mixing] but this effect is likely modest at best.
Understand the basics of HFOV. The pressure swings
in the trachea [compared to the alveolus] are
relatively large. This occurs at a frequency of 150900 cycles per minute. The relationship to alveolar
pressure is that the pressure swings in the alveolus
are as frequent, but completely blunted, i.e., the
mean pressure in the alveolus is the same as the
mean in the proximal airway/trachea, its just that
the pressure in the alveolus, compared to the
trachea, is totally damped. Because each tidal

Notes 2014
breath is much lower than dead space,
the mechanism of gas exchanged is
clearly NOT convective. Carbon dioxide
clearance is improved by increasing the
pressure amplitude as opposed to
frequency. The mechanisms at play are:
coaxial flow, Taylor dispersion [gas
mixing beyond a moving wave- front],
molecular diffusion and pendeluft
mixing. Which mechanism is most
important is not clear, and they probably
all play some role depending upon the
patient underlying pathophysiology.
When non-convective forces are at play,
alveolar ventilation is determined by the
frequency multiplied by the amplitude
[squared], thus
34
increasing the amplitude has the
greatest effect
on CO2 elimination. Increasing mean
airway pressure does not help with CO2
elimination, though it can with
oxygenation. Changing the bias flow
[the continuous flow of gas in the circuit
which is oscillated] has small effects on
CO2, but not nearly as much as the
frequency or amplitude. The diameter of
the airway can also affect CO2
clearance, so creating a cuff leak can
also lower PaCO2. The ultimate
physiological benefit of HFOV is that
tidal pressure, volume and atelectasis
swings are reduced. HFOV from the
perspective of the alveolus is best
described as CPAP with wiggle meaning
the pressure is elevated, with small
deflections about this mean.
There are feedback controls available
for example PRVC on the Servovents
[a.k.a. auto flow
on the Drager, also known as VC+],
in an attempt to mesh the best of both
worlds. It is a pressure preset
[pressure-limited, time-cycled] mode
that will vary the pressure limitation
up and down based on the compliance
of the respiratory system and the
measured volume achieved. It is like
having a little RT in your machine
changing the pressure to get the right
tidal volume.
ASV or adaptive support ventilation is
only available on the Hamilton. It
adjusts tidal volume for you using
feedback to minimize ventilator work;
further, it alters the I:E ratio to minimize
air-trapping. This is a fancy, algorithmbased mode of ventilation whereby the
clinician sets the minute ventilation,
PEEP and fraction of the minute
ventilation supplied by the ventilator.
The ventilator then supplies some test
breaths and calculates resistive and
elastic work of breathing and sets its

Notes 2014
own tidal volume and
respiratory rate to minimize
work and such that there is
always 1 inspiratory time
constant to 3 expiratory
35
time constants. Clinically, ASV has been
compared to physician preferences and
when the clinician preferences are high
lung volumes, ASV tends to select lower
lung volumes
with higher respiratory rates. When

low
compliance situations exist, ASV tends
to select higher tidal volumes than the
recommended 6 cc per kg. ASV may
help weaning when compared to older
weaning methods, but not yet
compared with newer SBTs.
3
h

Notes 2014
and limitation of right-ventricular
outflow
impedance during inspiration,
regardless of the NAVA level.
INTRINSIC PEEP, SYNCHRONY AND LIB
INTRINSIC PEEP OR AUTO-PEEP
New feedback controls PAV and NAVA.

Proportional assist ventilation with a bigger
effort, there is an increased flow [like
it showed that right ventricular performance is less
pressure assist or pressure support],
impaired during NAVA compared to PSV. Proposed
but also more pressure in an attempt to
mechanisms are preservation of cyclic intra-thoracic
unload the muscles of respiration. So it
pressure changes characteristic of spontaneous
responds with flow AND pressure. The
breathing
downside is that if the effort drops, your
MV drops, there are no minimums with
PAVS, so the patient must be awake and
cooperating. There may be a waveform
case of PAV on the boards. Consider a
patient receiving MV with varying
pressure and flow based on effort
[estimated from esophageal pressure
tracing]. PAV or proportional assist
ventilation calculates inspiratory
resistance and compliance from a test
breath. It monitors inspiratory flow
demand, calculates the work of
breathing [i.e. pressure requirements
for desired flow and volume] and
applies a set proportion of required
pressure. It is compared to power
steering on an automobile. The driver
selects the distance to turn wheel and
the system supplies pressure to reduce
effort. Like the automobile driver, the
patient must be reliable.
What about NAVA? It short-circuits the
normal pathway between the
generation of the trigger and the
ventilator. It uses direct diaphragm
contraction as the trigger. The greater
the EMG- diaphragm contraction, the
greater the breath. The studies are
small and observational. There was a
recent great study on NAVA and
cardiopulmonary interaction in the
Critical Care Medicine September 2014;

Notes 2014
What about intrinsic PEEP
[PEEPi]? What determines
PEEPi? Minute ventilation,
I:E ratio,
& time constant. In volume
control ventilation, peak
and plateau pressures go
up with PEEPi. In pressure
control ventilation, what
happens is that tidal
volume and flow decrease
with PEEPi [this can be
exceptionally important
during PRVC as tidal
volume falls with PEEPi, the
ventilator can respond by
increasing the pressure
delivered].
Classic sign of air-trapping expiratory flow does not
return to baseline.
for-your-buck is usually to lower the

tidal volume and/or RR. Reducing the
respiratory rate will prolong the E time
and mitigate the air- trapping. Adding
PEEPe will not fix the problem unless it
is a triggering problem [below].
To fix PEEPi, you can fix

one of the three
determinants decrease
tidal volume [or RR],
improve I:E, reduce timeconstants. The best bang-
What are the other consequences of

PEEPi? Well, the circuit pressure must
fall prior to an assisted [patienttriggered] breath; PEEPi adds a
threshold load in this situation as it
raises the
Think of air-trapping in a patient with

COPD with a rising PaCO2. The effect is
to increase intrathoracic pressure and
the classical signs are increasing Paw in
VACV, decreasing Vte in PACV and
incomplete expiratory flow. There may
also be decreased blood pressure and
high dead space with increasing PaCO2.
There will be a case of COPD with
increasing PaCO2 and decreasing BP
and the correct answer is to reduce the
RR, Vte or disconnect the circuit. In a
passive patient, PEEPtot may be
measured at end-expiration.

Notes 2014
circuit pressure. Put your hand on the
chest and
eye the ventilator. If there is effort and
no breath, this is a classic sign of high
trigger threshold load. You can remove
the PEEPi, or add PEEPe [extrinsic
PEEP] to lower the trigger threshold.
The breath rate might increase because
of the increased trigger sensitivity
dont be alarmed; youve done the right
thing.
VENTILATOR SYNCHRONY
What about synchrony? The response
of the ventilator lags behind patient
effort during both inspiration and
expiration. This causes fighting and
increased sedation. Look at the assisted
[patient-triggered] breath pattern and
compare it to a control breath
waveform [if there is one].
Look to the pressure graphic, it may
have a negative deflection!
Sometimes flipping to a pressurepreset mode can improve synchrony.
PAV and NAVA have theoretical
appeal.
Ineffective triggering is seen in patients
with airway obstruction and auto-PEEP
and is the most common cause of
ventilator asynchrony in the ICU. There
is a drop in airway pressure and a small
increase in inspiratory flow, but no
breath delivery. This is due to PEEPi as
above, and is treated as described
above.
Recognize double-triggering [DT]. The
second most common cause of
ventilator asynchrony. DT is seen in
patients with very high respiratory
demand with a short inspiratory time.
There is a very brief expiratory phase
noted, [less than half the duration of
the preceding inspiratory time] and
there is a second breath delivered. The
h
38
drop in airway pressure in early expiration triggers
another breath.
Auto-triggering is seen when there are frequent
drops in airway pressure that trigger breaths
[usually caused by a leak] and seen during low
respiratory rates. Auto-triggering [whether
pressure or flow is the trigger] can be caused by
cardiac oscillations. This phenomenon has been

Notes 2014
studied in cardiac surgery patients who
tended to
have larger hearts, larger
cardiac output and therefore
larger effects on intrathoracic pressure tracings.
The clue is typically cardiac
oscillations noted in both
the pressure and flow
tracings with followed by
rapid [usually less than one
second] breaths delivered
by the ventilator. A leak in
the ventilator circuit, water
in the ventilator circuit,
hiccups or chest tube are
other causes of autotriggering.
Recognize flow-starvation
[or flow asynchrony] on the
ventilator. Consider a
patient who is dyssynchronous with the
ventilator on VACV and the
pressure waveform shows
low [scooped out] Paw with
negative inspiratory
deflections. The flow is
constant at 45 L/min and
there is a consequent long
inspiratory time. There is an
intervention and the next
those who did not receive a protocol.
Measures
of respiratory muscle pressure, vital
capacity, etc have been used to predict
extubation success, but they are rather
poor indicators.
When to consider vent discontinuation?
There must be a reversal of ARF, the PF
ratio must be above 150-200, PEEP less
than 8, FiO2 less than 0.4, pH more
than 7.25, minimal inotropic support,
reliable inspiratory efforts. Once SBT
passed cough is essential, cough
velocity must be more than 1 L/sec.
Some perform the white card test
39
waveform bundle reveals normal
pressure waveforms that are higher and
higher flow rate. The EEP level of the
Paw has not changed, but the
inspiratory time is shorter confirming
the intervention was to make the
ventilator deliver higher flow rates.
Note that flow starvation may result in
double triggering as described above.
LIBERATION FROM THE VENTILATOR
What about discontinuation of the
ventilator? What is the best way to
liberate? Pressure support? SIMV? The
one that worked the best is a technique
that doesnt wean the daily
spontaneous breathing trial technique.
SBT should be 30-120 minutes [see
Tobins editorial in the Blue Journal The
Myth of Minimal Ventilator Settings
early 2012].
SBT via pressure support or T piece is
equally effective based on the trials
and 30 minutes versus 2 hours of SBT
are equally effective. Note that there
is data in COPD patients who failed Tpiece and then were immediately
extubated and placed on NIPPV that
showed improved weaning outcomes.
Other data reveals that patients
assigned to a weaning protocol did
better than
does a goober hit a card with cough? Also consider
suctioning frequency [if more than every 2 hours
probably too much]. Less important than cough is
actually gag reflex, cuff leak and alertness.
Evidence favors excessive secretions [suctioning
every 1-2 hours] as most likely to predict
extubation failure, above and beyond that of
altered sensorium. In conclusion one of the most
important factors, is the ability to cough and clear
secretions.
The use of a cuff leak to predict post-extubation
stridor is contentious with one study showing that a
returned volume within 110 cc of the tidal breath
predicted post-extubation stridor. The risk factors
for post-extubation upper airway obstruction are:

Notes 2014
female, trauma, repeated intubations,
and length of intubation duration.
There are conflicting data regarding the
use of steroids prior to extubation. The
risk of post- extubation laryngeal
edema is variable [2-22%] and the rate
of reintubation is less than 5%. It can
be difficult to predict which patients will
have this problem. Female gender and
traumatic injury increase the risk of
post-extubation laryngeal edema. In
one trial patients were indiscriminately
given 80 mg of MP 12 hours prior to a
planned extubation and there was less
clinically significant laryngeal edema
and less reintubation due to laryngeal
edema [8% versus 4%]. Some dont
believe these results. One more recent
trial gave steroids only to patients who
had a reduced cuff leak [the difference
between
the inspiratory and expiratory volume

should be
more than 10% of the tidal volume or
110 cc] but their cutoff was less than
24% of the total tidal volume and this
resulted in reduced edema and
reintubation [30 versus 8% - these
numbers still seem high to me].
40
The Kress trial looked at daily sedation

vacations and found that mechanical
ventilation was reduced by 2 days, ICU
time by 3.5 days and benzodiazepine
doses were cut in half [not true for
propofol]. The amount of selfextubations was not different between
the two groups.
However, the ABC trial in Lancet,
2008 showed that those patients
awakened and given an SBT did selfextubate more, but there was no
difference in re-intubation rate.
Continuous infusion of sedatives
tends to provide a more steady state
concentration in the blood than bolus
dosing.
Why does someone fail an SBT?
Search for causes, try again in 24
hours in the meantime there is little
benefit to changing settings.
Recognize that a Passy-Muir [PM] valve
can cause asphyxia and respiratory
arrest if used in conjunction with a
cuffed tracheostomy. It makes
exhalation impossible. Removal of the
PM valve should always be attempted
first if this is a consideration [rather
than drop the cuff] because the internal
diameter of an un-cuffed or deflated
tracheostomy can still obstruct the
airway. Only removing the PM valve will
be fail-safe. The incidence of aspiration
has been shown to decrease with the
use of a Passy-Muir valve [compared to
tracheostomies without a PM valve,
not compared to no tracheostomy]
because it may promote normal
expiration during gluttation. There is

Notes 2014
also reduced secretions and
improved cough with the
Passy-Muir as compared to
standard tracheostomies.
Air escaping through the
be able to clear secretions, have an
adequate
cough, have normal mentation and be
hemodynamically stable before being
candidates for a PM valve.
ARDS
ARDS must be of an acute onset. The

old definition of ARDS required a PF
ratio of less than 200 [less than 300
was acute lung injury]. There must also
have been bilateral CXR infiltrates and
there should have been no evidence of
LA hypertension.
The new, Berlin Criteria get rid of ALI
and define mild [300-200], moderate
[200-100] and severe [less than 100]
ARDS by PF ratio. Also, acute requires
7 days of symptoms onset or less and
there is no need to absolutely rule out
elevated left atrial pressure by any
means other criteria such as PEEP
and pulmonary compliance, etc. did
not predict clinical outcome.
Primary [or pulmonary] ARDS tends to
be patchy, less PEEP responsive [see
Blue Journal Gattinonni article 1998]. If
it does not evolve to SIRS/MODS then
the outcome is better than secondary
ARDS. Secondary [or extra-pulmonary]
ARDS occurs from severe hypotension,
pancreatitis, abdominal sepsis tends
to be more diffuse, more PEEP
responsive, but somewhat worse
outcomes.
The initial indices of oxygenation [PF
ratio, etc.] are NOT related to ARDS
mortality. In fact, in the ARDSNet trial,
initially the patients with the lowest PF
ratios tended to have the best
outcomes. The most common cause of
h
41
nose and mouth likely contribute to
secretion evaporation, as well as the
improved cough response. In patients
who are unable to be fully liberated
from mechanical ventilation, they must
death is worsening MODS and sepsis. Any nonpulmonary organ dysfunction, especially liver as
well as the failure to improve after 7 days of
treatment were negative predictors of survival.
So there is a disconnection between physiology
and outcomes. ARDSNET improved outcome, but
made gas-exchange worse! Small tidal volumes
resulted in worse gas exchange and worse

Notes 2014
compliance [perhaps from
increased atelectasis]
until the third day and
then things get better.
There is clearly respiratory
acidosis with small tidal
volumes. There is some
evidence that respiratory
acidosis helps cell injury.
Perhaps by unloading
oxygen from hemoglobin.
The mortality rate for
ARDS has decreased
since the early 80s [above
60% or so] to mid-30%
now in early 21st century.
There has been no
reduction from 1994 to
2006! Sepsis is the main
disease in ARDS/ALI. It is
a disease of old people
with pneumonia which is
probably why mortality
has plateaued. In 80 year
olds, there is a 50%
survival rate in ARDS.
A portion of patients who
survive long bouts of ARDS,
suffer from chronic critical
illness [generally defined as
respiratory failure lasting
weeks see Blue Journal
concise clinical review
with a certain pulmonary insult or in
small clinical
trials may not be applicable to all
patients with ARDS.
It is known that in ARDS there are
pulmonary microthrombi and a couple
of large sepsis trials have looked at APC
[xigris] which were initially positive,
have now fallen out of favor.
When does one use steroids in late
ARDS? There is no right answer, some
people do it. It was once thought that
ARDS would respond to high dose
42
2010] and require tracheostomy and
LTAC placement. In these patients:
most are not freed from mechanical
ventilation [reported liberation rates: 30
- 50% and if success is achieved, it's
almost always within 60 days]; at the
end of one year, fewer than half will be
alive [32 - 52%]; and 12% will be alive
and independent.
How can ARDS be treated?
Unfortunately, there is very little to be
done other than blocking
manifestations to buy time for the
patient to improve. The most
important thing to do is not to make
things worse. Appropriate infection
management as quickly as possible is
very important, as well as surgical
intervention/source control where
present. Minimize transfusions,
minimize excess fluids, and reduce
aspiration risk.
TREATMENT
What has failed in ARDS? Antiendotoxin antibodies, NSAIDS, antiTNF, ketoconazole, lysophylline, PGE.
Why is there so much failure? It may
have been that the human spectrum of
ARDS is exceptionally heterogeneous,
so drugs or therapies in a very specific
animal population
steroids [3-4 grams per day] it seemed to make
things worse in trials in the early 1990s.
Then steroids in ARDS seemed to get patients off
the ventilator 2 days early, but many went back and
there was no change in mortality. The steroids were
weaned off very quickly, so is it that they were
tapered off too quickly or did the patients have too
much myopathy? The answer is unknown.
What about immune-nutrition? The addition of
combinations of arginine, glutamine, nucleotides,
omega 3 fatty acids of feeding formulae. There
seemed to be a reduction in infectious complications
[2001 JAMA]. Then there were three RCTs since 1998
all three were positive ARDS patients came off the

Notes 2014
ventilator and had a mortality benefit.
By the mid-2000s it was essentially a
recommendation with grade A
evidence.
The ARDS Net OMEGA trial from CCM
2009 and stopped for futility with more
than 500 patients [more than all three
of positive trials combined]. The control
group in the ARDS NET OMEGA trial had
no fat, only carbs. In the three positive
trials, the control groups all had omega
6 fatty acids [did this cause harm?] as
O6 fatty acids are known to be proinflammatory. The answer isnt known.
So how do we buy time in ARDS then, if
we cannot mitigate the mediators?
Recall the hysteresis curve not too
high [volutrauma] not too low
[atelectrauma]. Keep the tidal volumes
43
low keep the [usually estimated]
transpulmonary pressure less than 30 cm
H2O [assuming normal chest wall
compliance]. In the seminal ARDSNet
trial [ARMA] the respiratory rate was
higher in the physiologically normal
tidal volume group [i.e. 6 cc/kg Vte
group]. Also in this group, PaCO2 were
higher, plateau pressures lower, PF
ratios and static compliance were
lower; but most importantly, the
mortality was also lower in the 6 cc/kg
group. The NNT was 10. Patients on the
lower tidal volume were off the
ventilator earlier. The ARDSNet [ARMA]
trial revealed an ARR of 9% for those
patients who received both a volumelimited [5-7 cc/kg IBW] and pressurelimited [Pplat less than 30 cm H2O]
approach with ARDS.
While there is a direct correlation
between plateau pressure and
mortality, in the ARDSNet [ARMA] trial,
the incidence of pneumothorax was
neither related to end inspiratory nor
end expiratory airway pressure
[probably because this is once removed
from the trans-pulmonary pressure]. In
normal lungs, tidal capacity is reached
at a trans-pulmonary pressure of about
30-35 cm H2O so this is used as the
upper limit for the plateau pressure; but
it is tidal volume or volutrauma that is
the culprit, not barotrauma.
Interestingly, very low tidal volumes
with extracorporeal carbon dioxide
scavenging did not show benefit on
mortality.
What is more important, plateau
pressure or tidal volume? Even when
the plateau pressure is kept less than
30, keeping the tidal volume as low as
possible seems to confer mortality
benefit. CCM review patients without
any ALI at outset, there was a doseresponse relationship between tidal

Notes 2014
volume received and the
progression to ALI.
Basically the dictum should
be that no Pplat is safe and
keep volumes low.
however, will augment right ventricular
afterload
as well as increase the risk of lung
fracture. The average reduction in
cardiac output with PEEP can be 13%.
Whats the best PEEP? There were
three big trials that addressed this. The
ALVEOLI trial [part of the ARDS Network]
was the American trial in NEJM [2004],
LOVS was Canadian in JAMA [2008],
EXPRESS was French in JAMA [2008],
2300 patients in total. All three RCTs in
high versus low were negative for high
PEEP in terms of survival. The details
are these: the intervention group was 46 cm H2O above the control.
Interestingly, survival was higher in the
EXPRESS trial for high PEEP in ARDS
[not ALI]. There are more recent trials
showing that if the lung can be
recruited, there is benefit to high PEEP.
In some patients, high PEEP [high teens]
there will be harm, in others, there may
be benefit. Overall, the trend suggests
high PEEP is better. The Canadian trial
was the largest in 2008 with almost
1000 patients. The meta-analysis in
JAMA 2010, found that high PEEP does
reduce mortality in sicker patients
[lower PF ratio], whereas the high PEEP
in ALI [PF 200-300] patients seemed to
do worse with high PEEP. If the lung
injury is less severe, the high PEEP may
harm the good lung. In the pooledanalysis, however, higher PEEP seems
to only improve PF ratio.
The use of PEEP to keep the lungs
above their lower inflection point has
been advocated to reduce
atelectotrauma, but these
measurements are difficult to make
PEEP AND ARDS
44
Increasing PEEP is an attempt to recruit

atelectatic, consolidated lung. Too
much PEEP
accurately and studies have been inconclusive [see
discussion above].
Understand patent foramen ovale. Autopsy studies
have shown an incidence of 20-30% of all people
may have some degree of PFO. If a patient is given
10 cm of PEEP, there is no change in cardiac output,
a decrease in PaO2 and an increase in calculated
shunt fraction, think PFO. PEEP per se should
reduce shunt fraction as it increases alveolar
ventilation. In theory, excessive PEEP can increase
dead space and then

Notes 2014
45
redirect flow causing

venous admixture and
potentially shunt but this
is a modest effect at best.
suggested harm [see Slutsky

commentary NEJM spring 2013]. This
mode of ventilation should now
generally be avoided in ARDS.
ADJUNCTIVE THERAPIES IN
ARDS
What about ECMO? CESAR, 180 patients

with severe ARDS and there was a
suggestion of improvement with ECMO
but this trial had many methodological
problems.
When you are failing on

maximal ventilator
management, what is the
next best step? APRV does
recruit lung & improves PF
ratio. It is also known
upside down SIMV, or
comfortable PCIRV. The long
I time can recruit alveoli.
There is little evidence to
suggest that it improves
outcome.
HFOV is basically CPAP with
wiggle in the alveolus. The
wiggle causes CO2 and
oxygen movement up and
down the tracheobronchial
tree. A 2010 meta-analysis
from 6 peer reviewed
studies showed reduced
mortality, less barotrauma,
PF ratio improves. But this
does not include the most
recent NEJM article which
arm. Again, there was no difference in
mortality,
there were 2.5 days fewer alive and off
ventilator in conservative arm and
more time out of ICU by about 3 days.
In the group that got less fluid and
more Lasix, there was less CNS
dysfunction [?
Less sedatives or less cerebral edema?]
All other organs the same.
What about paralysis? Papazian in
NEJM 2011. 24 hours of paralysis, and
then 3 weeks later there is a mortality
benefit. How can this be? The reason
is not known. Does paralysis improve
synchrony? Decrease the trans-
What about keeping the lungs dry?

The FACCT trial [NEJM 2006] showed
that there was no increased shock, no
increased renal failure, but over two
days alive and off the ventilator. The
FACCT protocol can only be instituted
when the patient is stable and no
longer in shock.
Specifically, the FACCT trial looked at
conservative fluid versus liberal as well
as central venous versus PAC. PAC
versus CVC arm showed that there was
no difference in management utensil. In
the conservative arm there was less
fluid and more Lasix. The net fluid
balance in the conservative arm was
less than the liberal arm. There was
CVP separation [8 versus 11-12]; in the
PAC arm Ppao was 13 versus about 16.
Tidal volume the same, but PEEP and
respiratory rate were slightly lower in
the conservative [Lasix]
pulmonary pressure and improve mortality in
ARDS? Be wary, however; consider a patient with
ARDS getting PACV, triggering all breaths. With
NMB, there is volume loss on following breaths.
There is loss of spontaneous effort, not
bronchospasm, mucus plugging or tension
pneumonthorax. In pressure assisted ventilation, if
the patient is working with the ventilator and the
patients contribution is removed, the patient will
lose volume and Mve.
What about esophageal balloon monitoring?
Consider a patient with obesity with high Pplat, but
end-inspiratory esophageal pressure is 17 cm H20
[normal less than 5 at end-inspiration] & Pplat is 37.
What to do? Just leave it alone. The trans-

Notes 2014
pulmonary pressure is only 20 at endinspiration [37 minus 17]! At Duke,
they raise their Pplat to BMI. That is, if
BMI is 40, Pplat can go to 40!
Understand basic effects of prone
positioning. The time-course of when to
prone a patient is not known. Proning a
patient typically results in an immediate
improvement in the PF ratio, but the
effect may be delayed as well. There is
no relation between the severity of
ARDS and the degree improvement in
PF ratio when a patient is placed in
prone. Sometimes the beneficial effect
of prone may reverse as the abnormal
ventral areas of the lung shift anteriorly.
Prone positioning doesnt change
perfusion of the lung much, but closing
volume is likely reduced in the
46
dorsal lung units when compared to
the supine
position. In 60-75% of patients with
ARDS, prone positioning improves the
PF ratio. In animal models, pronation
has been shown to protect against the
development of lung injury, but this
unknown in humans. The time when to
flip someone and when to return to
supine is not known nor is the
frequency of flips known. Prone does
increase the risk of new pressure sores
on the ventral position of the body. In a
2001 study by Slutsky, the prone
position for 7 hours per day for 10 days
had no effect on mortality, but did
improve mortality in the most severe
ARDS. However, the PROSEVA trial in
May 2013 showed that in patients with
a PF less than 150, prone for 16 hours
per day, every day, for up to 28 days or
longer, improved mortality. This
resulted in a 17% absolute risk
reduction in mortality for the prone
position. A meta-analysis in the CMAJ
this past summer echoed these findings
especially the 16 hour rule.
What about iNO? iNO selectively dilates
well- ventilated portions of the lung to
improve VQ and this improves
oxygenation. It also tends to lower
physiological dead space for a similar
reason and carbon dioxide elimination
is improved but this effect is seen only
in patients with a pre iNO value of
PaCO2 more than 50 mmHg. There is
no change in overall mortality. At small
doses the risk of Met-Hb is low. iNO at
high inhalational levels increase the risk
of methemoglobinemia.
Increasing from 20 ppm iNO to 80 ppm
will not improve mortality, shorten
ventilation time or increase the PF ratio.
20 ppm is basically the maximal dose
with little effect above that level,
certainly not 80 ppm. At 80 ppm there
is an elevated risk of
methemoglobinemia or nitrogen dioxide

Notes 2014
toxicity. In 75% of patients
receiving iNO [with ARDS]
there will be a modest
improvement in their PF
ratio. The average increase
Continuous lateral rotational therapy
was
developed in the late 60s and employs
special beds that can only be rented
and not purchased. The rotation is 40
degrees in each direction and studies
[all funded by the bed makers] have
shown numerous beneficial effects
though no large, rigorous trial has been
performed.
47
in the PF ratio is 16 mmHg. There may
be a higher risk of renal replacement in
patient on iNO. It is argued that routine
use of iNO in ARDS patients should not
be done.
The strongest evidence base for NIV is with
AECOPD; Asthma has been disappointing. If a
COPD patient can protect the airway, but fails an
SBT, NIV can be attempted. Two general scenarios
one, you failed to correctly predict that a patient
would be liberated from invasive ventilation,
should the patient be re-intubated or tried on NIV?
There is data to support the use of NIV in the COPD
population. This jives with the data about NIV preintubation. But if the patient
NON-INVASIVE POSITIVE PRESSURE VENTILATION
What feature of NIV with pressure

supported breaths causes the most
problems? Flow cycling causes the
most problems. Pressure support
delivers a flow to a pressure target.
When the flow reaches a minimum
value, the breath terminates based on
flow cycling. The problem occurs
because of the leaks inherent in
facemasks and particularly bad if the
patient has a BPF [common board
scenario]. So breath delivery will be
prolonged. Sometimes triggering
mechanism may not be as sensitive.
There is no set inspiratory time with
pressure support.
Pressure support ventilation with

variable flow [like NIPPV] tends to
interact with the patient a little better.
Comparatively, pressure-ACV allows a
back-up rate and allows inspiratory
time selection. Having control of the Itime will allow the clinician to address
a leak. By contrast, NIPPV or PS, which
gives variable flow control, and the
patient controls I-time. The interface is
important, the larger the mask can
cause more problems in terms of leak.

Notes 2014
cannot protect their
airway, cannot cough or
clear secretions, then
back to el tubo.
Patients must be
monitored closely while on
NIPPV. Improvement in
PaCO2 and pH within 30
minutes to 2 hours
predicts success
[avoidance of intubation
and death]. A higher level
of consciousness at
presentation also predicted
success in one trial.
There is some evidence for
failed extubations and
heart failure. The end-point
is typically avoiding
intubation. Pulmonary
edema has fairly good
evidence in terms of CPAP
it improves gas exchange
and prevents intubation
[C3PO Trial in NEJM]. The
addition of inspiratory
support during pulmonary
edema seemed to make
people more comfortable,
but there was suggestion
that this may increase MI
risk, but this was probably
rapport between the patient and the
ventilator.
The complications of NIV are: leak,
discomfort, eye irritation, drying,
congestion, gastric insufflation,
hemodynamic compromise.
LIFE-THREATENING ASTHMA
Life-threatening asthma [LTA] is

defined as: respiratory arrest, need for
MV, pH less than 7.3 from retention,
LOC. Ultimately as a result of intense
airway spasm & mucous congestion.
48
a fluke; therefore, BiPAP is an effective
form of therapy during pulmonary
edema.
What about acute hypoxemic RF [e.g.
PNA, early ALI]. There isnt much data
supporting its use here. NIV does seem
to prevent intubation in the
transplanted and immunocompromised
for some unclear reason.
What about NIV in the chronically
hypercapneic patients? The thought is
letting the respiratory muscles rest
at night with NIV. Studies show that
the PaCO2 is lowered during the day,
but there are little good clinical data
otherwise [this goes for
neuromuscular, sleep disordered
breathing, chronic COPD].
Starting NIPPV. Start low, titrate
inspiratory pressure to patient comfort,
can start with 5-10 cm H2O of PS.
Dont go above 20 cm H2O, this blows
open gastro-esophageal junction.
Titrate PEEP [EPAP] per triggering and
patient effort.
PEEPi can be detected with an
esophageal pressure probe, and
inspiratory threshold load can be
seen. NIV requires additional time
from the RT, not really the RN. The
first 8 hours of NIPPV requires 100
minutes to establish a
How many people die of asthma in the US every
year? There are about 5000 deaths [ARDS is more
than 100,000]. There are 1.5 million ED visits for
asthma each year in the US. About one in 5 are
admitted and 4% of those admitted require ICU
level care. African-Americans have twice the
average mortality rate and ED/inpatient asthma
care and account for 50% of the total cost of
asthma.
There is rapid onset LTA [type 2] where people
essentially die within 2 hours. Slower onset [type
1] is more than 2 hours in presentation.
ASTHMA TREATMENT

Notes 2014
Bronchodilators IV or inhaled? MDI or
nebulization? Properly aerosolized
beta-agonists are better than IV. An
MDI versus a nebulizer are equally
effective if the patient can work with
the MDI/spacer. Remember that proper
MDI use requires expiration to FRC,
inhalation of the drug via a spacer to
TLC and holding for 10 seconds.
High volume nebulizers can deliver
higher doses with tidal breathing. The
standard dose by nebulizer is several
fold higher than the MDI, so the patient
can breathe tidally, the efficiency is
therefore quite low, but the patient can
breathe slowly.
What about dosing? Continuous usually
requires 5 mg per hour. There is
suggestion that continuous is slightly
better the patients should be closely
monitored.
49
What about anti-cholinergics? There is
reduction
in hospitalization, improved PEFR and
reduced costs with the use of inhaled
anti-cholinergics.
The use of steroids they work;
multiple studies have shown steroids to
be beneficial especially in severe
asthma. IV or PO? They both have
equal effects the bias is to give it IV,
but the literature shows that the oral
form kicks in at the same time as IV.
What about steroids by MDI? This
method is possible if the patient can
cooperate by MDI as described above
[holding for 10 seconds].
What about the use of oxygen? It is
helpful if low, but the PaCO2 can rise
as it does in COPD [BMJ. 2010 Oct
18;341]. Target saturations in the low
90s. The reason for the worsening
hypercapnia is due to: respiratory
muscle fatigue, diminished minute
ventilation from hypoxic drive,
worsening VQ mismatch from release
of hypoxic vaso-constriction, and the
Haldane effect. The biggest
contributor is the VQ mismatching. The
release of the hypoxic drive to breath
is a small and transient reason for CO2
retention in these patients. There may
be a boards question where a CO2
retainer is given zolpidem for sleep. It
is not the ambien that is causing the
carbon dioxide retention.
Magnesium blocks muscle contraction, there
is a Cochrane review favoring IV magnesium
in severe asthma.
What about the use of mechanical
ventilation for the patient failing the
aforementioned therapies? There are
special asthma problems on the
ventilator avoid VILI. With airway
obstruction there is heterogeneity of
time-constants, and therefore there
may be over-distension of normal parts

Notes 2014
of the lung. Therefore, treat
like ARDS to prevent overdistension. There is the
problem of PEEPi as well.
A respiratory acidosis rarely causes a
lot of harm,
and there is some literature to suggest
that it may improve outcome. Its OK to
have a respiratory acidosis if it protects
the lung. So sacrifice some minute
ventilation.
Triggering dys-synchrony can be
visualized by a delay between the drop
in the esophageal pressure and airway
pressure. What is the normal delay
between the pleural pressure and
airway pressure? It moves at the speed
of sound [MACH1] super fast. So
pleural and airway pressure should
drop at exactly the same time. The
application of extrinsic PEEP will not
change the total PEEP, but it will help
the trigger load.
What about heliox? It is a low-density
[but high viscosity] gas compared to air
and it may reduce PEEPi. It has never
been shown to improve outcomes. It
will foul the monitors as most monitors
are calibrated for air/oxygen and its
unpredictable. Heliox will decrease the
output of nebulizers because of a
decrease in particle size and therefore
inhaled mass of the nebulized drug.
Flow or dose of nebulizers must be
increased when combined with heliox
to maintain the same nebulized output.
The reduced density of helium will
convert some turbulent flow to laminar
flow [via Reynolds number] and also
reduce the driving pressure required to
move turbulent airflow itself. These
latter effects likely increase the
deposition of drug delivery via
inhalation and improved gas exchange.
Interestingly, carbon dioxide diffuses
better through helium than air which
aides in carbon dioxide excretion. There
h
50
All the pressures in the chest will rise
with PEEPi. This will cause the plateaus
to rise. It increases the triggering load
[see above discussions].
is evidence in spontaneously breathing patients
with severe airflow obstruction [COPD] that there
are clinical benefits with heliox [even though the
viscosity of Heliox is greater than air]. The greatest
improvement in flow [reduction in effective airway
resistance] occurs with a helium concentration of
40%, and flow increases linearly to helium of 80%.
Most departments have heliox- oxygen mixtures in
40-60% concentration for

Notes 2014
hypoxemic patients [higher
FiO2]. If the patient
is not hypoxemic [e.g. a
central airway obstruction]
then a 20% O2-80% He
mixture may be used.
One last mechanical
ventilation issue. Aerosols
are retarded by the
endotracheal tube. There
needs to be extradosing/very high doses in
the mechanically ventilated
patient. The MDI may be
used if the patient on the
vent is paused, but there
still needs to be more than
2 puffs, probably 6-10
puffs. Also, make sure the
MDI is in the inspiratory
limb so its not exhaled
away.
OUTCOMES OF LIFE-THREATENING
ASTHMA
Type 2 [rapid] needs more
steroids, fewer hours of MV,
lower death rate. Slow or
type I is the opposite. If you
51
are young and intubated for asthma, six
year mortality is quite low, but if you
are older survival is only 60%.
PULMONARY EMBOLISM
BASICS OF DIAGNOSIS & RISK
Stasis, vessel wall injury and

hypercoagulability are Virchows triad
flying over an ocean in economy class,
compared to business or first, there is
an increased risk of clot.
What is the relationship between
temperature and PE? The vast majority
of patients [86%] are typically without
fever, once above 103 degrees, PE was
found in less than 2%.
What about the ECG? A new RBBB is
concerning especially in a hypotensive
patient other features include:
S1Q3T3, right axis deviation, most
common feature is sinus tachycardia.
On CXR, a wedge-shaped pleural-based
infiltrate is known as Hamptons Hump,
there may be atelectasis with
parenchymal densities, enlarged right
descending PA [marker of PH],
decreased vascularity, cardiomegaly,
but also normal in up to 25%.

Notes 2014
Troponin levels, BNP levels do correlate
with
severity of RV dysfunction. Negative
troponins are a powerful predictor of
survival; so is a negative BNP. In
contrast to a positive troponin, a
positive BNP does not predict poor
outcome. In a patient without
underlying cardiopulmonary disease,
50-60% of the pulmonary vasculature
must be occluded before pulmonary
hypertension arises, this is why there
may be a totally normal RV on echo and
by ECG. So a normal right heart on TTE
does not rule out PE, it may suggest PE,
or it may diagnose or exclude
alternative etiologies. RV dilation, RV
HK, shift of the IV septum, dilated PAs,
dilated IVC may be seen on TTE. If
there is hypotension from PE, there
must be RV findings.
A normal ABG does not rule out a PE.
38% of angiographically documented
PE, had completely normal ABG.
The d-dimer [latex agglutination is the
old one] the new rapid ones are useful
if the d-dimer is not elevated.
The use of end-tidal CO2 may be
helpful is there is a sudden, abrupt drop
in ETCO2 which suggests an acute PE
from the loss of cardiac output;
successful thrombolysis of the PE will
result in an increase in ETCO2.
Ventilation perfusion scans are of good
value in non-intubated patients with a
normal CXR. Looking for large,
segmental defects in perfusion [6 view
study is what is needed]. Most V/Q
scans are of intermediate probability.
Computed tomography of the lung has
multiple names [e.g. CAT scan, CT scan,
CT angiogram, spiral CT, helical CT,
multi-detector CT]. The CT scan may
detect clots into the segmental
arteries. The sub-segmental arteries
are more difficult to see, but clots can
4
h
be there as well. Treatment of sub-segmental PE is a

controversial topic.
Negative studies are negative when combined with
other clinical /imaging parameters. PIOPED II
evaluated CT angiogram, if angiogram positive

Notes 2014
then treat, if negative, but
high clinical pre-test
probability then there is
diminished predictive
power.
A negative bilateral LE US
does not rule out PE. A
majority of patients with PE
by CT have negative LE
ultrasounds.
27% of patients had CT
angiogram contraindications [high creatinine
is the most common
cause]. PIOPED III looked at
MRA/MRV for PE it is
technically inadequate for
the diagnosis of PE. Very
short breath hold, new MRI
technology may improve
this.
TREATMENT
It is safe to withhold
treatment is there is a low
pre-test probability and
the d-dimer is negative.
AC with heparin should be
used for at least 4-5 days
as transitioned to
Coumadin regardless of
the INR [the INR will
become therapeutic prior to Coumadin

protection].
100 mg of tPA IV over 2 hours is the

dose for treatment of PE if using
thrombolysis. Indications for tPA is
persistent hypotension, severe,
persistent hypoxemia maybe. There is
expert opinion for front loading a
patient who is actively dying that is
front load with 40-60 mg and then the
other half over 2 hours.
The use of half-dose for sub-massive
PE is still controversial. Based on the
small MOPPETT II trial, it may improve
long-term RVSP by TTE. The [ACCP]
2012 consensus guidelines suggest
systemic tPA in sub-massive PE
patients only if there is a high risk of
developing into massive PE [a grade 2C
suggestion based on low quality
evidence/expert consensus]. There
have been a couple of recent trials
including: TOPCOAT
[U.S.] which is a much smaller study
compared to PEITHO [European - 12
times larger with 1,000 patients]. In
TOPCOAT there was no mortality
benefit as also seen in the PEITHO
study, but the sample size was too
small to make definite conclusions
about mortality or bleeding risk. The

Notes 2014
follow up was little longer (90 days vs
30 days in
PEITHO), but it difficult to recommend
systemic tPA in sub-massive PE unless
patients are in impending
hemodynamic collapse.
54
The CAP pathogens tend to be somewhat esoteric
on the board examination; CXR findings can be too.
Recognize that the bulging fissure sign is not
pathognomonic for klebsiella and that antibiotic
coverage should not be tailored based on
Mechanical disruption or surgical

interventions may be considered, when
the patient has contraindication to
medical thrombolysis.
The IV septum can be compressed
during PE. The IV septum may be used
as a marker of fluid resuscitation. Keep
the aortic diastolic pressure adequate
which is the coronary pressure head to
the right heart. Inotropes should be
used, less fluids [limit 500 cc to 1L].
Always think about filter placement,
upper extremity clot is felt to be less of
a problem for hemodynamic instability.
When IVC filter? Contra-indications to
AC, onset of bleeding, ongoing clots,
hemodynamic instability in patients
who will not be given tPA. Patients with
filters should be anti-coagulated when
possible; returning to an active lifestyle is important as filters increase
DVT risk. Also, retrievable filters
should be considered. 40-120 days
may make removal difficult.
In calf-vein thrombosis, only AC if there
are symptoms. If no symptoms, and no
longer term risk, probably no
treatment, consider serial US exams.
For upper extremity thrombus it is now
recommended to provide AC! Not
necessary to remove catheter, as you
can treat through the catheter. The
patient may ambulate with PE/DVT,
unless there is symptomatic DVT, then
this should resolve prior to ambulation.
COMMUNITY-ACQUIRED PNEUMONIA FOR THE BOARDS

Notes 2014
radiographic findings.
Consider a patient with
severe sepsis; broad
spectrum antibiotics
should remain until the
patient improves. The
bulging fissure sign is
most commonly seen in
klebsiella pneumonia, but
can commonly be seen in
streptococcal pneumonia
as well. Further it has
been reported in
acinetobacter infection as
well as TB, H flu and
Yersinia pestis!
RSV
RSV has significant
morbidity and mortality in
elderly and transplant
patients. There is one
serotype with two
subtypes A/B. RSV is
highly seasonal and
reinfection is
exceptionally common.
Fomites are the way it is
transmitted and there are
hospital and ICU
outbreaks.
There may be prolonged shedding in

severe
disease especially in those with comorbidities, those on steroids, and
immunocompromised hosts who may
shed for weeks to months.
About 25% of patients with influenza
develop pneumonia, fever is nearly
ubiquitous. Influenza itself can
uncommonly cause shock. The RT PCR
for influenza is the most rapid and most
specific. All means of testing can detect
A and B types.
Individuals can be infected more than

once during the same virus outbreak.
The DFA or EIA for RSV is not very
sensitive. The rapid antigen test is
terrible in adults, the culture is slightly
better [40-65% sensitive] but takes
days. Antigen tests of the upper
respiratory tract can be negative while
BAL positive. RT-PCR has high
sensitivity.
55
Severe disease may occur in

immunosuppressed patients like lung
transplant or bone marrow suppression;
early treatment of URI may prevent
progression to viral pneumonia.
Secondary pulmonary co-infection with
bacteria uncommonly occurs with
RSV.
In selected adults, ribavirin may be
helpful, Immunotherapy with IVIg may
be helpful. But, inhaled ribavirin is a
challenge to deliver and IVIg is
expensive and has not been shown to
change outcome.
INFLUENZA
Influenza can be spread by large
droplets within 6 feet. Certainly hand
contact. Aerosol from NIPPV may spread
influenza. Incubation is 1-4 days, virus
shedding may occur 24 hours prior to
symptoms and through days 5-10 of
symptoms.
Rapid tests are immunoassays that are positive or
negative, but sensitivity and specificity are not
great. Sensitivities range from 10-80%!
Specificities much better 90-95%. So a negative
RIDT does not exclude influenza.
Rimantadine is not effective against B influenza.
Zanamavir gets both types. Oseltamivir and
Zanamivir are the treatments of choice. 300 elderly
patients with influenza in Canada. Those who went
to the ICU had a 50% mortality. Early use of
oseltamivir and early ED visit improved outcomes.
Do steroids work in the severely ill with influenza?
There were 5 studies done retrospectively with

Notes 2014
many confounding factors that favor
not giving steroids. Primary influenza
pneumonia and secondary bacterial
pneumonia are the big complications.
Pregnant women and obese patients are
at risk for complications as well as the
common causes for bad outcome [e.g.
old, cardiopulmonary disease, comorbidities]. Mortality in 2009 H1N1
pandemic was 17%, 39% if on ventilator.
Shock was also quite common. The
pandemic influenza stains affect the
entire tracheobronchial tree [1918 H1N1
and 2009 H1N1], but the seasonal
influenza does not do this.
FUNGAL PNEUMONIAS
Recognize and treat blastomycosis. The
scenarios will often describe a persistent
pneumonia unresponsive to multiple
rounds of antibiotics with a specific
geographic locale. Consider a
56
patient in Chicago with pulmonary and CNS
symptoms. There is little on history,
however, to differentiate blastomycosis
from histoplasmosis other than blastos
penchant for the CNS [as well as skin
and bones]. Blasto clues are often skin
lesions or infection of ones dog. Blasto
itself in the early days was known as
Chicago disease. Stiff neck is rarely
seen in fungal meningitis and this total
clinical picture is somewhat consistent
with TB, though the rapid progression of
symptoms argues against it.
Blastomycosis
and Histoplasmosis do not cause
peripheral eosinophilia and blasto is a
budding yeast.
Blastomycosis is asymptomatic in 50%.
It is treated with itraconazole and
amphotericin depending on severity.
One answer option may be caspofungin
for treatment, but echinocandins are
not adequate for the treatment of
endemic fungi or aspergillus. Caspo
should really be reserved for bad
candidemia. Further caspo does not
enter the CSF.
Histoplasmosis is without symptoms in
more than 95% but can cause a plethora
of symptomatic syndromes including
acute and chronic pneumonia, acute and
chronic disseminated disease, fibrosing
mediastinitis and cavitary pneumonia.
Like blasto, histo is treated with
itraconazole or amphotericin depending
on severity.
Sporotrichosis causes dermatitis in
immuocompetent patients and
pulmonary and disseminated disease in
immunocompromised patients.
Pulmonary sporotrichosis usually
causes upper lobe cavitary disease [so
too can blasto and histo].
Another scenario may be a patient who
has traveled throughout Southern
Ontario, Tennessee and Alabama
[blasto], Iowa and Ohio [histo] and

Notes 2014
Arizona/California [cocci]
and develops bilateral
pulmonary infiltrates and
respiratory failure after
clinch the diagnosis of
coccidioidomycosis. Cocci
is typically asymptomatic, but can
cause a CAP like picture. Up to onethird of CAP syndromes in endemic
areas are actually due to this endemic
mycosis. Asians and African Americans
have a higher risk of disseminate
coccidioidomycosis.
Cocci is treated with fluconazole or
ampho depending on severity. It does
not respond to echinocandins.
TUBERCULOSIS
Fun facts: the rate of TB in the UK is
about 4 times that in the US, surgery
may be used as an adjunct for resistant
TB and HIV is the most important risk
factor for progression of latent to active
TB.
1.7 million people die each year in the
world from TB. In the UK there is much
easier travel from high risk populations.
Sick patients with TB in the ICU have a
very high mortality they die of
multiple organ failure and shock. Note
that disseminated TB in the ICU can
present with gram positive rods in the
blood, because TB is well-known to
enter the blood [thats how it
disseminates] and it is a gram positive
rod.
The normal response of the body is to
form a granuloma with central caseous
necrosis and this sits in a precarious
balance and is altered by nutrition,
immune state, drugs, HIV, etc.
HIV is the fuel on the fire of
tuberculosis. HIV patients cant really
form granuloma and have higher
burden in the airway, they cavitate and
collapse into the lung. HIV is the most
being given a course of prednisone. If

there is peripheral eosinophilia and/or
large spherules on lung biopsy, these
57
important risk factor in the US for progression to

active disease.
Culture is the gold standard for diagnosis - AFB
smear provides an indication of infectiousness, and
TST and IGRA cannot distinguish between active and
latent disease. The TST depends upon degree of
induration. The QFT gold depends on lymphocytes
from host reacting to known antigens of TB. IGRAs
are likely better in those

Notes 2014
who received BCG. In active
TB, the TST and IGRA
are only about 70%
sensitive.
50-80% of active TB have
positive smears. Traditional
cultures take 3-8 weeks,
broth can grow 1-3 weeks.
Molecular probes can be
applied to any tissue. They
have very high specificity
but sensitivity varies with
burden of organisms. What
about the use of nucleic
acid amplification [NAA]
tests? Always use the NAA
tests in conjunction with a
smear. If both positive, this
is very predictive. If NAA
positive but smear
negative, presume TB
pending culture if 2 or
more NAA positive. If NAA
negative, smear positive,
this is likely a false
negative NAA due to
sputum inhibitors that
prevent NAA amplification.
If both NAA and smear
negative, its hard to say.
NAA detects 50-80% in
smear negative but culture
Who should be tested? Enigmatic
pneumonias,
compromised hosts, during outbreak,
those who fail treatment with betalactam, travel history within 2 weeks
and nosocomial PNA of unknown
etiology. Antigenuria is 60-95% sensitive
and highly specific. Nosocomial
outbreaks are 50-60% sero-group 1 [as
opposed to travel where it is essentially
all serotype 1]. Azithro, levoflox or
moxiflox are therapies. There is data to
support levofloxacin as superior to
azithromycin; there is an increasing
incidence of azithromycin failure in
legionella.
58
positive. Basically a single negative NAA
is not definitive to exclude TB if there is
a moderate to high clinical suspicion.
The initial therapy is 4 drugs IRPE.
Second line therapies are serious. MDRTB resistant to at least IR, XDR-TB
resistant to IR, any quinolone and at
least one injectable [amikacin,
kanamycin, capreomycin]. The highest
rates of these bad bugs are in the
former Soviet Union and China.
LEGIONELLA
This often occurs in outbreaks or after
returning from an enclosed space such
as a cruise.
Legionella comes from the environment
and may affect normal and abnormal
hosts. There are 16 serotypes and 7090% of disease are from serotype 1.
The gram stain has polys but no
organisms.
Culturing L. pneumophilla requires nonroutine media, urine antigen may
persist for days after anti-microbial
therapy is begun, only serotype 1 [70
or so % or L. pneumophilla] is detected
by the urinary antigen test, but
importantly, once the urine test is
negative, a full treatment course must
still be undertaken.
ANTHRAX
Anthrax is a disease of herbivores, and transmission
to humans by contact with infected animal or animal
part. Bacillus anthracis was previously very common
in livestock before a vaccine was developed by
Louis Pasteur. Human disease is consequently rare
save for areas of the world where vaccine is rare. It
is a gram positive rod, non-motile. Human infection
is cutaneous, GI or inhalational.
Cutaneous anthrax occurs when the spores get
under the skin with a 20-30% death rate.
Cutaneous anthrax begins as small papules that
progress to deep ulcers and then eschars, there is
regional lymphadenopathy. In 2009 there was an
outbreak in Scotland from cutaneous heroin

Notes 2014
injection. These people died quickly in
resistant shock with huge fluid
requirements. GI anthrax is from
ingestion of contaminated meat. GI is
more lethal with severe inflammation
and necrosis of the gut 50-100%
mortality.
Inhalational Anthrax is more of a
bioweapons concern. The incubation
period can be up to one month. When
inhaled spores cause disease, lots of
badness happens. The spores are
transported to the local lymph nodes
where the toxins are produced. This
leads to hemorrhage, lymphedema,
systemic unrest, shock and death.
Hemorrhagic mediastinitis and pleural
effusion are the most common thoracic
manifestations,
59
with CXR demonstrating a wide
mediastinum, but
a fairly normal parenchyma. Those
patients who progress to shock
frequently have non- cardiogenic
pulmonary edema [ARDS picture] as
well as meningeal signs with meningeal
hemorrhage being common.
Bronchospasm is not reported.
The mortality is 100% untreated and
90% treated. There is no risk of personto-person transmission so standard
isolation is required. Pulmonary anthrax
does not require respiratory isolation,
because it is contracted by spore
inhalation from the environment. Health
care providers do not require PEP, spores
are very hardy in the environment and
are not inactivated by drying or sunlight.
The therapy is cipro or doxy plus one or
two additional agents [rifampin,
chloramphenicol, clinda, pen or amp or
vanc, imi, clarithro], consider steroids
for severe edema or meningitis. The
vaccine should be administered
subcutaneously at diagnosis and 2 and
4 weeks later. Post vaccination Ig is
available for anthrax when
documented. Typically those
prophylactically vaccinated are in the
army.
THE PLAGUE
What about the plague? It is caused by
Y. pestis. The plague is common in the
SW of the US and transmitted by fleas.
Bubonic plague has a 15% mortality. It
is a small gram negative coccobacilli
with a safety pin appearance. Bubonic
plague presents with a cervical bubo,
petechiae if it progresses to septicemic
plague; blood cultures must be
specifically looked for in the
microbiology lab. Bubonic plague can
progress to pneumonic plague.
Pneumonic plague does not require
many organisms to get sick. 70%

Notes 2014
mortality, there is
person-to-person
transmission so the
person droplet transmission is a less
common
means of contracting pneumonic
plague.
Bubonic plague differential is staph,
strep, glandular tularemia, cat scratch
disease. Pneumonic plague differential
is anthrax, tularemia, meliodoidosis,
CAP, flu, hantavirus, hemorrhagic
leptospirosis. Septicemic plague
differential is also broad gram neg,
meningococcus, RMSF, TTP.
Treatment requires streptomycin,
gentamicin or doxy, cipro,
chloremphenicol, patient must be in
respiratory if pneumonic plague
considered and for these patients,
health care providers require postexposure prophylaxis with doxy, cipro,
chloramphenicol, TMP. Health care
workers do not need IVIg PEP.
DAH SYNDROMES
The definition of massive hemoptysis

varies between 100 and 600 cc of BRB
per day, but most accept 200 mL in 24
hours. The management includes
airway stabilization with intubation if
needed, cough suppression, coagulation
correction, sometimes antibiotics and
localization with a bronchoscopy or CT
scan with angiography. These patients
often take a kidney hit from all of the
dye that they receive to localize and
coil the bleeding. Setting them up for
dialysis might be indicated preemptively. There are multiple, multiple
causes of hemoptysis.
Recognize
that
heart
failure
[especially from mitral stenosis]
can
present
with
significant
hemoptysis.
patient should be in respiratory

isolation. Person-to-
60
Consider a patient admitted with insidious onset of

dyspnea and then is intubated as his hemoptysis
worsens. An ECG reveals RVH with LAE and CXR
shows Kerley b lines with mediastinal haziness and
lower lobe infiltrates, large pulmonary arteries and
a straight left heart border. There is also a double
density sign with splayed carina both suggesting an
enlarged LA.

Notes 2014
The diagnostic test is a TTE
because this patient
has mitral stenosis. This
patient was found to have
DAH on BAL, but this is
due to rupture of
hypertensive pulmonary
veins rather than capillary
inflammation. The
difference between mitral
stenosis and pulmonary
veno-occlusive disease
would be the size of the
left atrium. In North
America, severe MS is
often due to subclinical,
childhood rheumatic fever
which progresses
insidiously over decades.
Other causes of DAH
should be considered
much less likely given the
degree of
cardiopulmonary
abnormality on CXR and
ECG. It would be quite
strange to see this picture
in a patient with ANCApositive vasculitis or
catastrophic APLS or
inhaled cocaine all of
which may present with
DAH. Isolated pulmonary
capillaritis which is a
small vessel capillaritis
syndrome was described following
influenza
infection. Hydrocarbon exposure may
also increase this risk, especially
smoking. GPS has been known to
relapse with resumption of smoking.
Hemoptysis is the most frequent
presenting symptom though there are
exceptions
some present only with renal
symptoms and most patients on
presentation have an active urine
sediment. Renal biopsy is the goldstandard for diagnosis and linear
confined to the lungs can only be

detected by surgical lung biopsy.
61
Recognize that smoking crack can

cause diffuse alveolar hemorrhage.
Consider a young patient with sudden
onset disease and hemoptysis. The
CXR reveals bilateral alveolar
infiltrates in the lower lobes. There is a
slight 19% peripheral eosinophilia. A
big clue is melanoptysis coughing up
soot in these board questions.
Importantly the UA is normal, which
essentially rules out the pulmonary
renal syndromes. As an aside, crack is
extracted from cocaine via an ether
solution. It is called crack because it
pops when you heat it.
Understanding Goodpastures syndrome
[GPS]. Consider a young patient who
smokes who presents with hemoptysis
and nephritic syndrome on a UA. The
anti-GBM antibody is positive and so too
is p-ANCA. It is important to note that
the absolute value of the GBM antibody
does not correlate with disease
severity, though its level can be
followed to measure treatment. A
subgroup of patients with GPs
syndrome also have positive p-ANCA
and these patients are more likely to
have multi-organ involvement. The
trigger for GPS is unknown, but the
original
staining is seen as opposed to the granular staining
seen in other nephritic diseases. The treatment for
GPS is plasmapharesis, cyclophosphamide and
steroids. This is dropped mortality to less than 20%.
Recognize that negative pressure pulmonary
hemorrhage [NPPH] can mimic DAH. Consider a
patient who receives anesthesia via an LMA and
then goes into respiratory distress in the recovery
room with hypercapnia and blood in the LMA and
then ET tube. His CXR shows patchy consolidation
much more in the right with volume loss and a
bronchoscopy shows blood everywhere with
increasing blood on lavage. He likely developed

Notes 2014
NPPH trying to breathe around a
misplaced LMA. Intubation corrected
this.
Pulmonary capillary stress fracture
occurs in thoroughbred horses and can
occur in human athletes. During
exercise, the trans-capillary pressure
can reach 35 mmHg in people!
ventilation, but with a clinician-set

decelerating
flow waveform. In this situation, the
flow will not vary and there will not be
a square-wave pressure tracing.
62
There are variations of pressuretargeted mechanical ventilation. One

such variation is known as PRVC, autoflow or volume control plus which
essentially allows the ventilator to alter
VENTILATOR WAVEFORM SCENARIOS
inspiratory flow and pressure to achieve
Recognize pressure assist control
a given volume. When this physiology is
ventilation based on the ventilator
applied to pressure support it is known
waveforms. There is a machine and
as volume support. Another variation
patient triggered, pressure-targeted
of PACV is known as BiLevel, BiPhasic or
[there is a square wave pressure
Airway pressure release ventilation
waveform] and time-cycled [each
which allows for spontaneous breaths
breath is the same length] breath.
on top of a prolonged inspiration.
Note that flow and volume are varying.
Consider a COPD patient who is
In the patient-triggered breath, flow
switched to pressure support with
increases to maintain the square-wave
dropping volumes and dyspnea, what
pressure. Also note that the value of
to do? With airway obstruction, and
flow at which the breath cycles varies
pressure support, the pressure target is
meaning that it is cut short based on
hit quickly. Then as flow gets passed
the time-cycle. The mimic of this mode
the obstruction, it tapers slowly. So
would be volume control with
breaths become very long and there is
volume/flow-limited
air-trapping. Shorten the I- time [also
true with NIPPV].
Recognize auto-PEEP. The presence of
expiratory flow prior to machine
inflation indicates dynamic
hyperinflation and intrinsic PEEP. Since
intrinsic PEEP occurs with increases in
airway resistance and/or rapid
respiratory rates, sedation and
paralysis is a good therapeutic choice.
Additionally, decreasing airway
resistance with medications, increasing
rather than decreasing inspiratory flow
rates to allow more time for exhalation
(ie, decreasing I:E ratio) would also
work.
Understand pulmonary mechanics and
waveforms including the esophageal
pressure [Pes] tracing. The patient has
developed pulmonary edema with high

Notes 2014
peak pressure. You must
know that this high peak
pressure is the result of
poor pulmonary
of poor chest wall from poor pulmonary
compliance. A high Pplat in the setting
of a high Pes means that chest wall
compliance is impaired. Whereas as
high Pplat with an unchanged Pes
means that pulmonary compliance is
worsened [e.g. pulmonary edema].
Understand ATC on the ventilator. What
ATC is doing is trying to achieve a
smooth, pressure, square wave. But the
resistance of the tube adds to this such
that a square wave at the tube opening
may be more triangular in the proximal
trachea. It calculates the pressure
waveform distortion based on the
length and resistance of the
endotracheal tube. It is done to add
comfort.
63
compliance and therefore a high
plateau pressure. Then you are also
given Pes so you can differentiate the
effects
ferrous iron on Hb is oxidized to the ferric state, the
Hb curve is shifted leftwards such that tissue
hypoxemia can be profound as the Hb will not
unload oxygen.
Other agents can cause Met-Hb as well including
dapsone, primaquine, and nitrates including iNO.
The incidence of Met-Hb may be as high as 1 per
1000 TEEs that use benzocaine. Pulse oximetry is
totally inaccurate when measured oxygen
OTHER PULMONARY BOARD EXAM SCENARIOS
Understand leukocyte larceny. The

patient has a WBC count of nearly
300K and a PaO2 in the 30s. Normally,
placing an ABG on ice will inhibit
oxygen consumption by thrombocytes
and leukocytes, but in extreme values,
even rapid placement on ice cannot
inhibit oxygen consumption. The only
way to accurately measure PaO2 is to
add an inhibitor of oxygen consumption
such as potassium cyanide or NaF. The
former is better studied. Heparin must
be added as well.
Recognize and treat
methemoglobinemia. The patient
receives topical cetacaine for a TEE
and has falling oxygen saturations to
the mid-80s. Her membranes are
cyanotic and ABG reveals blue blood
with a PaO2 over 100 mmHg. She
needs IV methylene blue. When

Notes 2014
saturation in the setting of
Met-Hb and typically
reads in the mid-80s.
Iatrogenic met-Hb can lead
to abrupt, profound
symptoms including
cyanosis, convulsions and
death. Congenital forms of
Met-Hb can lead to
asymptomatic cyanosis.
The dose of methylene
blue is 1-2 mg per kg IV
over a few minutes.
Excessive methylene blue
can worsen symptoms and
produce hemolysis
[especially those with
G6PD deficiency].
Methylene blue further
impairs pulse oximetry to
detect oxygen saturation
so it should be discarded.
Recognize and treat
tropical pulmonary
eosinophilia. The patient is
a young woman from India
and is recently diagnosed
with asthma, though she
continues to be short of
breath. She has enhanced
bronchovascular markings
on CXR and lower lobe
mottled opacities. She has
64
a marked leukocytosis 33K with a
prominent peripheral eosinophilia. Stool
for O and P are negative twice, and IgE
is markedly elevated and ABPA titres
are negative. She likely has been
sensitized to either W. bancrofti, Brugia
malayi, Brugia timori which are
parasites that live in Africa and
Southeast Asia. Mosquitos are the
vectors and the larvae travel via the
lymphatics to the pulmonary system
and cause congestion and eosinophilia.
There are 6 diagnostic criteria for TPE:
1. Residence in an endemic area 2.
Insidious onset over weeks to months 3.
Prominent evening symptoms 4. Marked
peripheral eosinophilia 5. Markedly
elevated IgE and 6. elevated anti-filarial
antibody. The treatment is that of
asthma and diethylcarbamazepine [150
PO BID for 3 weeks].
Understand the treatment of
decompression sickness complicated
by arterial air embolism. The patient
ascends too quickly from 20 feet below
when he is frightened. He does so
against a closed glottis. During the
rapid ascent, the nitrogen bubbles
expand, and when done so against a
closed glottis, the alveoli rupture and
cause barotrauma. Further, this
patient also has
h
a headache, seizes and then has decerebrate
posturing indicating that the air also
entered his pulmonary veins and then
traveled to the left heart and brain.
Treatment of central arterial air
embolism classically involves putting
the patient in the left lateral decubitus
position and in trendelenberg to trap
the air bubbles in the right atrium but
this technique is questioned.
Hyperbaric therapy with 100% oxygen
should certainly be considered to
reduce the amount of air within the
vasculature.
Treat fusobacteria necrophorum
infection. The patient presents with a
pharyngitis and then progresses to liver
abscesses and iliac crest inflammation.
From this you are supposed to know
that this is Lemaires Syndrome which
is characterized by pharyngeal
infection, septicemia with rigors 4 to 5
days after the local infection, lateral
neck tenderness and swelling,
metastatic abscesses (especially to the
lung), suppurative arthritis, jaundice,
and renal failure. F. necrophorum is a
gram negative anaerobe that requires
treatment with prolonged clindamycin,
surgical incision, and anticoagulation.
The pathogenesis is unclear but
involves an inciting pharyngeal
infection of the pharyngeal or
peritonsillar space, septic
thrombophlebisits and metastases to
various organs commonly the pleural
space, kidneys, liver and bone. It
usually occurs in very young people.
Overwhelming septicemia and death
can rarely occur.
49
3.There
CRITICAL
CARE NEPHROLOGY
ACUTE
KIDNEY INJURY
are various means of
more than 5 per LPF, the likelihood is very high
classification of AKI. There is the RIFLE
classification [Risk increase Cr by 1.52 x or oliguria for 6 hours, Injury
increased Cr by 2x or oliguria by 12
hours, Failure
increase in Cr by 3 x or oliguria for 24
hours or anuria for 12 hours, Loss
which is persistent AKI with complete
loss of kidney function for more than 4
weeks, ESKD which is complete loss for
more than 3 months]. But there is also
AKIN stage 1-3 which are kind of similar
but should be within 48 hours because
the AKIN only applies to acute kidney
injury. AKIN stages 1-3 correspond to
the RIF of RIFLE.
for ATN. So muddy brown casts are very

helpful for diagnosing ATN. This study then
went on to describe a sediment score.
CLASSIC FRAMEWORK AND URINARY INDICES

The classical framework is pre-renal,
post-renal and intra-renal. Note that
pre-renal does not mean hypovolemic;
it means salt-avid physiology which
may occur in any state of fluid balance.
The 3 key tools beyond the history are:
a renal US, UA and urinary indices. In
pre-renal disease, the UA should be
bland with maybe hyaline casts
[Hyaline casts are formed by Tamm
Horsfall protein]. In ATN, there may be
tubuloepithelial cell casts or muddy
brown casts or coarse granular casts.
By contrast, if there are lots of RBCs,
RBC casts, WBCs [sterile pyuria] and
especially WBC casts think of AIN.
In one study the density of muddybrown casts was used to predict ATN. In
the absence of granular casts per LPF,
the likelihood of ATN is very, very low. If
www.heartlung.org
JE
K

Review Notes
The higher the sediment score,

the worse the AKI stage. The
sediment score was based on
RTE [renal tubular epithelial
cells] per HPF, and number of
granular casts per LPF.
When you start looking at patients in

whom there have been interventions
[including saline infusion], or in patients
who are not oliguric then the value of
these indices are greatly degraded.
FeNa cutoffs are 1 and 3%, FeUrea is 35
and 50% for pre-renal physiology
versus intra-renal physiology as the
culprit for AKI.
Consider a patient with AKI in

the ICU with a creatinine bump
on pip-tazo, valsartan and
plavix; there were 20-25 WBC
with minimal protein and blood.
This is likely an interstitial
nephritis. The UA is benign with
valsartan and Plavix can cause
a micro-angiopathy in the
kidney.
DRUG-INDUCED KIDNEY DISEASE

The spectrum of drug-induced kidney
disease within the ICU is wide.
Systemic capillary leak with pre-renal
azotemia may occur, there may be
changes in intra-renal blood flow [e.g.
ACEI, ARB, NSAIDs including COX2,
tacrolimus, vasopressors, contrast,
exenatide].
Urinary indices are helpful but

have caveats FeNa and
FeUrea. The clinically valuable
urinary electrolytes lie in the
patient with oliguria [without
CKD], first presenting to the
hospital.
lung.org
There may be direct glomerular injury

from NSAIDs, gold, penicillamine,
captopril, pamidronate, all interferons
or microangiopathy from ticlopidine,
Plavix, OCP, gemcitabine, mitomycin C.
JE
K

Review Notes

Notes 2014
Tubulointerstitial injury may be a
consequence of
direct toxicity, classically this is
contrast, aminoglycosides [AGs] a
cumulative effect after 5-7 days of AG
therapy, vancomycin, amphoB,
pentamidine, cisplatin, tenofovir,
bisphosphantes, osmotic
nephropathies such as hetastarch, IVIG,
mannitol, but also AIN such as betalactams, quinolones, sulfonamides,
vancomycin, PPIs may be the most
common cause of AIN, allopurinol,
diuretics.
Crystal deposition can occur from MTX,
acyclovir, indinavir, atazanavir,
sulfadiazine, TMP-SMX. Acyclovir
toxicity tends to present with
neurological depression and renal
failure with needle-shaped crystals.
Acyclovir neurotoxicity usually presents
within 24 to 72 hours of initiation of the
drug; signs include changes in
cognition, level of consciousness, action
tremor, multifocal myoclonus, asterixis,
hallucinations, and delusions. Seizures
may occur and the dysfunction can
progress to coma [usually with an
abnormal electroencephalogram]. The
treatment is cessation of the drug and
use of hemodialysis [effective because
of limited protein binding of the drug
and its low molecular weight, 225
Daltons] which shortens the duration of
toxicity. By contrast, uric acid
nephropathy may be seen in a
recurrent multiple myeloma patient,
just started on salvage chemotherapy
with AKI. Normally, uric acid
nephropathy doesnt occur unless
serum levels are well above 15, but
para- protein nephropathy from
myeloma can cause uricosuria which
raises the risk. The real kicker is the
presence of rhomboidal or rosetteshaped crystals in the urine which are
68
uric acid [note, uric acid is needle-shaped in
arthritis!] The treatment is rasburicase.
RP fibrosis from methyldopa can lead to urinary
obstruction.
CHOLESTEROL EMBOLI
Recognize cholesterol embolic phenomenon.
Consider a patient with an infrarenal stent placed

Notes 2014
for a AAA and two weeks
later develops profound
weakness, fluid overload
and a creatinine over 12 and
a potassium of almost 8.
There may be livido
reticularis and peripheral
blood eosinophilia [more
than 80% have
eosinophilia]. This
embolization can occurs
days to weeks following the
manipulation of the aorta.
Anticoagulation in these
patients has been
associated with accelerated
atheroembolism and should
not be used. There is no role
for thrombolysis in the
treatment of
atheroembolism, but in
thromboembolism of the
kidneys, within 6-12 hours
there may be a role. The
prognosis of atheroembolic
disease is quite poor. Less
than 50% who require
dialysis recover renal
function and mortality at
one year is 22-80%.
HEPATORENAL SYNDROME
Prevention of HRS is important - teat
SBP in the
patient with cirrhosis and ascites
promptly! The patient should receive
antibiotics that are renally dosed
because of AKI. Patients with SBP, who
also have a serum creatinine >1
mg/dL, blood urea nitrogen >30
mg/dL, or total bilirubin >4 mg/dL
should receive 1.5 g albumin per kg
body weight within 6 hours of
detection and 1.0 g/kg on day 3.
Evaluate acute kidney injury in the
setting of decompensated cirrhosis
with large ascites. Consider a patient
who was aggressively diuresed but his
creatinine continued to rise. His urine
Type 1 is the acute, fulminant form,

classically with a precipitating factor.
Type 2 is a smoldering, insidious type
Cr 1.5-2.5 range that doesnt change
much. The diagnosis requires a Cr
more than 1.5 or GFR less than 40, no
shock, lack of improvement after
stopping diuretics and with plasma
expansion, no nephrotoxins, no
proteinuria, & no obstruction.
69
Classic type I HRS typically has a urine

Na less than 10, oliguria,
hyponatremia. Lack of infection is no
longer a criteria as infection is the most
common precipitant, other triggers
include large volume paracentesis
without albumin, aggressive diuresis, GI
bleeding, new liver insult [e.g. ethanol,
hepatitis].
How is HRS managed? Albumin
replacement often with invasive
monitors, liver transplant if eligible.
Vasoconstrictors +/- TIPS. Systemic
vasoconstrictors are often tried
midodrine with octreotide but the
data is marginal. There is reasonable
evidence that IV infusions of NE are
helpful, the best evidence is
vasopressin and terlipressin [not
available in the US].
sodium is 38 with blood and trace protein in the
UA. Hepatorenal syndrome is diagnosed when the
patient is removed from diuretics for 2 days and
receives 1 gram per kilogram per day of albumin
and there is no improvement in creatinine. Urine
sodium should not be used to differentiate HRS
from pre-renal azotemia because their physiologies
are essentially identical. Further, diuretics given in
house raise urine sodium. Large volume
paracentesis should be tried only after albumin is
given. Octreotide and midodrine can be tried to
treat HRS after the diagnosis is made, but firstly
albumin must be administered to rule out a
hemodynamic culprit.
MANAGING AKI IN THE ICU

Notes 2014
Most cases of AKI in the ICU have no
specific intervention. There are general
supportive measures: avoiding further
injury, adjust medication doses,
manage nutrition, electrolyte and acidbase balance, correction of anemia and
coagulopathy.
The most common cause of AKI in the
ICU is ATN. Therefore intelligent use of
diuretics is required. Remember that
GFR only applies to the steady state, so
acutely changing creatinine cannot be
used to calculate a GFR, it could be
much less than 10 for example is the
creatinine increases from 1.0 to 2.0,
even if the MDRD GFR is calculated as
much higher.
70
What is the intelligent use of diuretics
in AKI?
The PICARD study suggest that
mortality may be worse with diuretics.
High dose diuresis can lead to deafness
by 4x. Use diuresis if the patient is
volume overloaded but do not waste
time delaying other interventions. With
200 mg of Lasix + thiazide [or Lasix
drip 10-80 mg per hour], the patient
should make urine within 30 minutes or
so. If not, the patient may need
dialysis.
DIALYSIS FOR AKI IN THE ICU
What are the indications for dialysis?
Clinical uremia, diuretic resistant
volume overload, intoxications,
refractory electrolytes, acidemia.
Questions about dialysis in AKI early
versus late, continuous versus
intermittent and the intensity of dialysis
[i.e. the dose of dialysis].
Timing of dialysis is difficult to know. If
there is a BUN over 150, the patient
should be dialyzed, but this is very old
data. What about subtle differences,
like a BUN more than 80 for a few days?
Early is probably better. The problem is,
what defines early? Some use time to
BUN or creatinine elevation, some use
time of admission. Based on timing of
admission to ICU, late dialysis based on
number of days did worse [late = more
than 5 days]. The answer is still
somewhat unclear.
What about modality? Continuous
[CRRT] versus intermittent [iHD]. iHD
has good-to-excellent, solute control
depending on how often it is done. iHD
provides more rapid solute clearance
and there is hemodynamic instability
only if volume is removed. In terms of
CRRT, CVVHDF has excellent solute
control and greater solute clearance
than CVVHF. Remember that
hemofiltration is convective clearance

Notes 2014
of solutes while
hemodialysis is diffusive
clearance of solutes. The
continuous modalities offer
requires special fluids and training plus
the need
for anticoagulation.
SLED is sustained low efficiency dialysis
[or EDD extended daily dialysis], it is a
conventional iHD machine run over 7-8
hours, sometimes every day. SLED/EDD
versus CVVH is quite comparable, but
need for AC is the same. Acute PD is not
used in the US. It is less precise than
CVVHDF and typically used only for
patients who are on PD as an
outpatient.
There is still much debate about
modality, but what is known is that
CRRT is superior to peritoneal dialysis
in acute, infection-associated renal
failure [mortality rate of 47 versus
15%] a study in Vietnam in 2002
NEJM. There is little, good-quality
evidence supporting the use of iHD
over SLED, EDD or CRRT. There are
conflicting meta-analyses as to the
benefit or not of CRRT.
Is there a mortality effect? There were
3 meta- analysis and the RR for CRRT
modalities versus iHD shows no
consistent effect on mortality,
recovery of renal function or
development of chronic kidney
disease. There was much less HD
instability in patients on continuous
modalities. The MAP was 5 mmHg
higher in the continuous group. In
patients receiving continuous
modalities there is a much better fluid
removal over a period of days. Fluid
overload is associated with worsened
outcome.
Good scenarios for CRRT liver failure
[cerebral edema effect], hypotension,
severe volume overload, and excessive
h
71
precise volume management, and
better hemodynamic stability, but he
disadvantages are that CRRT demands
a different dialysis machine that
volume intake that is obligatory [i.e. from IV
infusions].
Dialysis dose and AKI outcomes. Remember the
dose is how much dialysis is more, better? The
truth is unclear. The two best papers Palevksy and
Belloma in NEJM 2008 and 2009 did not show a
benefit to more dialysis in the RENAL and ATN
studies. One trial [Schiffl NEJM 2002] showed an
improved survival in daily dialysis compared to
every other day, though this trial is criticized
because the dosing was less than optimal.

Notes 2014
Another trial looked at
dosing in CRRT and found
an improved survival when
comparing an effluent of 20
cc/kg/hr to 35 cc/kg/hour,
BUT NOT improvement
when comparing 35 to 45
cc/kg/hour [i.e. optimal
effluent of 35 cc/kg/h]
So the summary is: IHD or
CRRT as per preference,
intensity iHD 3 times per
week as long as kt/V [i.e.
the dose] is more than 1.2
per session.
CONTRAST NEPHROPATHY
The prevention of contrast
nephropathy [RCIN] is far
from clear, but evidence
suggests the use of normal
saline, potentially IV NAC
and the use of iso-osmolar
radio-contrast may reduce
the incidence. In small
trials, ACEIs have been
shown to be protective!
Risk factors include ESRD,
heart failure, diabetes, and
How to approach hyponatremia? 1. Is it
real 2. Is
water excretion appropriate and 3. Is
ADH
secretion appropriate?
Is this real? Look to the serum
osmolality. If it is normal or high,
consider pseudo-hyponatremia, uremia,
transpositional hyponatremia from
glucose, mannitol, or glycine. Each
increase in serum glucose by 100,
above 200 mg/dL lowers Na by 1.6. If
the serum osms are low, then it is true
hyponatremia.
Is water excretion appropriate? Look at
urine osms or specific gravity [if you
multiply the last two digits of the
NSAID use. There is little study on this

topic, but normal saline is superior to
oral hydration and normal saline is
superior to half-normal saline for the
prevention of RCN.
72
The rate is 1 cc/kg/hour for 6-12 hours

before and after the contrast load. The
addition of Lasix to saline infusion is
detrimental, however, there is no
evidence that stopping chronic Lasix in
a patient will protect the kidneys when
trying to prevent RCIN. Fenoldapam is a
selective, peripheral dopamine agonist
with effects of renal blood flow, but an
RCT showed that it did not prevent
RCIN. Low-osmolar [non-ionic] contrast
agents seem to prevent RCN compared
to the classical ionic ones. Prophylactic
hemodialysis to prevent RCN made
things worse.
ELECTROLYTES IN THE ICU

HYPONATREMIA [EXCESS FREE WATER]
Pathophysiologically, hyponatremia
should really be considered hyperaquemia or an excess of free water.
The diagnosis should focus on why the
patient has too much free water.
specific gravity by 30, this approximates urine
osmolality]. If urine osmolality is less than 100
[spec gravity between
1.003 and 1.004], the patient is appropriately
urinating free water [ADH activity is low]. Reset
osmostat rarely lowers urine osms below the high
120s. If urine osms are high, then ADH release is
high.
Then ask volume status. Appropriate ADH release
occurs when the patient is volume down. If the
patient is volume down, then look to the urine
sodium, if it is low, then there are non-renal volume
loses, if high, consider renal loses such as diuretics,
adrenal insufficiency, cerebral salt wasting should
be entertained which occurs following craniofacial
injuries. If the patient is hypervolemic, then the ADH
is maladaptive and this is fairly obvious from the
clinical examination. Euvolemia is the most difficult

Notes 2014
situation. If urine osms are high, SIADH,
renal failure, hypothyroid, isolated
glucocorticoid deficiency [anecdotally
rare, but one paper suggested up to
30% of SIADH is actually isolated
glucocorticoid deficiency]. If urine osms
are low, consider polydipsia, reset
osmostat even if patient appears
euvolemic consider sending serum
uric acid to help distinguish SIADH
[SIADH results in hyperuricosuria which
lowers serum uric acid level].
73
ADH is stimulated in response to
multiple
disorders and physiological stimuli
pain, nausea/vomiting, malignancies,
pulmonary disease, CNS disorders.
Drugs affect water metabolism, drugs
have multiple mechanisms ADH
analogues like DDAVP, oxytocin;
increased ADH secretion by drugs
[nicotine, anti-psychotics, certain antidepressants, carbamazepine];
increased renal sensitivity to ADH
[NSAIDS]; drugs that cause
hyponatremia by unknown mechanisms
such as haloperidol, SSRIs, MDMA, PPI.
How is excess free water treated? The
etiology, rapidity and volume status
are important. If hypovolemic and no
symptoms, always give NS. If
euvolemic or hypervolemic and no
symptoms, free water restrict. If mild
symptoms, give saline
& furosemide [poor mans hypertonic
saline] or just furosemide [depending
on volume status], and consider V2
antagonists [though V2 antagonists are
rarely the answer in the boards]. If
symptoms are severe [i.e. neurological
in nature], the answer is hypertonic IV.
A 100 cc bolus of 3% will safely
increase the Na by 2-4 mEq/L in most
patients; another rule of thumb is that
in general, a bolus of 1 mL/kg [or by
hourly infusion] of 3% saline will raise
the serum sodium by 1mEq/mL.
Regardless, very close monitoring is
important.
AVP receptors V1a is vascular smooth
muscle cells & myocardium; V2 is the
renal collecting duct. V2 antagonism
induces a brisk loss of free water.
Conivaptan is a mixed antagonist, all
others are pure V2. Conivaptan is only
available IV.
Consider a woman with GERD,
depression, and confusion for 2 days.
Spontaneous UOP is 50 ml/hour. Serum

Notes 2014
sodium is 116. Urine osms
high. On the boards, fluid
restriction is never
appropriate with CNS
symptoms. The answer is
mostly sodium and potassium. A large
portion of
idio-osms are generated by 74 hours.
This is what dictates acute and nonacute. If you dont know how long it has
been, rapid correction can lead to CPM.
The patient should not increase more
than 10-12 in 24 hours or 18 in 48
hours. Rapid achievement over 2-4
hours of 2-4 mEq/L is safe and
advisable. Caution is advised with
patients who may develop a brisk water
diuresis [i.e. superimposed ADH
secretion from hypovolemia].
CPM can be extra-pontine, so it is not
called CPM anymore. It is now called
osmotic demyelination syndrome
[ODM]. It rarely occurs in patients
whose serum is more than 120 mmol/L.
It typically develops after an initial
improvement in mentation - potentially
by weeks. There can be focal motor
deficits, quadriparesis, paralysis,
respiratory depression, pseudobulbar
palsy, coma, up to 33% extra-pontine.
MRI with hyper-intense lesions on T2,
non-enhancing.
Those with: ethanol abuse, malnutrition,
hypokalemia, elderly women on
thiazides, burns, liver transplant are at
excess risk for ODM.
Consider a patient with cirrhosis who
loses lots of ascites through and
abdominal incision and presents over
two weeks with worsening hypotension
and new hyponatremia to 103. He
received hypertonic saline and corrects
12 mEq over the course of 4 hours.
Alcoholics and cirrhotics are particularly
susceptible to ODM because of their
inherently low intra-cellular osmolality.
Whats interesting is that despite his
3% saline IV. The brain defends itself

from excess water by increasing intracellular sodium, potassium and idioosms. The acute response is
74
diuresis and rapid correction of serum sodium

[presumably due to release of ADH and free water
diuresis], his urine osmolality is shockingly high
[above 500] suggesting that there is SIADH.
However, the urine osms are high because the
kidneys start to dump urea that accumulated in the
pre-renal state, so even though the osmolality of the
urine is high, it is essentially free water. He still
needs free water replacement to prevent his serum
sodium from jumping too quickly.

Notes 2014
Consider a schizophrenic
with a tonic-clonic
seizure, Na 116, urine osms
92, plasma osms 240. He is
making about 120 cc per
hour.
Schizophrenia should raise
concern for SIADH, and
polydipsia. His urine
sodium of 35 suggests that
he is not volume down, hes
making good urine and he
is diluting the urine quite
well; this patient has
polydipsia. With reset
osmostat, this degree of
low Na is rarely achieved.
HYPERNATREMIA [FREE WATER
DEFICIENCY]
Pathophysiologically,
hypernatremia should really
be considered hypoaquemia or a deficiency of
free water. The diagnosis
should focus on why the
patient has too little free
water.
Hypernatremia is very
common in the ICU and
almost always the MDs
fault. It has a negative
and 2. The hypernatremia is due to
free water
deficiency [treat the calculated free
water deficit with enteral free water
flushes or D5W IV].
Special circumstances of
hypernatremia include DI which
requires DDAVP replacement, but this
uncommon outside of the
neurosurgical population.
Tonicity balance? This is a
physiological/rational approach to
determining the treatment fluid for
dysnatremic patients. The tonicity is
75
prognostic value. It is the combination
of a loss of drive for free water, loss of
access to free water and hypotonic fluid
loses [e.g. fever, NG suction, increased
MVe, mechanical ventilation, urinary
concentration defects such as DI,
diuretics, solute diuresis from glucose,
mannitol, urea, TPN].
How to manage hypo-aquemia? Give
back free water! The patient needs
replacement for what is lost and also for
ongoing loses. If the patient is
hypovolemic, give saline back, also give
free water. If the patient is
hypervolemic, give free water. It is a
fallacy to withhold or limit free water
in a patient who is volume overloaded.
In one liter of free water, less than 150
mL stays in the intra-vascular space.
This will not contribute to vascular
congestion, but it will replenish the
intra-cellular free water deficiency.
Further, free water replacement will
help mitigate thirst, which is one of the
most unpleasant experiences recalled
by ICU survivors.
In patients who are hypervolemic and
hypernatremic [common in the ICU],
there are two separate problems to be
addressed. 1. the hypervolemia is due
to excess total body sodium content
[treat with sodium restriction and Lasix]
the amount of sodium and potassium in the infusion
fluid as well as the sodium and the potassium
coming out of patients. If the input tonicity is more
than the output tonicity, then serum sodium will
increase. If the input tonicity is less than the output
tonicity then the serum sodium will drop. There are
fancy formulae for this approach.
HYPOKALEMIA
Low serum potassium is a combination of intake
problems, excretion problems and trans-cellular
shifts [which is only an acute issue].
B2-adrenergic agents shift potassium into cells,
re-feeding can shift potassium into cells and

Notes 2014
alkalosis can as well, but the latter is
a mild effect.
You can use the TTKG to narrow the
differential for hypokalemia. Note that
the urine is the numerator of the TTKG;
so the TTKG is a measure of renal
potassium excretion. The lower the
TTKG, the more potassium avid are the
kidneys. If the patient is hypokalemic
and the TTKG is more than 2, then the
beans are wasting too much
potassium.
If the patient is hypertensive and there
is renal K+ wasting based on the TTKG,
then the patient should be checked for
aldosterone and renin activity. If both
the renin and aldosterone activities are
high [a ratio less than 10] then the
patient is: on a diuretic, has CHF, renovascular HTN, malignant HTN, reninsecreting tumors,
76
these entities are known as secondary hyperaldosteronism. Note that these workups are often initiated in search for
renin-secreting tumors, or renal artery
stenosis etc. but that being on a
diuretic can mimic this physiology. If
there is high aldosterone activity with
a suppressed renin activity, this is
primary aldosteronism. If both low
aldosterone and low renin occurs, then
consider the patient is on exogenous
mineralocorticoids, has Cushings
syndrome, has lots of licorice intake, or
other weird things like Liddle and
Geller syndrome.
Other hypoK+ pearls: always treat
magnesium too [JASN Oct. 2007 v.18;
p.2649]. Be aware of hypokalemia as a
part of re-feeding syndrome - low
phosphate may occur too. Be very
wary of the kaleriuretic effect of
bicarbonate. Sometimes patients have
a metabolic alkalosis and the clinician
plans of using acetazolamide. If the
patient is also hypokalemic, then there
can be a pronounced reduction of
potassium. If you dump bicarbonate
through the kidneys, you will dump
potassium like crazy. This can also be a
problem if giving bicarbonate for
contrast nephropathy or in patients
with DKA [typically discouraged in
both].
HYPERKALEMIA
Like, low serum potassium, a high
concentration of serum potassium is a
combination of intake problems,
excretion problems and trans-cellular
shifts [which is only an acute issue].
Shifting potassium out of cells is known
to occur in critically ill patients low
insulin states, solvent drag in
hyperosmolar states and mineral
acidosis [NOT organic acidoses e.g.
lactic and ketoacidosis]. Further,
cellular destruction causes

Notes 2014
hyperkalemia, not
increased cellular
turnover. Also, during the
clotting process, cells leak
their contents such that
This becomes a problem only when
there is a wild
elevation in WBC or in platelet count
[serum potassium may be 1-2 mEq
higher from cellular leak]. In this
situation one can try rapidly separating
plasma from cells and measuring the
electrolytes.
In terms of problems with potassium
excretion, if the patient is hyperkalemic
and the TTKG is less than 6, then the
beans are holding onto K+ too avidly;
this is almost always a drug-effect.
ACE, ARB, aldactone, trimethoprim
which has an amiloride-like effect, BB
can do it, heparin IV and SQ and LMWH
are also common, but forgotten
culprits, others are succinylcholine,
ketoconazole and digitalis toxicity.
As above, trimethoprim acts like
amiloride a potassium sparing
diuretic. Amiloride blocks the apical
distal nephron sodium channel which
decreases the trans-membrane voltage
and favors potassium retention within
the cell.
Studies have shown that the
administration of high dose
trimethorprim raises serum potassium
by 0.5 to 1 mEq in general with case
reports of significant elevations. The
latter may occur in patients with
hepatic disease as the liver clears
trimethoprim
How is high K treated? First, to treat
severe, high potassium is to control
cardio-toxicity [calcium chloride IV],
then shift potassium into cells [betaagonists, insulin] and then remove the
potassium [diuresis, kayexelate, and
hemodialysis]. These three things
occur in concert. The kayexelate story
77
whenever electrolytes are measured
from serum [as they usually are], the
potassium levels are slightly higher
than when measured from plasma [0.2
to 0.4 mEq higher].
is interesting because of the FDA warning
regarding kayexelate crystal formation in the gut
causing necrosis; in one series this complication
was noted in 2 of 117 [roughly 2%] patients given
kayexelate. The sorbitol may be the culprit.
Consider only using kayexelate for a potassium
above 6.
HYPERCALCEMIA

Notes 2014
What causes hypercalcemia
in the ICU? Usually
malignancy [multiple
mechanisms including PTHrp secretion, bony
involvement, excess vitamin
D synthesis]. Consider also
thiazides, rhabdomyolysis
recovery and
immobilization.
When do you suspect MilkAlkali syndrome? This can
be seen with very high
calcium levels in fitness &
vitamin enthusiasts or those
taking
fist-fulls of Tums for various
reasons. The most common
cause in the ICU is,
however, malignancy.
Patients with mild to
moderate hypercalcemia
often have primary
hyperparathyroidism, but
this is not typically an
indication to be in the ICU.
Moderate to severe is
frequently malignant but
may also be milk alkali
syndrome. If the calcium is
high, check a PTH to make
sure that PTH is
appropriately suppressed.
immunogenic work up is entirely
normal. The
skin lesions are the calciphylaxis and
are thought to be the result of calcium
deposition into superficial arteries that
then necrose. These lesions are
typically not biopsied as they rarely
heal. Systemic calcinosis is the
calcification of the organs. The lungs
may certainly be involved. The
alveolar-septal walls are often filled
with granular and linear calcific
deposits these findings may also be
seen in pulmonary vessels and airways.
78
Treatment: There may be normal or
high phosphate, metabolic alkalosis,
and certainly AKI. PTH and 1,25 D levels
are usually suppressed, management is
supportive saline and removal of the
source. Generally, treatment involves
sodium-containing hydration and this
promotes calciuresis. Loop diuretic
should be used sparingly if at all but
only when volume replete. Steroids can
be used if there is a granulomatous
process driving the calcium from excess
vitamin D production. Calcitonin has a
rapid tachyphylaxis, but it is good
because it acts quickly, especially when
hypercalcemia is quite severe.
Bisphosponates IV are important,
zoledronic acid is typically used.
Hemodialysis is last resort.
Recognize and treat calciphylaxis and
systemic calcinosis. There may be a
scenario set up like a pulmonary renal
syndrome. However, the patient has
chronic renal failure [on peritoneal
dialysis] and presents with bilateral
infiltrates, hypoxemia and lower
extremity raised, violaceous lesions.
Notably the stem does not state that
the lesions are biopsied and the
parathyroid hormone level is given as
normal, despite a fairly high serum
calcium. Further, collagen vascular &
The bone scan is the best diagnostic procedure
here, though it does have poor sensitivity and false
positives can occur in lymphoma. The best
treatment is prevention. In ESRD, it is important to
keep the calcium phosphate product less than
70. Corticosteroids and immunosuppressants
should be avoided. Mortality is high, usually
secondary to sepsis.
HYPOCALCEMIA
It is quite common in the ICU with many causes,
probably sepsis is the most common cause
followed by provision of blood products. There is a
physiological hypoparathyroidism of sepsis.

Notes 2014
There is little data to suggest that
replacement is beneficial and some
data [in animals] that it makes things
worse. Calcium should be replaced if
there is instability, seizures, tetany, or
essentially any symptoms. Calcium
gluconate or chloride can be used [the
latter having 3 x more elemental
calcium usually requiring a central
line]. Post para-thyroidectomy may
result in hungry bone syndrome
especially in patients with end-stage
renal disease and prolonged, refractory
secondary hyperparathyroidism. A
calcium drip may be needed in this
situation.
Like hypokalemia, with hypocalcemia,
you cannot correct it until
hypomagnesemia is also corrected.
HYPOMAGNESEMIA
Hypomagnesemia is typically caused
by diuretics and ethanol abuse,
nutritional deficiency, but
79
also aminoglycosides, ampho B, but also PPIs that
is a mixed mechanism [it is probably
from dumping mag into the bone]!
You can do a FEMg in those without
renal loss it will be less than 4%, if more
then it may be renal dumping.
Replacing with IV or PO Mg. Amiloride,
triamterene and spironolactone will also
raise magnesium levels.
HYPOPHOSPHATEMIA
Hypophosphatemia can be evaluated
by a FEPO4, with the cutoff being 5%.
The kidneys will hold on to phosphate
when there is trans-cellular shift of
phosphate [e.g. insulin secretion,
respiratory alkalosis, re-feeding, hungry
bones, sepsis], but also decreased
intestinal absorption [e.g. phosphate
binders, calcitriol deficiency].
FEPO4 will be high [more than 5%] in
hyperparathyroidism, Fanconi
syndrome, and osmotic diuretics or
solute diuresis the latter being very
common in the ICU.
Causes of severe hypophosphatemia
seen in the ICU are ethanol
withdrawal chronic depletion, and refeeding syndrome with D5; DKA which
is from losses and pH shifts; TPN with
re-feeding; acute respiratory alkalosis,
correction of chronic respiratory
acidosis, and the diuretic phase of
severe burns. Perhaps the most
common cause is respiratory alkalosis.
Usually, if the PaCO2 goes back up, the
hypophosphatemia will also reverse.
ACID-BASE IN THE ICU

METABOLIC ACIDOSIS
There are 5 basic mechanisms of

metabolic acidosis renal acid
excretion problems such as renal
failure or distal RTA, renal loss of
bicarbonate such as a proximal RTA, an

Notes 2014
extra- renal loss of
bicarbonate that is in
excess of the kidneys
ability to regenerate bicarb
of metabolic acid in excess of the renal
ability to
regenerate bicarb [e.g. lactate, dlactate, ketoacids, HCl, cationic and
sulfated aminoacids, toxic ingestions],
and lastly dilution acidosis which is the
result of excess chloride [see Stewart
approach below].
ANION GAP ACIDOSES
What are the causes of a high anion
gap added anions, but also consider
hypoK+, hypoMg++, hyperPO3,
hyperSO4 [mediators of AG in renal
failure which can get up to 20 or so],
hyper- albuminemia. If AG is more than
25- 30, it is almost always an organic
acidosis. The clinically important anion
gap acidoses can be remembered by:
KULT [ketones, uremia, lactate and
toxic alcohols].
What about the causes of a low AG?
This is important in hypo-albuminemia
because the patients expected AG will
drop by about 2.5 for each decrease in
albumin by 1 g/dL. Other causes of a
low AG are cationic paraproteinemia,
hypoPO4, hyper: Ca, K, Mg, Li
intoxication, and severe hypernatremia.
TOXIC ALCOHOLS
Remember that the osmolal gap is the
difference between measured and
calculated serum osms. If it is more
than 10, theres an unmeasured osm
floating around. If this is in conjunction
with a metabolic acidosis, then the
culprits are ethylene glycol, methanol,
ethanol, propylene glycol,
formaldehyde and paraldehyde. If
there is not a
metabolic acidosis, but only an osmolal
gap, think about: mannitol, glycine,
sorbitol, maltone [IgG infusion],
[e.g. diarrhea, pancreato-biliary

fistulae, pancreas transplants, ileal
conduits], increased generation
80
isopropyl alcohol [rubbing alcohol will also give

you ketones as acetone] and pseudohyponatremia.
With toxic alcohol ingestion, the cause of the
osmolal gap is different from the cause of the anion
gap. With ethylene glycol, the osmolal gap is
ethylene glycol, and the AG is glycolic, oxalic and
hippuric acids. With methanol, the OG is

Notes 2014
from methanol, but the AG is from
formic and
lactic acids. With
propylene glycol, the OG is
from propylene glycol, but
the AG is from L and D
lactic acids.
Where does ethylene
glycol come from? Antifreeze and other solvents.
There is typically CNS
depression early with HTN
and this is followed by
acidemia, CV collapse and
renal failure from oxalate
crystals in the urine. The
diagnosis is made by a
high AG with high OG.
Leukocytosis is common.
The treatment of ethylene
glycol, prevent further
metabolism with
fomepizole and increase
the conversion of
glyoxalate to glycine with
pyridoxine and thiamine.
Then you want to remove
toxic metabolites with iHD
or PD this is
recommended only when
there are neurological
changes, renal failure or
severe acidemia.
titrated to maintain a blood level of 100
to 200
mg/dL) or 4-methylpyrazole
(fomepizole, 15 mg/kg loading dose,
followed by 10 mg/kg q12h). Although
ethanol has been the standard
treatment for ethylene glycol
ingestions, ethanols kinetics are
unpredictable, patients must be
rendered intoxicated, and it may
predispose to hepatotoxicity and
hypoglycemia.
4-methylpyrazole is a more specific
inhibitor that is not associated with
h
81
Charcoal hemoperfusion is not
indicated for the treatment of ethylene
glycol toxicity.
Consider a patient who presents
following a suicide attempt with
obtundation, renal failure, rhomboid
crystals in the urine and a combined
osmolal and anion gap. The
combination of a combined osmolal and
anion gap is [in clinical reality] limited
to propylene glycol, ethanol, ethylene
glycol and methanol. The presence of
oxalate crystals in the urine clinches it
for ethylene glycol toxicity.
Hemoperfusion is only indicated when
the molecule of interest cannot be
dialyzed. This is common in very large
molecules, those that are highly protein
bound or those with a high volume of
distribution [tissue- bound]. There are
complications with charcoal
hemoperfusion so this is not ideal. As in
all toxic ingestions, GI decontamination
with activated charcoal should be
instituted immediately once airway
protection is established [within one
hour of ingestion]. The major toxicity in
ethylene glycol ingestions is not due to
the parent compound but is due to its
metabolites glycoaldehyde and glycolic
acid. This metabolism may be inhibited
by the administration of either IV
ethanol (loading dose, 0.6 g/kg,
followed by a continuous infusion
these toxicities, but it is substantially more
expensive. Although it is only approved for the
treatment of ethylene glycol intoxication, 4methylpyrazole is likely to be beneficial in the
treatment of methanol intoxication as well.
Hemodialysis is indicated when the ethylene glycol
concentration is > 50 mg/dL. In patients being
concomitantly treated with IV ethanol, the rate of
ethanol infusion must be increased to compensate
for enhanced clearance by dialysis.
Other high AG/OG acidosis is methanol and
propylene glycol. The latter is from IV lorazepam
drips. 10 mg/hour for more than one day results in a

Notes 2014
lactate acidosis and renal failure. They
often need dialysis.
Consider a patient with ethanol abuse,
pH 7.2//24//100, AG 31, serum ketones
1:4, lactate
3.5. UA with granular casts and calcium
oxalate crystals. The patient has a high
AG and high osmolal gap as well. When
both gaps are present it means either
that the toxic alcohol ingestion is
caught early, or there is co-ingestion
with ethanol which slows the
metabolism of the toxic alcohol.
Methanol or ethylene glycol are the
common causes from the community.
NON-ANION GAP ACIDOSES
The urine AG measures
ammoniagenesis. The normal urinary
anion gap is negative because there
are unmeasured urinary cations [mostly
ammonium] getting rid of daily acid
intake. In other words, the urinary
chloride normally exceeds the sum of
the urinary sodium and
82
potassium. This normal concentration of
ammonium is in the 30 mEq range
[from our acidic protein intake]. When a
metabolic acidosis from bicarbonate
loss occurs, the distal nephrons work to
excrete more acid in the form of
ammonium such that the urinary gap
becomes more negative [50 mEq or
more negative].
If there is systemic acidosis with a
negative serum anion gap,
hyperchloremia and the urinary gap is
positive, then the distal nephrons are
behaving badly and there is either a
type 1 or 4 RTA in the works. The
difference between the latter is usually
potassium level as type 4 RTA occurs in
the setting of aldosterone resistance
such as diabetes [where renin and
angiotensin levels are low], ACEI,
NSAIDS, heparin, and various adrenal
abnormalities so there is commonly
hyperkalemia. Type 1 RTA [distal], by
contrast, occurs in the setting of lupus,
active hepatitis, Sjogrens syndrome
and hypergammaglobulinemia.
Type 1 [distal] and type 2 [proximal]
RTA both classically have hypokalemia.
Urinary pH is generally unhelpful when
making the distinction between the
aforementioned because it is regulated
by multiple other factors, though an
alkaline pH in the setting of systemic
acidosis, in general, supports the
diagnosis of an RTA.
Type II RTA occurs from proximal
nephron bicarbonate wasting, if you
give bicarbonate, they dump
bicarbonate, but they can acidify their
urine. Type II RTA, therefore, also has a
negative urinary anion gap [like gut
bicarbonate loss], because the distal
nephrons can still generate ammonium.
As an aside, type III RTA was once used
to refer to the condition of a combined
Type I and Type II RTA [an inability to

Notes 2014
generate ammonium in the
distal nephron and an
inability to resorb proximal
83
bicarbonate, respectively]. Thats when
you consult nephrology [JASN: Aug.
2002 vol. 13; p2160].

Notes 2014
ICU RTA causes [type I autoimmune,
ampho B,
transplant rejection, sickle cell disease,
volume depletion], [type II or proximal,
renal bicarbonate wasting multiple
myeloma, acetazolamide use, Sjogrens,
transplant rejection] and [type IV or
aldosterone deficiency or resistance for
example from ACEi, NSAIDs, heparin,
aldactone].
Why should acidosis be treated? HD
stability, myocardial performance,
arrhythmias, and chronically muscle
catabolism. The treatment of metabolic
acidosis is to treat the cause. Sodium
bicarbonate can be used for patients
with diarrhea, renal failure, etc. If the
patient is on TPN, acetate and lactate
and citrate should be titrated. Dialysis
may be needed.
METABOLIC ALKALOSIS
Metabolic alkalosis occurs from the loss
of hydrogen for example from NG
suction, antacids, renal proton loss from
diuretics, mineralocorticoids,
hypercalcemia, post- hypercapnia
compensation, low potassium or refeeding. There may be retention of
bicarbonate from administration of
sodium bicarbonate or milk-alkali. There
may be metabolism of organic anions
from massive transfusion or recovery
from organic acidosis. Or there may be
contraction alkalosis from chloride loss.
In the proximal tubule bicarbonate is
reabsorbed [facilitated by hypokalemia,
hypercapnia and volume contraction], in
the distal nephron bicarbonate is
regenerated [facilitated by hypokalemia,
hypochloridemia, hyperaldosteronism]
these factors maintain the metabolic
alkalosis.
Metabolic alkalosis should be treated
because of hypoventilation, reduced
oxygen release from Hb, hypokalemia,
6
h
hypocalcemia, muscular and cardiac

complications. Correct the generating and
maintaining factors. In the hypovolemic patient
sodium and potassium should be infused. It
corrects volume contraction, it lowers renin

Notes 2014
and aldo activity, corrects Cl depletion,
corrects K
depletion, and increases
GFR. The use of acid is
rarely needed [0.1N HCl],
one also could consider
giving acetazolamide,
spironolactone, amiloride.
Lastly, dialysis or CVVH.
Consider a patient with
cryptic metabolic alkalosis.
The history and physical is
paramount. In patients
without a history of
hypertension, it is almost
always surreptitious use of
diuretics or inducedvomiting. However for
board exams the patient is
usually hypertensive with
stigmata of uncontrolled
hypertension such as AV
nicking on retinal exam,
suggesting
hyperaldosteronism.
The urinary chloride can
help differentiate
hyperaldosteronism from
diuretic use. Diuretics
initially increase urinary
chloride, but then with
hypovolemia the urinary
chloride level falls [many
hours after last dose].

Hyperaldosteronism [e.g. Conns
Syndrome] presents with urinary
chloride that is high [above 25]. If the
patients urinary chloride is high and
has not received a diuretic recently, the
patient likely has hyperaldosteronism.
Consider a patient aggressively

diuresed for heart failure with Na 142, K
3.7, Cl 86, CO2 40.
7.57//41//78. He is still with edema.
What to do, spironolactone or
acetazolamide? The preference is
spironolactone, not use acetazolamide
because the patient is already
hypokalemic and this will worsen with
acetazolamide administration.
Aldactone has a better synergistic
effect with Lasix and aldactone also
results in acid retention. The patient
has a mixed metabolic and respiratory
alkalosis which is the most common
acid-base problem of treated heart
failure.
MIXED DISORDERS
What about the gap/gap or delta gap?
In a simple disorder, the increase in AG
should match the drop in the measured
venous bicarbonate mEq for mEq.
Making this comparison is the gap- gap.
If the AG increased more than the drop
in

Notes 2014
bicarb, then there was a pre-existing
metabolic
alkalosis i.e. the measured venous
bicarbonate is too high as compared to
the elevated anion gap.
Remember, when determining the
change in AG, you must correct for
albumin. So if the AG is 25 with a
normal albumin, it means that there are
13 excessive anions that must be
accounted for assuming a normal AG
of 12. The 13 cryptic anions must be
matched by a drop in venous
bicarbonate. Thus, assuming a normal
bicarbonate of 25, the measured
venous bicarbonate should be 12-14. If
the measured value of bicarbonate is
much higher, there is a pre-existing
metabolic alkalosis, if the measured
bicarbonate is much lower there is a
pre-existing non-gap acidosis.
A common board scenario is the triple
acid-base question. Consider a young
patient with IDDM, with pneumonia
[respiratory acidosis], DKA [AG
metabolic acidosis], and vomiting for
one week [metabolic alkalosis]. How is
this approached? Find the primary
disorder there is acidosis with a low
bicarbonate and a PaCO2 below 40
mmHg, but Winters Formula [1.5
bicarbonate + 8 (+/-2)], shows that the
PaCO2 is too high. Thus there is a
metabolic [AG] and respiratory acidosis.
But applying the gap-gap will reveal
that the measured bicarbonate is too
high for the increase in anion gap, thus
there is a metabolic alkalosis [from
puking].
BRIEF
TO PH
OVERVIEW OF THE STEWART APPROACH
What is the mechanism of post-fluid

resuscitation metabolic acidosis? It is a
narrowing of the strong ion difference
based on the Stewart model.
h
86
Note that if you add normal saline [154 Na, 154 Cl]
to the plasma [average Na & Cl around 140 and
105, respectively], the excess chloride in the NS
relative to the serum chloride will increase the
serum chloride out of proportion to the increase

Notes 2014
in serum sodium. In other
words, the difference
[i.e. the strong ion
difference] between sodium
and chloride will narrow.
This will favor more protons
dissociated into the plasma,
and less bicarbonate anions
[to maintain electroneutrality] and therefore a
more acid pH. This is often
explained as the NS
diluting away the
bicarbonate but, per the
Stewart approach, the
bicarbonate is disappearing
per its dissociation constant
in order to maintain
electroneutrality [as the
negatively charged chloride
increases].
COMMON BOARD ACID-BASE
SCENARIOS
Board exam scenarios
additive disorders
metabolic acidosis plus
respiratory acidosis which
presents with severe
acidemia, low bicarbonate
[from cryptic lactate] and
increased PaCO2 e.g.
cardiac arrest.
87
There may be metabolic plus
respiratory alkalosis
severe alkalemia with high
bicarbonate and low PaCO2 e.g.
vomiting with hepatic failure, or treated
acute heart failure.
There may be counterbalancing
disorders where the pH is near normal.
Consider respiratory acidosis with
metabolic alkalosis with variable acidity
high bicarbonate, high PaCO2 for
example acute on chronic lung disease
and CHF treated with diuretics.
Respiratory alkalosis and metabolic
acidosis variable acidity with low
bicarbonate and low PaCO2 for
example chronic renal failure with GNR
sepsis.
There could be an anion gap metabolic
acidosis and metabolic alkalosis with
variable acidity, variable bicarbonate,
variable PaCO2. The increase in the AG
will exceed the decrement in
bicarbonate. There is classically shock
with lactic acidosis [or DKA] and
superimposed vomiting.
Then there are triple disorders such as
respiratory acidosis, metabolic acidosis
and metabolic alkalosis chronic lung
disease with DKA and
h
vomiting [as above patient] and respiratory
alkalosis, metabolic acidosis and
metabolic alkalosis for example ASA
overdose in a patient receiving
diuretics.
There may be a combined gap, nongap metabolic acidosis in a burn
patient. The patient has an albumin of
2, an anion gap of 15 and a measured
venous bicarbonate of 8. His predicted
anion gap is only 5 because of his low
albumin which means that his
measured anion gap of 15 leaves 10
anions unaccounted for. If his gap-gap
is 10, his measured venous bicarbonate
should be 15, but it is 8 which means
that there is a bicarbonate wasting
process as well. He is hyperchloremic
with lactic acidosis.
Recognize inconsistent an
uninterpretable data in acid-base.
Consider a patient with a venous
bicarbonate of 13 and an AG of 27.
Further, the PaCO2 is 40 this would be
both a metabolic and respiratory
acidosis and the pH should be rather
low. However the pH listed is normal. In
the presence of an anion gap acidosis
with a bicarbonate that low, the pH
could only be normalized if the PaCO2
is lowered substantially. Even if this
patient were a chronic retainer, at a pH
of 7.40 and a PaCO2 of 40, the venous
bicarb should be in the mid-20s
[because the chronic retainer would
chronically raise their bicarb such that
an anion gap acidosis of 27 would bring
the measured bicarbonate down to the
mid-20s].
Therefore, this data is wrong.

Notes 2014
63
SEVERE ACUTE PANCREATITIS

ENZYMES
The goals of this section are to
recognize common presentations and
etiologies or pancreatitis, describe the
staging roles of CT scans, when to use
IAP monitoring, when to use early
enteral feeding, when to use
antibiotics, and define when
intervention is important.
The common causes of acute
pancreatitis are alcohol, gallstones,
trauma and hyperlipidemia. Only about
10-20% have life-threatening course.
The death is from complications and
typically occur less than 7 days or
more than two weeks [two peaks of
critical illness]. The real morbidity and
mortality occurs in a small group of the
severe pancreatitis. World-wide the
number one cause of severe
pancreatitis is gallstones, in the
U.S. it is alcohol.
There is no real definition of severe
pancreatitis, but be wary of: APACHE II
more than 8, three or more Ransons
criteria, and acute pancreatitis plus
local complications.
Acute pancreatitis is a lot like sepsis
with a lot of IL1, IL6 and TNF. Symptoms
occur and patients present to the ED
prior to the peak of cytokine production.
There is a window of time between
cytokine production and end-organ
dysfunction. Presentation to the ED
typically occurs in the first 12 hours,
maximal cytokine production typically
occurs around 24 hours, and end-organ
badness flares around 60 hours.
Therefore there is not more than 60
hours for one to prevent end-organ
dysfunction with intervention.
Renal failure can falsely elevate amylase.

The initial amylase is typically 3x normal [80
is usually normal] but can be normal in up
to one third. The
4. CRITICAL CARE
GASTROENTEROLOGY
Notes 2014
pancreas may be burned out, the
patient may be late in presentation.
Both amylase and lipase can be
measured, there is no reason to
measure both after a procedure.
Trigs, glucose, LDH, calcium, liver
enzymes should also be measured.
Procalcitonin may predict comorbid infections, but not certain.
The severity assessment is via the
Ransons Criteria. At admission
Age, WBC, blood glucose, AST,
serum LDH and again at 48 hours
[48h criteria are essentially about
fluid requirements].
There is a simple bedside score
five easily obtained measures such
as serum urea more than 9, age
more than 60, mental status
disturbed, SIRS, pleural effusion. If
more than 3 then consider severe
acute pancreatitis.
IMAGING
64
What does the CT scan tell you or what
are the classical indications for
scanning? The following should prompt
a CT scan: if the diagnosis is in doubt,
severe clinical pancreatitis, high fever
with leukocytosis, Ranson more than 3,
APACHE more than 8, lack of
improvement at 72 hours, or acute
deterioration after improvement.
The single best way to prognosticate
severe acute pancreatitis? CT scan with
IV and oral contrast with more than
50% necrosis. This comes from
Balthazar. A is a normal pancreas, B is
focal or diffuse enlargement of the
pancreas, C is B but with peripancreatic inflammation, D is single
fluid collection, E is two or more fluid
collections and/or gas.
What are the data between Ranson and
Balthazar in terms of infection and
mortality? Balthazar is probably better
in terms of prognosis for both
h
mortality prognosis and infection. The
data is from the early 1990s.
Organ failure in necrotizing pancreatitis
- the most common organ to fail in
response to pancreatitis is the
respiratory system [35%], followed by
CV in the 20% range, GI just below
that, hepatic and renal about 15%.
Balthazar did refine the scoring from
letter grade to points [A0, B1, C2, D3,
E4, with percent necrosis as the
addition] from 2002 and this is the CT
severity index. 7-10 on the CT severity
index is a bad prognosticator.
What about an ultrasound? Gallstones
should be suspected in all patients and
therefore all patients need an US and
biochemical tests to assess the
common bile duct; endoscopic US may
also be helpful.
MANAGEMENT
Management issues require
identification and severity assessment,
patients must be assessed for risk of
rapid deterioration [e.g. elderly, obese,
ongoing volume requirements,
substantial necrosis]. Consider intraabdominal pressure [IAP] elevation,
and pain control such as systemic or
epidural analgesia [there was no
difference between the two groups in
one study].
The lungs are typically the first system
to fail, there is diminished FRC. These
patients require lots of fluids,
sometimes up to 500 mL per hour.
There is a 23% incidence of acute renal
failure as well.
What about IAH and compartment
syndrome [ACS] in pancreatitis.
Elevated intra-abdominal pressure [IAP]
is roughly 12 mmHg, but ACS usually
defined at 20-25 mmHg with organ
dysfunction. There is still a large
proportion of patients requiring
abdominal decompression after 10 days. UOP, MAP,

etc. tend to improve after decompression. If
decompression occurs beyond 7 days, there is a low
survival percentage. Intra-abdominal hypertension
[IAH] and ACS are
rare in mild disease; thus,

measure IAP in all patients
with severe acute
pancreatitis [APACHE more
than 7, MODS more than
2], if IAP more than 18
mmHg on first assessment,
then this should be
monitored continuously.
IAH management is to
consider limiting
crystalloids, decompress
stomach, consider fluid
drain, and consider surgical
intervention.
A patient with jaundice and
severe acute pancreatitis
should have an ERCP
within 24 hours of
presentation absolutely.
The key is the presence of
jaundice. Severe acute
pancreatitis without signs
of obstruction then urgent
ERCP is not needed. This is
based on 7 RCTs of ERCP, 3
of which were
methodologically sound.
Overall complications no
difference, mortality no
difference, no effect on predicted

severity.
A new patient with severe, acute
pancreatitis a fever, elevated WBC
should receive which antimicrobial
therapy? A popular clinical practice is to
place such patients on a carbapenem
such as imipenem, but there is no clear
benefit from this practice. The
Isenmann trial and Dellinger trials were
the latest and best [cipro/flagyl and
merrem] and showed no difference in
outcome. The fever, white count and
systemic unrest are a result of severe
pancreatic inflammation and not
infection per se.
The practice of antimicrobial therapy
for severe, acute pancreatitis comes
from Pederzoli in 1993 which gave
imipenem - 70 some patients and the
study was criticized methodologically.
What about probiotics? Lancet 2008
looked at 298 patients. There was no
benefit to probiotics and suggestion of
harm.
Patients with severe acute pancreatitis
[SAP] can get lipids to their hearts
content. They should get early enteral
nutrition. There is suggestion that total
enteral nutrition has a lower mortality.

Notes 2014
It takes a long time to get organized
necrosis.
When should intervention be
considered? If an FNA is positive [i.e.
infected necrosis or pancreatic
abscess], if the diagnosis is uncertain
[rare with CT], if there is persistent
biliary pancreatitis, or a concurrent
surgical problem such as ischemic
enterocolitis.
An FNA is performed in the majority of
patients if an FNA is positive for
infection, intervention was done
[surgical] still those people have a very
long length of stay. There are many
forms of intervention here, classically it
was surgical. It was previously felt that
early intervention with debridement
would be beneficial, but this is totally
wrong. Now, the idea is to never
operate on these patients as mortality
will increase.
There was an RCT in NEJM 2010 which
used a minimally invasive approach.
Complications and death were lower
when the initial approach was
percutaneous drainage. At MGH, the
vast majority of patients had lots of
gram negatives, but usually
polymicrobial. AGA recommendation is
to always perform an FNA prior to
intervention.
Recognize and treat severe,
hypertriglyceridemia- induced
pancreatitis. Consider a young woman
with type II DM with severe pancreatitis
and SIRS who is intubated for declining
mental status. Her triglycerides [trigs]
are above 6000. This type of
pancreatitis is relatively rare as a cause
[less than 4% of causes] but can occur
in patients with trigs above 1000.
Triglycerides this high are usually the
result of some genetic disorder or lipid
metabolism [of Frederickson]. As an
important aside, a triglyceride level of
more than 500 can cause a falsely low
6
h
amylase level [to confound the diagnosis of

pancreatitis], but lipase should be elevated. Recall
that on the endothelial cells of muscle and fat,
lipoprotein lipase degrades triglycerides into FFA
and glycerol. Importantly, heparin and insulin
activate lipoprotein lipase and lower serum
triglyceride levels, so these therapies are an
important part of treating severe
hypertriglyceridemia. The dose of heparin should

Notes 2014
be prophylactic dose
because there is fear of
hemorrhage into the
inflamed pancreas. If
heparin and insulin do not
lower the triglyceride
content quickly,
plasmapharesis can be
tried. It is important to
avoid propofol in these
patients and consider comorbid hypothyroidism.
IN SUMMARY
What are the conclusions?
Identify risk factors by
severe disease CT staging
too. IAP monitoring should
be considered in the most
ill. EBM does NOT support
ABx, there is no role for
ERCP in patients without
jaundice, no role for early
necresectomy. A positive
FNA should have drain. Less
invasive drainage is
favoured.
GI BLEEDING IN THE ICU
GI mucosal and motor

dysfunction is common in
the ICU. Bleeding adversely affects

outcome and increased LOS by 8 days
and mortality by 1-4x or so.
Stress-related mucosal damage may

develop within 24 hours in the majority
of ICU patients [depending on how this
is defined]. Deb Cook looked at this in
1994 and in 2001 there was an
increase in mortality when ICU-related
bleeding occurred. How is the disease
defined?
Endoscopic evidence is highly sensitive.
Endoscopically, this is seen as multiple
subepithelial petechiae, which progress
to superficial erosions. Most occur in
the fundus it looks like skinning your
knee on the concrete. An isolated
lesion in the proximal stomach is
usually a vascular malformation and
not-stress related, they tend to not be
localized vascular lesions.
While clinically evident ICU bleeding
occurs in 5- 25% of ICU patients [i.e.
coffee grounds in NG, melena,
hematocheezia], what should be used is
clinically important GI bleeding. What is
this? HD instability, measured decrease
in Hb, or need for blood transfusion [2
U]. The natural history of this disease
is 1.5%!
really a risk for stress ulcer bleeding in the ICU

from a few observational studies.
PREVENTION
MECHANISM
Gastric acid secretion, mucosal blood
flow diminution, duodenal reflux [bile
and bile salts are very caustic to the
stomach]. All must be somewhat
present.
In critical illness, splanchnic hypoperfusion results in insults to all 4
protective mechanisms of the gastric
mucosa: reduction of bicarbonate,
reduced blood flow, decreased motility
and acid back diffusion. Free radicals
are also important. Prostaglandins
accelerate healing, have direct
cytoprotective effects, and increase
blood flow.
Gastrin, histamine and acetylcholine all
facilitate gastric acid secretion
[remember basic pathophysiology
cartoon] via activation of the protonpotassium pump. The target pH of the
stomach is what? The pH is 1-2
normally! Its essentially a sterile
lumen. Get the pH above 4 to prevent
stress ulcers.
There are only two risk factors in the
Cook study that were significant
mechanical ventilation for more than
48 hours and coagulopathy. Notably,
hypotension was not quite significant.
The multiple regression odds ratio for
bleeding was
15.6 for mechanical ventilation, 4.3 for
coagulopathy, 3.7 for hypotension
[NS], and 2.0 for sepsis [NS] as well as
other NS risks. There were 847
patients who were high risk [MV,
coagulopathy]; 31 of them had
clinically important bleeding [above
drop in BP, need for pRBC] and 2 of the
1405 low-risk patients had clinically
important bleeding [3.7 versus 0.1%].
So, the most important risk factor for
stress related mucosal injury is MV for
more than 48 hours! H. pylori is not
How do we prevent stress ulcer bleeding? What is

the best preventative therapy? The data
blockers, but it is better than placebo

for prevention.
suggests that continuous infusion of H2
blockers
66
is the best prevention! This

is an old, old study, but
convincingly, infusion was
required to get the gastric
pH more than 4. H2-blockers
have many drug-drug
interactions; summarizing
10 RCT, H2 blockers by
infusion, are better than
placebo by about 50%.
What about immediate
release powder with
bicarbonate for oral
suppression versus
cimetidine continuous
infusion there was no
difference in bleeding,
gastric pH or PNA!
In 1997 PPI versus H2
versus sucralfate, 67
patients no difference.
Sucralfate has a mechanical
effect, does not alter pH, it
was not better than H2
What about H2 blockers versus PPI.

The pH was higher with high dose PPI,
but no difference in GI bleeding
[continuous H2 infusion versus high
h
dose PPI]. PPI is a pro-drug that is
inactivated by protons. They work only
on activated pumps, so they are not
instantly acting drugs. The max effect
takes 3-4 days! The enteric coating of
the PPI is critical, the PPI must be in
the enteric capsule; otherwise, they
are inactivated by acid.
PPIs have some 2C19 and 3A4 effects.
There is a genetic polymorphism to the
metabolism of this drug. In Asians,
there are slow metabolizers and have
an extended half-life. There shouldnt
be important clinical effects from this.
In clinical practice, PPI is commonly
supported over H2 blocker [Conrad
2005 CCM] they used a definition of a
positive NG aspirate PPI had a lower
risk of positive NG aspirate, but no
difference in overt bleeding [the Cook
definition] nor was there a difference in
pH.
CCM 2010 meta-analysis showed that
PPI has no important advantage in
either clinically important bleeding or
the development of PNA. H2 blockers
are cheaper, considered the standard.

Notes 2014
CCM 2013 meta-analysis found that in
critically ill
patients, proton pump inhibitors seem
to be more effective than H2 blockers in
preventing clinically important and
overt upper gastrointestinal bleeding.
The robustness of this conclusion was
limited by the trial methodology,
differences between lower and higher
quality trials, sparse data & possible
publication bias. No observed
difference between drugs in the risk of
pneumonia, death, or ICU length of
stay.
Paul Marik looked at enteral feeding
[CCM 2010]. His analysis showed that
there was no benefit to H2 blockers
while using enteral nutrition, nor was
there a difference in PNA overall.
However, PNA risk may increase in
those who receive both enteral feeding
and H2 blockers. It is still felt that
enteral feeding is not a substitute for
stress ulcer prophylaxis until further
study confirms this.
PEPTIC ULCER DISEASE & CONCLUSION
Prevention of stress-related mucosal
injury is not the same as active peptic
ulcer bleeding PPIs are held as the
standard for treatment of GI ulcerrelated bleeding. PPIs enhance healing,
decrease rebleeding and reducing
surgery.
Conclusions stress ulcers do occur,
but they are uncommon and in patients
with known risk. H2 blockers are
efficacious, PPI have not been studied
as well in the realm of prevention. PPI
will increase pH. Enteral nutrition is not
a substitute. Patients with peptic ulcer
bleeding require PPI prior to endoscopy
and this will change the stigmata of
bleeding, rebleeding and need for
surgery is reduced with PPI for ulcer
bleeding; it is unclear if H. pylori is a
71
cause of stress-related mucosal injury when
identified.
GI BLEEDING IN CIRRHOSIS
Consider a cirrhotic with melena, large esophageal

varices and red wale sign. What is the most
important management step?

Notes 2014
70% of cirrhotics have
varices; there is a varix
development rate of 6% per
year of cirrhosis. Not every
patient with varices will
bleed, but 30% will. Variceal
bleeding commonly requires
at least 4 units of pRBCs,
and in a sizeable percentage
of patients the pRBC
requirement can be dozens
of units.
INITIAL APPROACH
The initial approach to the
bleeding cirrhotic should be
to assume its variceal in
nature. About 50% of
variceal bleeds stop
spontaneously. The risk of
bleeding relates to variceal
size [LaPlaces law] and
endoscopic stigmata. The
red wales sign are little red
blebs that put the patient
at high risk for re-bleed.
The overall risk is directly
related to liver
insufficiency. Currently, in
house mortality is 20%;
twas 30-40%, 20-30 years
there is a normal liver. There could be a
gradient
across the liver from intra-hepatic
disease and varices arise regardless of
the cause. Finally, there is a posthepatic cause for example from
thrombosis of all of the hepatic veins at
the same time. Additionally, right heart
failure, but that is typically longstanding heart failure.
Esophageal and gastric varices are
commonly feared, but small and large
bowel varices can happen - they are
clinically much less significant.
72
ago. There is still a 30-50% one year
mortality after the first bleed, if there
is additional renal failure, or other organ
failure, then the risk of in house
mortality goes way, way up.
There is a good body of literature as
varices are fairly common. Antibiotics
have a survival benefit. Endoscopy will
lower bleeding risk and re-bleeding risk,
but interestingly no survival benefit.
Vasoactive drugs reduce transfusions
and bleeding, non-selective betablockers reduce bleed and rebleed.
Sucralfate and rFVIIa have no data, PPI
have no data, but are suggested as
many patients have multiple reasons
for bleeding. 25% of patients will have
non-variceal causes of bleeding. So
strongly consider PPI. Can consider
tranexamic based on the CRASH-2 trial
[in trauma patients, not GIB], but in
upper GIB with modern anti-ulcer
treatment, there was no significant
difference in outcome [Cochrane 2012].
PATHOPHYSIOLOGY
There is too much pressure in the portal
circulation via prehepatic, hepatic, and
posthepatic causes. Prehepatic would
be thrombosis of portal vein; varices
develop, but
What are the important numbers for portal
hypertension? A hepatic vein pressure can be
obtained by wedging the portal vein. If it is greater
than 12 mmHg, this is a problem. A value between
5-12 mmHg is abnormal, but does not lead to
significant varices. The goal is to get the pressure
less than 12 mmHg.
PRESENTATION AND MEDICAL MANAGEMENT
Clinically there can be massive hematemesis,
melena. Large IV access is important. FFP is given
liberally for an INR more than 1.5. Target a
hemoglobin of 8-10 [old data] as going above 10
can worsen bleeding. NEJM early 2013 showed that
restrictive blood transfusions [Hb of 7 g/dL]

Notes 2014
lowered mortality in patients with early
cirrhosis and active GI bleeding.
73
performed in a patient suspected of
having an
upper GIB; NG lavage is deemed
antiquated in modern management of
UGIB [Gastro Endos 2011 vol 74 page
981]. Please stop performing this
procedure on patients suspected of
having an UGIB; there is truly no clinical
utility.
Erythromycin can be helpful to clear
clots give 125 mg IV or 10 mg IV
reglan [though the data for this practice
is poor]. These patients should be in an
ICU certainly for monitoring. 20% are
infected at admission, 70% will develop
an infection in house! Broad spectrum
antibiotics for 5-10 days will lower
infection rate and mortality rate. The
greatest benefit is in Childs C and preEGD. Suggest cefotaxime, levaquin,
unasyn.
What about splanchnic vasoconstrictors
for example vasopressin. It restricts
blood flow via V1a receptors at the
celiac and SMA, but also to the
peripheral arterioles. Octreotide is more
specific for the splanchnic beds, mostly
over 5 days. Bleeding control was
better with octreotide versus
vasopressin with fewer side effects.
What about octreotide alone? The best
is octreotide plus endoscopy no
survival improvement, but secondary
end-points are better.
ENDOSCOPIC INTERVENTIONS
What about NGT placement? When swallowing,
the intra-luminal esophageal pressure
aspirates [blood or grounds] did not have a high-risk
can get up to 60-70 mmHg which is
endoscopic lesion] table 2 RUGBE investigators in
much greater than any NG tube
[Gastro Endos 2004 page 174]. Further, while a
placement; but the sensitivity [83%] and
positive lavage facilitates early endoscopy, there is
specificity [33%] of blood or coffee
no clinically meaningful improvement in outcome
grounds on NG tube aspirate for
when NG lavage is
predicting high-grade endoscopic
lesions is fairly poor [i.e. 17% of
negative NG lavages - no grounds, no
blood - had a high risk bleeding lesion
on endoscopy and 67% of positive NG

Notes 2014
The risk of bleeding goes by
varix size and stigmata.
Should banding be done?
The tip of the endoscope is
placed near the varix, there
is suction and then there is
banding onto the varix. It
may be technically
challenging with lots of
blood, but can be done. If
in the first 6 hours. The risk for rebleeding is
increased with PV thrombus, HVPG more
than 20 mmHg, and infection [probably
because of systemic hemodynamics].
Typically, when the risk is high, then GI
will go back and take a second look for
further therapy. If there is still bleeding,
then proceed to TIPS evaluation.
The other option is to break out the
Minnesota or Blakemore tube. The
difference is the number of lumens
[quadruple lumen versus triple lumen,
respectively]. The patient must be
intubated and sedated. They can be
placed orally or nasally, but epistaxis
can be caused because of the size. The
gastric balloon must be increased in size
to 250 cc, it is air-filled and it must be
pulled up on the GE junction and taped
to a football helmet or hockey mask.
This may create a hiatal hernia, or a
tear. If the gastric balloon is in the
esophagus, there can be perforation.
The next step is TIPS which is a
Seldinger technique through the IVC,
cannulate the right hepatic vein and
then, through the liver making a pass
into the right portal vein; there needs to
be an US Doppler before TIPS to make
sure its patent. Once in, they can pass a
wire through to the portal vein, they
then place a mesh stent to prevent the
parenchyma from closing down. It is a
technically difficult procedure.
the varix cannot be seen, injection of

sodium morulate can be used. Banding
has reduced infection, reduced ulcers
and less strictures as compared to
sclerotherapy. Banding is therefore
preferred when visualization is good.
74
10-20% of varices will re-bleed within 5

days [reason octreotide is given for 5
days]. Re-bleed is typically defined as 4
U pRBCs or hypotension
What about comparing the modalities for the
prevention of variceal re-bleeding? TIPS and surgery
out-perform endoscopy, but endoscopy has a lower
rate of encephalopathy. Mortality is roughly the
same between all three. There are new Teflon stents
that have much better one year patency 86 versus
47%, but they are technically more difficult.
What about outpatient prevention of rebleeding? So
the patient makes it through the hospital.
Patient needs outpatient repeat EGD in 2-3 weeks
[repeat banding prn], then q 2-3 weeks. PPI BID to
lower esophageal ulcer bleeding [co-

Notes 2014
morbid] and avoid NSAIDs.
Beta blockers
decrease re-bleeding too.
The prevention of recurrent
bleeding from known
esophageal varices [board
scenario]. Consider a
patient with a history of
COPD-asthma [so the use
of beta-blockers is
contraindicated]. Thus the
correct answer is repeat
esophageal variceal bandligation to reduce rebleeding. There is mention
of the use of nitrates as
well, but monotherapy
tended to worsen outcome
in cirrhotics. The use of
TIPS is reserved for those
en route to an OLT but if the
patient is an active
alcoholic, OLT is not in the
future. Some centers
perform splenorenal shunts,
there is probably less risk of
HE with a distal splenorenal
shunt.
IN SUMMATION
75
So with all the therapy when the
variceal bleeder comes in, 70% will stop
bleeding with medical management,
then get an outpatient EGD. 20% will
have early re-bleeding in house that is
stopped with repeat EGD in house. If
worsening bleeding or more varices in
house, patient should go for Blakemore
+/- TIPS [latter occurs in about 10% of
variceal bleeding].
In summary, variceal bleeding outcome
depends upon cirrhosis severity &
hemostasis achievement. Emergent EGD
after resuscitation has more than a 90%
efficacy [i.e. 10% go on to need TIPS in
house]. Antibiotics improve survival!
Octreotide is preferred over vasopressin
and if there is early re-bleeding, then
EGD should be repeated in house. For
uncontrolled bleeding, and selective rebleeders, balloon tamponade can be
done as bridge to TIPS. Secondary
prevention is as follows: follow up EGD,
BB, PPI and no NSAIDs.
FULMINANT HEPATIC FAILURE
Acute liver failure [ALF] is the onset of

sudden coagulopathy with an INR of
more than 1.5 and encephalopathy
within 8 to 26 weeks in a patient

Notes 2014
without prior liver disease thus
cirrhosis,
chronic hepatic disease, chronic
alcoholic hepatitis, HBV reactivation,
sepsis, etc. are exclusions [e.g.
cholestasis of sepsis] as these are preexisting conditions.
OVERVIEW
There are only 1-10 cases per 100,000
every year, so less than 3000 cases per
year in the US. The leading cause of
ALF in the US is Tylenol [APAP] OD, it is
not alcoholic hepatitis.
Identification of the cause of ALF is
therefore paramount because of the
varying treatment. Consider sending a
drug screen, acute Hep A and Hep B
serologies, but these are less common.
Rarer causes include autoimmune,
pregnancy- related may occur [third
trimester], HSV/CMV can do it, but rare.
Everyone should get an US Doppler as
this may be a result of hepatic venous
outflow. APAP overdose is far and away
the most common cause. The second is
unknown, the third is drug-toxicity.
Deterioration can be rapid and
unpredictable with the need for a
transplant evaluation acute liver
transplant has a 75% one year
survival. In a patient with ALF,
transplant evaluation includes: HIV,
CMV, HSV, hepA, IGM, HBsAg, HBcAb,
blood cultures, PPD, CXR, Liver US,
EKG, TTE.
What is the evidence based
management of ALF? There is benefit
for NAC for APAP and non-APAP acute
liver failure, H2 blocker [reduce the risk
of GIB, but also PPI likely of benefit] and
CVVHD as well as mannitol [for high
ICP, but small trials].
There is a low incidence of ALF so this
disease is very hard to enlist patients
7
h
into randomized controlled trials. Survival with

APAP, Hep A, ischemic do fairly well. Idiosyncratic
drug reaction, indeterminate dont do so well.
Remember, indeterminate, fulminant hepatic
failure or fulminant hepatitis B should receive
rapid evaluation for emergent transplant.

Notes 2014
Amanita toxicity treatment is lavage, IV
penicillin,
and dialysis. Autoimmune
ALF can be treated with
steroids. Budd-Chiari
treated with AC. Entecavir
should be used for acute
hep B but outcomes are
poor. Fulminant herpes is
treated with acyclovir.
Pregnancy C-section, but
usually post-partum
problems are the main
morbidity.
Wilsons disease should be
treated with rapid copper
chelation, but transplant is
inevitable.
TYLENOL TOXICITY
APAP overdose is treated
with lavage, charcoal, NAC.
APAP is about 50% of all
cases of ALF. How is the
history with APAP? More
than 6 grams at once. 4-6
grams over many days.
Prolonged fasting [depletes
glutathione], nausea,
vomiting, chronic ethanol
abuse increase the risk.
There is high AST and ALT [in the 1000s]

with normal bilirubins. For example, if
you see a bilirubin of 17, it is less likely
APAP toxicity. There is coagulopathy,
metabolic acidosis, and renal failure.
The Rumack nomogram is from the
early 80s and is only good for a single
time point ingestion.
Remember that, while uncommon in
APAP toxicity, clinical jaundice will
falsely lower APAP levels because it is a
colormetric assay.
NAC enhances glutathione stores, give
it until the INR is less than 1.5. Oral
dosing is suggested, but acetadote is
the IV form. The dosing is 140 mg/kg,
then 70 mg/kg every 4 hours to 72
hours or INR less than 1.5. The dosing
is different for
IV. It is 150 mg per kg for the first dose,
50 mg/kg for 4 hours, and 125 mg/kg
over 19 hours. It is an IV sulfa drug and
therefore there are concerns of sulfa
allergies.
Less than 1% die from acute APAP. Half
of the worst outcomes cases are
therapeutic misadventures. The serum
APAP levels tend to be low, having a
low level doesnt mean that it wasnt
APAP especially if the drug was take
over a period of many days. Usually
the patients are

Notes 2014
taking multiple products with APAP in
them, with
the majority being OTC. Many are on
Vicodin, etc. and then taking OTC
medications.
The encephalopathy tends to be worse
in non- intentional cases of APAP OD,
maybe with delayed presentation or
narcotic co- administration from the
un-intentional nature. These patients
should be in the ICU because of the
rapid neurological deterioration. INR is
important to follow as is creatinine.
ABG, lactates, ammonia are also
somewhat helpful. Blood sugar should
be followed closely. Liver biopsy
should not be performed as it does not
change recovery/mortality.
Infection is very prevalent as the
hepatocytes are being lost [which is the
largest source of Kupffer cells]. Some
suggest daily surveillance cultures as
infection in these patients is usually
bacterial in nature 80% with
bacteremia, fungal in about 20%.
Cephalosporins and vancomycin should
be considered, but avoid
aminoglycosides.
Fever can worsen HE so cooling is
suggested [avoid APAP for
temperatures]. Coagulopathy is
common and may appear like DIC. Only
about 10% have severe bleeding
despite coagulopathy. FFP and should
only be given if there is severe bleeding
or a procedure is planned.
NON-TYLENOL RELATED ACUTE LIVER FAILURE
There is no benefit for steroids in nonAPAP liver failure, nor enteral
decontamination, charcoal
hemoperfusion or albumin dialysis. In
acute alcoholic hepatitis [not ALF],
there is improved outcome with
steroids for a high discriminant
function.
h
78
Lee et al Gastroenterology 2009 found that there
was more transplant free survival in NAC group
especially in those with early stage
encephalopathy, but no improvement in overall 3
week survival with IV NAC versus placebo. There
was a trend to improvement in length of stay.

Notes 2014
NAC is well-tolerated with
an 8% increase in
nausea and rash. In kids
with ALF, there was not a
survival benefit with NAC;
transplant can be a
consideration. In acute
alcoholic hepatitis [again,
not ALF] there is suggestion
of benefit for NAC [NEJM
Nov. 2011].
CEREBRAL EDEMA
Cerebral edema is the main
cause of mortality in ALF. It
typically occurs only in
acute cases. In ALF it is the
result of ammonia,
ischemia, cytokines, toxins;
there is astrocyte swelling
and there is loss of cerebral
autoregulation.
The increase in glutamate to
glutamine in the astrocytes
leads to brain swelling as
water is pulled into the
brain/astrocytes.
Grade I encephalopathy
[EN] is slow mentation with
minimal or no change in
LOC. The patient should be
when placing and ICP monitor? There
is a 4-6
hour window for this procedure, but no
data. Be wary in those with
thrombophilia as there may be an
increased risk of clotting.
Once the monitor is in, the CPP should
be above 50 and ICP not more than 25
for 5 minutes.
Hyperventilate to temporize, give
mannitol, pentobarb coma but watch
out for hypotension and pressors [may
reduce cerebral blood flow].
What to do about Grade IV HE? Grade
IV is comatose, unresponsive to pain,
79
placed in experienced nursing care in
quiet room. Grade II there is
disorientation, drowsiness, with
asterixis, inappropriate behavior. Grade
III are incoherent, typically intubated,
usually with renal failure, ICP
monitoring should be considered at
grade III.
Hemodynamic management of these
patients is difficult, CVVHD will most
certainly be needed. Levophed is
preferred over vasopressin and try to
minimize PEEP. An ICP monitor is
suggested because CT is insensitive
and physical exam is poor in terms of
managing the patients ICP.
Neurosurgeons will push back as there
is no data to support ICP monitoring,
and typically theres a bleeding risk.
In those with ICP monitoring [Liver
Transplant 2005] all were intubated,
majority were on dialysis, and majority
were listed for transplant. In the ICP
group, there was more ICU intervention
[more mannitol, more barbs, and more
pressors] but no change in outcome.
There was no difference in listing rate
or transplant rate.
Michigan consults neurosurgery for all
grade III or IV HE. Does one use rFVIIa
to lower bleeding risk
with decorticate or decerebrate posturing. Liver
transplantation [ALT], ABO compatible whole liver
usually takes about 2-3 days for critically ill ALT.
Again 75% one year survival.
What about hypothermia? There is pilot data with
hypothermia to 33 centigrade, often heavy
propofol is needed because of shivering. What
about worsening coagulopathy, infection risk?
Rewarming method? There is no study as yet.
PROGNOSIS
The old modified Kings College Criteria from the
early 2000s, differentiated APAP from non-APAP
ALF. For APAP OD bad prognosis was lactate more
than 3.5 in 4 hours, or lactate more than

Notes 2014
3.0 in 12 hours, pH less than 7.3, INR
more than 6.5, creatinine more than
3.4, or stage 3 or 4 HE. For non-APAP,
prognosis was best predicated by INR
more than 6.5, or 3 of: INR more than
3.5, age more than 40, Bili more than
17.5, Jaundice more than 7 days.
Understand the MELD. It was initially
derived to predict mortality in patients
with portal hypertension awaiting TIPS.
It is composed of three components 1.
INR 2. Creatinine 3.
Bilirubin. It is a burdensome calculation
but became predictive essentially of all
patients with chronic, advanced liver
disease in terms of survival. In patients
with a MELD less than 16, medical
treatment is superior to liver
transplant. The MELD is now used for
allocation of livers. A
80
MELD of 25-29 has a median survival of
200 days,
30-37 is 110 days and more than 37 is
about 50 days median survival. MELD
is also used to predict survival in
chronic cirrhosis with various other
insults such an infection and variceal
bleeding. The use of the MELD score
has led to a 15% reduction in mortality
in the chronic liver failure population
because of improved organ allocation.
MELD is not used in patients with acute
fulminant hepatic failure awaiting
transplant.
GI BOARD SCENARIOS
Understand caustic ingestion [see also

section 9]. A 15 year old swigs a
whiskey bottle and it is 25% sodium
hydroxide. The next step is admission
to the ICU and an emergent endoscopy
to evaluate the extent of injury.
Alkaline ingestions tend to be more
injurious than acid and tend to injure
the esophagus while acid ingestion
tends to injure the stomach. Emesis,
drooling and stridor portend a more
serious injury to the GI tract.
Esophageal necrosis and perforation
carries a very high mortality. Deep
esophageal ulcer can result in
esophageal strictures at 14 days,
superficial injury has a good outcome.
The patient should have an urgent
endoscopy once perforation is ruled
out. Stage 3 injury [the worst kind of
burn] carries a high mortality, and high
risk of stricture. Some centers perform
esophagectomy for stage 3 burns. The
injury to the mucosa is instantaneous
so lavage and neutralization is futile.
Emesis and charcoal are
contraindicated given the risk of
perforation. The role of antibiotics is
controversial, so too are steroids.
Perforation is a contra-indication to
steroids.

Notes 2014
The treatment of acalculous
cholecystitis. Consider an
elderly CVA patient with
COPD and CAD who has an
aspiration pneumonia but is
81
persistently febrile despite being on
broad spectrum antibiotics. His ALP and
bilirubin are slightly elevated which
prompts a RUQ US and this shows a
thickened gall bladder wall with
h
sludge without a dilated common duct or
cholelithiasis. The treatment of choice
is a percutaneous cholecystostomy
because mortality rates in some studies
are as high as 50%. It is thought that
acalculous cholecystitis is a form of
ischemia reperfusion injury, which
occurs in the critically ill. Risk factors
for development of this disorder include
biliary stasis secondary to ileus, fasting,
dehydration, narcotics, or mechanical
obstruction of the biliary tree,
mechanical ventilation, total parenteral
nutrition, recent surgical intervention,
and vasopressor support.
Patients often present with
undiagnosed fever or subtle signs of
infection without a clear source after
careful evaluation. However, fulminant
sepsis secondary to acalculous
cholecystitis is described.
Recognize and treat ascending
cholangitis. Note that the patient has
Reynolds Pentad which is Charcots
triad plus altered mental status and
hypotension. Charcots triad is fever,
jaundice and RUQ pain. The RUQ US
reveals a dilated common bile duct
without evidence of stone.
This is still ascending cholangitis and
this patient needs an ERCP which is
very effective at treating biliary sepsis.

Notes 2014
7
h
PHYSIOLOGY OF SEPTIC SHOCK

Septic shock is about 40% gram
negative, 50% gram positive and the
rest fungi. Septic shock is the leading
cause of death in non-coronary ICUs.
The usual hemodynamics of sepsis is
typified by hypotension with increased
cardiac output.
Nevertheless, there is impaired
myocardial function that lasts for 24-48
hours.
The nidus of infection such as a
pneumonia or abscess allows an
organism to proliferate or enter the
blood stream. Exotoxins can be
released into the blood stream.
Endotoxin can be released or a
structural component of the bug itself
[e.g. teichoic acid, LPS] which cause a
cytokine cascade. All organs can be
affected.
THE CARDIOVASCULAR RESPONSE
Sepsis causes a stunned myocardium.
The EF tends to increase over a period
of 10 days. The survivors of sepsis
tended to drop their ejection fractions
the most. The end-diastolic and endsystolic volumes increase during
sepsis. Parrillo NEJM 1993 looked at the
left ventricle during the acute phase of
septic shock and found LVEDV to LVESV
values of 225 ml to 150 mL with MAP of
40, CVP of 2, HR of 150 and a cardiac
output above 11. The EF was in the low
30s. During the recovery phase, LVEDV
to LVESV was 150 to 75 mL and EF of
50%, HR of 70 with a cardiac output of
5.3. Dilation of the left ventricle seems
to confer a mortality benefit, this may
be a compensatory response to
maintain stroke volume. It is thought to
be due to improved LV compliance during

sepsis such that LV volume at end-diastole
can increase. But there is also a
myocardial depression seen that is due to
TNF and IL-1.
5. CRITICAL CARE INFECTIOUS

DISEASE
Notes 2014
Microvascular endothelial
dysfunction is also a problem in
severe sepsis.
TREATMENT - ANTIBIOTICS
The treatment of septic shock
requires: early antibiotics. In mice,
withholding antibiotics from the
onset of infection leads to rapid
death at about 15 to 18 hours.
In humans, for every hour after the
onset of hypotension, the incorrect
or absence of appropriate
antibiotics increased mortality by
7.6% per hour.
TREATMENT VASOACTIVE SUBSTANCES
High dose versus low dose
vasopressin have differential effects
on splanchnic blood flow. Low dose
vasopressin [i.e. the doses used in
the VASST trial] tends to increase
splanchnic flow.
Low dose dopamine increases heart
rate and contractility and causes
some vasodilation while high dose

dopamine causes more
vasoconstriction. Norepinephrine
increases heart rate and contractility
about the same as dopamine but has a
much more pronounced vasoconstriction effect. Dobutamine is
essentially all HR, contractility and
vasodilation. Isoproterenol does so to a
greater extent [compared to
dobutamine]. Epinephrine has potent
HR, contractility and vasodilatory and
vasoconstrictive effects across the
board, whereas vasopressin and
neosynephrine are pure
vasoconstrictors [AVP a little more so].
What did the VASST trial show? There

was no difference between NE and NE
+ vasopressin in severe sepsis and
septic shock. There was not an increase
in ischemic events with vasopressin as
feared. The interesting subgroup was
those with low NE requirements at the
onset [less than 15

Notes 2014
mcg/min] who had vasopressin added
tended to improve outcome. This was
not what the authors had expected and
there was no good explanation for
these effects.
What about the SOAP II trial [NEJM
2010]? There was no difference
between dopamine and norepinephrine
in the overall outcome.
However, dopamine may worsen
cardiogenic shock. Overall there was a
trend to improved outcome with NE.
Dopamine lead to significantly more
arrhythmia.
What about the CATS trial? This was in
Lancet. It compared NE plus
dobutamine to epinephrine in severe
sepsis and septic shock. There was, in
effect, no significant clinical difference
between the two arms.
TREATMENT EARLY GOAL DIRECTED THERAPY
2014 has been a bad year for Dr.
Rivers. Both the PROCESS and ARISE
trials in NEJM have effectively
deconstructed the EGDT algorithm,
showing that blind application of his
initial protocol to all severely septic
patients does not improve outcome this
day in age.
In the Rivers trial, [NEJM 2001] in the
first 6 hours, the protocol group
received 5 liters of fluid, compared to
nearly 4 liters in the control group. In
both groups about 25% received
vasopressors of some sort. Whats
interesting is that in the PROCESS trial,
the ratio of fluids to pressors seemed to
switch in all of the study arms [as
compared to the original Rivers study];
that is, in 2014 there has been a trend
to use much less fluid and much more
vasoactive substances when treating
septic shock as compared to the
original 2001 Rivers trial. Further, as
h
compared to the Rivers trial, the baseline mortality
in PROCESS was lower.
The recent TRISS trial [NEJM 2014] confirmed what
many had believed since 2001, that transfusing to
a Hb target of 10 g/dL does not confer benefit in
severe sepsis and septic shock.

Notes 2014
TREATMENT STEROIDS
The Annane study [JAMA]
looked at patients with
severe sepsis and septic
shock. The mortality rate of
the control group was 60%.
There was an improvement
in mortality with the
provision of steroids in the
setting of relative adrenal
insufficiency [as determined
by a cosyntropin stimulation
test]. Then CORTICUS
showed no increase in
survival when steroids were
given in septic shock;
instead steroids improved
time on the ventilator but
increased risk of
subsequent sepsis.
COMMON INFECTIOUS
HIV
In terms of HIV, the

common pulmonary
disorders are
pneumococcus [most
common & tends to be more severe],

haemophilus, PCP, TB and the
atypical/viral syndromes.
The uncommon pulmonary disorders in
HIV are: aspergillus, histo/cocci, staph,
toxo, lymphoma and Kaposi.
The rare pulmonary disorders in HIV are
CMV, MAC and HSV. CMV can cause
pneumonia in transplant patients, it is
very, very uncommon to be a pathogen
in HIV/AIDS patient.
Other causes of dyspnea in AIDS CHF,
PH, PE, drug toxicity, neoplasms, septic
emboli.
PCP is a fungus that is transmitted via
respiratory secretions. Therapy with
monoclonal antibodies [e.g. adalizumab
- for MS] have been linked to PCP! CD4
counts are almost always below 200 in
patients with PCP [but 10% can have a
CD4 above 200]. The risk of PCP
with lower CD4 counts [54%
DISEASEincreases
SCENARIOS
ON THE BOARDS
with a CD4 count less than 50; 35%
with a CD4 count between 50-200].
AIDS patients tend to develop
pneumatoceles. Sometimes there will
be an infiltrate in one lung and a PNTX
in the other. In general, PCP presents
with

bilateral GGO and thin-walled cystic
lesions. Pleural effusion and
lymphadenopathy are really uncommon
in PCP. PCP in HIV-negative patients
typically have a poor prognosis and the
presentation typically occurs in the
setting of steroid tapering [patients
who have been on more than 20
mg/day of prednisone for more than
two months].
What is the diagnostic approach of
choice for PCP? Sputum or BAL stain or
immune- fluorescence. A serum or
lavage PCR for PCP is sensitive, but not
specific. So if its negative, its helpful.
Induction of sputum can diagnose PCP
if the RT is good.
If the patient has a PaO2 less than 70
on air or an Aa gradient more than 35
mmHg, the patient should get steroids,
plus 21 days of Bactrim.
Pentamidine can cause pancreatitis
and insulin release leading to low BG.
Consider a patient with known AIDS, a
WBC count of 3,000 with 3%
lymphocytes [total of 90 lymphocytes,
and maybe half of these are CD4] with
cough, fever, infiltrate and recent
Central Valley exposure. This patient
may have coccidioidomycosis. The best
way to obtain cocci on respiratory
secretions is a pap smear [better than
KOH and calcofluor]. Blood and bone
marrow cultures for cocci are rarely
diagnostic and take days to grow.
Similarly, CSF studies are poor. The
gold standard tends to be serologic
tests, but even these can be negative
in advanced AIDS patients.
Moving on to toxoplasmosis. This is a
CNS mass lesion in a patient with AIDs.
Its almost always either toxoplasmosis
or lymphoma. If the CD4 count is high,
it is lymphoma. Toxoplasma PCR is
quite specific, but only 50% sensitive and becomes

negative quickly with therapy.
You need to obtain a CSF for toxoplasma and EBV
titre the latter making it more likely lymphoma. If
CSF cannot be obtained, two weeks of
FEVER & RASH
treatment for toxoplasma and re-
imaging should
76
be done. If there is no
improvement in the mass
lesions, it is most likely
lymphoma.
The treatment for
toxoplasmosis is
sulfadiazine and
pyrimethamine. Steroids
are used if there is edema
and anti-convulsants only if
the patient seizes. Other
causes of brain-lung
disease in patients with HIV
include: TB, nocardia,
rhodococcus.
Anti-retroviral therapy for
the intensivist giving
inadequate doses can
harm the patient as
resistance can occur
quickly. It is better for the
patient to be on no HAART,
than to give the patient
inadequate doses.
This can be a viral exanthema

[measles, varicella, mono] or a
bacterial exanthema [staph,
pneumococcus, meningococcus,
leptospira, ehrlichia, rickettsia].
On the boards, summer sepsis

shouldnt die without doxycycline as
this could be Rocky Mountain Spotted
Fever [or Borrelia or Ehrlichia]. The
rash can easily be missed and the
remaining symptoms are very nonspecific. Tick exposure, south or
southeast US is more common. But
look for low platelets and
hyponatremia, liver enzymes might be
slightly high. Immuno- fluorescent
stain of a skin biopsy is the diagnostic
test of choice.
Understand the tick-borne zoonoses
often present with dermatologic
manifestations. Consider a patient with
profound ARDS over 10 days that is not
improving on fairly broad anti-microbials.
You learn that the patient pursues
outdoor activities in New England; it is
the summertime. Because all of the
following can cause profound ARDS
Lyme, Ehrlichia, RMSF, and Babesia
and because it is typically clinically
impossible to distinguish between
these infections and because coinfection with these zoonoses can
occur in up to 16% of patients [the
white-footed mouse
Vibrio vulnificus is a gram negative found in warm

marine water. It can cause a primary wound
infection if it gets into a cut [large, violaceous
plaques and bullae] and also a septicemia in
cirrhosis patients who ingest raw oysters.

which is the mammalian reservoir for
these diseases is 40% co-infected] it is
prudent to treat for all of them.
The treatment of Lymes [Borrelia
burgdorferi], Ehrlichia, and RMSF
[Ricketsia ricketsii] is tetracycline or its
derivatives. The treatment of lifethreatening babesiosis [B. microtii] is
clindamycin and quinine. The whitetailed deer which is the definitive host
for these organisms is increasing in
population so these diseases are
becoming more common.
Ehrlichia can occur anywhere from New
England to the South and South Central
US; it may present with petechial rash,
but this is uncommon.
Rickettsial disease is ubiquitous,
despite the name Rocky Mountain.
Key to the history were outdoor
pursuits, anemia, thrombocytopenia
and sometimes hyponatremia with
elevated liver enzymes.
Recognize ecthyma gangrenosum and
that the most common cause is
pseudomonas. Other causes of this
foul-looking rash are staph, HSV,
candida and mucor but pseudomonas is
the most common cause. This rash is
typically seen in immunocompromised
patients, but also in diabetics, burn
patients and the severely malnourished.
These lesions are most commonly found
in the perineum, but also on the
extremities, trunk and face. They
typically begin red purpuric macules
that then become vesicular, indurated,
bullous, pustular and finally
hemorrhagic. Then within 12-24 hours a
gangrenous ulcer begins to form. Biopsy
typically reveals bacteria invading the
veins. Patients with ANC less than 500
rarely survive. Treatment is with an
anti-pseudomonal antibiotic and
surgical debridement.
www.heartlung.org
JE
K

Review Notes
diplococcus that causes meningitis and

septicemia in epidemics. Mortality was
once 70- 90% and it is now 10%.
Cefotaxime and ceftriaxone [third gens]
are effective therapy for this bug.
Adequate treatment of N. meningitidis
has been seen with just 3-4 days of
h
antibiotics without failures, though
treatment is usually 7 days.
Vancomycin is usually initially added,
but can be stopped once N.
meningitidis is found. Other treatments
have not shown benefit including:
activated protein C, heparin in patients
with DIC and dexamethasone is really
only beneficial in patients with H.
influenzae or S. pneumococcus as
above.
INFECTIOUS DISEASES WITH
NEUROLOGICAL SYMPTOMS
77
Pneumococcal meningitis
has a high resistance rate,
this should be initially
treated with vancomycin
and ceftriaxone. If the
patient is over 50, the
patient may have listeria
and ampicillin should be
added.
Recognize ciguatera toxin. This is a type

of poisoning that occurs in people who
have eaten fish within 5 hours of
symptom onset. The toxin is heat
stabile, so cooking does not inactivate
it. It is common in fish from warm
areas such as Hawaii and Florida [e.g.
red snapper]. The toxin arises from
diatoms, but the fish bio-concentrate it.
The first symptoms are invariably
gastrointestinal and this then
progresses to neurological symptoms
that include paresthesia of the lips and
extremities and then even paralysis of
the respiratory muscles [the patient ate
red snapper and then was intubated
about 6 hours later]. Death has been
reported, but is rare.
The use of
dexamethasone? In
children, there is a
reduction in neurological
complications [mostly in H.
influenza meningitis], in
adults, it is controversial
but probably should be
used in severe disease in
pneumococcous or H.
influenza with the first
dose or before the first
dose of antibiotics.
N. meningitidis. This bug is
a gram negative
lung.org
JE
K

Review Notes

Notes 2014
Mackerel and tuna [scromboid fish] can
cause
scromboid poisoning is a histamine
reaction from the spoiled fish that
involves characteristic histamine
responses in the body including
flushing, cramps and diarrhea,
headache, mouth burning, nausea and
vomiting. Bronchospasm, and
wheezing may occur, treatment is
supportive and death is rare.
Diplopia and dysphagia 2-24 hours
following the ingestion of canned foods
should prompt an investigation for
botulism toxin. There is typically a
descending weakness that is bilateral
and symmetrical. There is no fever.
Botulism is caused by enteric entry of
botulism toxin and is treated by
enemas and lavage to remove the
toxin. Anti-toxins may also be
administered.
The third most common cause of food
poisoning in the US [after Salmonella
and staphylococcus] is
C. perfringens food poisoning. It is
associated with poorly cooked meats or
meat products such as gravy. Diarrhea
and cramping abdominal pain 7 to 19
hours after ingestion is typically how
this presents. It lasts for a few days. In
rare cases, the gut can become
necrotic and perforate which leads to a
very high mortality. If in the blood, C.
perfringens may cause massive
hemolysis.
Recognize Tick Paralysis in a young
women who recently returned from
camping. The distractor is that she ate
predominantly from canned food.
However, she has no bulbar symptoms
which would make botulism rather
unlikely. Tick paralysis is treated by
removing the Tick which is usually
persistently feeding and releasing a
neurotoxin into the patient which
causes a bilateral and symmetrical
h
91
paralysis with preserved sensation. Patients rapidly
improve once the Tick is removed.
Recognize West Nile virus infection. Consider a
young, transplant patient on multiple immunesuppressives who develops proximal muscle
weakness, then bilateral lower extremity with
unilateral upper extremity flaccid paralysis then

Notes 2014
bilateral upper extremity
flaccid paralysis. Both
polio and WNV can present
exactly like this also with
bulbar and respiratory
symptoms. The kicker is
his LP with reveals 10 WBC
with a PMN predominance
and high protein, this rules
out GBS and makes it
more likely infectious.
Further GBS presents with
ascending paralysis and
frequent cranial nerve
findings, rather than
proximal in this patients
case. GBS should also
reveal markedly slowed or
blocked peripheral nerve
conduction which is not
seen in this patient.
Treat tetanus. The patient
sustained a fall and cut 1014 days prior to presenting
with trismus and muscle
spasms, neck stiffness and
leg stiffness. She should
be treated by being placed
into a quiet, dark ICU room
and administered
benzodiazepines,
metronidazole, and
Fever is not characteristic of C. diff; it is
present in
10-15% only. Unexplained leukocytosis
is common with 50% having more than
15K and 25% with WBC above 30K. C.
difficile is NOT most reliably diagnosed
by stool culture. Many asymptomatic
adults and children have this bug
cultured in their stool, but it does not
mean disease. Further, 20% of
asymptomatic patients in the hospital
harbor this bug in their intestine,
though the majority of these isolates
are nontoxic forming.
92
tetanus anti-toxin. The patient should
NOT be treated with elective intubation
as this is a profound stimulus for
muscle spasms. The patients are often
given an elective tracheostomy at a
later date. In some patients,
neuromuscular blockade is required to
treat the spams.
C. DIFFICILE
There is a broad range of symptoms,
the more severe it is, the more
abdominal symptoms and risk of
dilatation and perforation of the gut.
C. diff has been associated with every
kind of antibiotic [including
metronidazole and vancomycin] and
even anti-neoplastic agents that have
anti-microbial activity [e.g.
methotrexate].
C. diff causes 25% of all antibiotic
associated diarrhea, but is the most
common cause of nosocomial
diarrhea. C. diff may present in a
whole host of manners from mild
diarrhea and unrest, to profound
fever, leukocytosis, shock and death.
C. diff of the proximal colon may
present without diarrhea [10% of
patients] and this is the kind that can
present as mega-colon, perforation
and death.
The most reliable test to diagnose C. diff is the
tissue culture for cytotoxicity of cytotoxin B.
However, this test is expensive and takes days.
Most labs use ELISA from the stool for toxin A and
B but not all labs detect toxin B which will miss
patients with C. diff. Additionally, even after three
negative ELISAs for A and B, some patients will
still be detected by the cytotoxicity assay [this can
be 5-10%] so in patients with a high suspicion with
negative stool ELISAs, the cytotoxic assay may be
required. Currently, the approach to diagnosis is
one stool for PCR, but no more. They will have a
positive PCR for a long time even after treated.

Notes 2014
What is the treatment of choice for
severe C. diff in the ICU? Oral
vancomycin plus IV flagyl [CID 2007].
FEVER AND HEMOLYSIS
Recognize and treat falciparum malaria.
The patient returned from West Africa
and has high fevers, anemia,
thrombocytopenia, hyponatremia and
ARDS. Blood smear reveals intracellular parasites. The patient has a
parasite load above 11%. Parasite
density can help monitor outcome.
Greater than 5% is quite bad and in sick
patients, should be checked every 4
hours. Bacterial pneumonia can
complicate things with falciparum
malaria, but ARDS from micro-vascular
unrest is a common event in malaria
and may affect the lungs. A quinoline
[quinine, quinidine] or artemisinin plus
a tetracycline derivative [or doxy
93
or clinda] is the treatment. The
quinoline
derivatives are the most effective
against falciparum. IV quinine is no
longer available in North America, but
IV quinidine is. It carries toxic effects
QT prolongation, arrhythmia,
hypotension, and hypoglycemia. It acts
by antagonizing heme polymerase
which causes heme to build up and
heme is toxic to the parasite.
Artemisinin binds iron and creates free
radicals which is also toxic to the
parasite.
Pentoxyphylline is a phosphodiesterase
inhibitor that lowers cytokine levels,
and was previously recommended, but
no longer. Exchange transfusion was
once suggested to reduce parasite
load, but it did not improve mortality
either.
Recognize other infectious causes of
massive intra-vascular hemolysis.
Consider a patient who is profoundly
septic after eating old gravy with
abdominal symptoms and serum that
is blood red. Note that beta-hemolytic
strep will cause hemolysis on an agar
plate, but not in vivo.
Massive hemolysis of an infectious
variety has a narrow differential
including C. perfringens, falciparum
malaria, babesiosis, and bartonellosis.
Bartonellosis (Oroya fever) is caused
by a pleomorphic Gram-negative
bacillus and is transmitted by sand
flies. It is endemic to mountain valleys
in Peru, Columbia, Ecuador and Bolivia.
Babesiosis is a malarial-like illness
transmitted by ticks from animals to
man; it can cause massive intravascular hemolysis. Cases have been
reported in the United States from
Long Island, Nantucket, and Marthas
Vineyard. As above, it may present
with ARDS and commonly co-occurs
with other tick-borne diseases. Rash is

Notes 2014
not a common feature of
Babesiosis.
Noninfectious causes of
massive intravascular
94
hemolysis include paroxysmal nocturnal
hemoglobinuria, the hemolytic uremic
syndrome, which can sometimes follow
Escherichia coli and

Notes 2014
shigella infections, paroxysmal cold

hemoglobinuria, and certain snake
venoms.
Which nosocomial infection has increased over the

last 10 years? CRSBIs, VAPs, CAUTIs have all
decreased; C. difficile has increased.
WEIL SYNDROME
The top 4 healthcare-associated infections are:

coagulase negative staph, staph aureus,
enterococcus and candida species.
Recognize and treat leptospirosis.

Consider a patient who is a young
veterinary technician who likes to camp
and hike in the southern US. L.
interrogans is a spirochete and a
zoonotic infection. Dogs, horses,
rodents, cattle and swine may all carry
the disease. It is transmitted via
contact with animal urine or infected
soil and water. It is common in tropical
locations such as Hawaii and among
those who are exposed to freshwater
and to animals. Leptosporosis in
pregnancy, especially early pregnancy
can be devastating to the fetus.
Conjunctival suffusion is the classic,
specific finding but it is not terribly
sensitive. The disease may be mild, or
fulminant with the organs most
affected being the liver, kidney and
spleen. A severe form is Weil syndrome
which is hepatosplenomegaly,
jaundice, hyponatremia and renal
failure can be present.
Cases of leptospirosis that require the
ICU have a greater than 50% mortality
and those with ARDS and or CNS
involvement may have a higher
mortality. While the disease is usually
mild and self-limiting, severe cases
should be treated with doxycycline;
ceftriaxone will also work.
Rocky Mountain spotted fever may also
present with Weil syndrome, but it will
be associated with a tick bite and a rash
of the palms and soles. It is caused by
Rickettsia Rickettsii it is typically
associated with marked
thrombocytopenia.
NOSOCOMIAL INFECTION
The top cause of CRBSIs is

coagulase negative
staph, for CAUTI its E. coli,
for VAP its staph aureus and
for skin and soft-tissue
infection its s. aureus as
well.
TRANSMISSION & ISOLATION
The most common cause of
transmission of nosocomial
infections is by contact.
Large droplets travel a few
feet and drop with gravity,
small droplets [less than 5
microns] stay airborne for
hours.
Seasonal flu is spread by
large droplets. Therefore,
the patient must be 3-6
feet away from anyone not
in a standard surgical
mask.
Standard isolation requires
hand hygiene, droplet
precaution requires hand
hygiene, private room if
possible and a surgical
mask within 3 feet of the
8
h

Notes 2014
patient. Contact precaution requires
hand hygiene, a private room, gloves
and gown.
Airborne precautions require hand
hygiene, private room and an N95. All
airborne isolation rooms require
negative pressure with no air
recirculation unless HEPA-filtered.
Pertussis, mumps, and N. meningitidis
are all droplet precautions. There are
three common diseases that require
airborne: measles, chickenpox and TB.
Meningococccus post-exposure
prophylaxis. The only people who
require PEP are those who are
intimately exposed to the patient with
direct fluid contact in an enclosed
space for 60 minutes or longer.
Because C. difficile is a spore-former, it

is hardy and only killed by bleach.
Nosocomial bacteria that can cause
infection via environmental
contamination are: C. diff, VRE, MRSA,
acinetobacter, pseudomonas, norovirus,
HBC, HCV, aspergillus, mucor and
rhizopus rhizopus likes to get into the
blood stream and cause infarction.
CATHETER-RELATED BLOOD STREAM
INFECTION

Notes 2014
Short-term catheters are contaminated
via the
skin, long-term catheters via the lumen.
The prevention of CRBSI should include
the use of a scrub cap. This is because
it was a part of a bundle that was
studied and shown to improve CRBSI
rate [like blood transfusion in the EGDT
trial]. It is hard to know which aspects
of the bundle improve outcomes, but
its there.
The use of coated catheters may
improve CRSBIs if the use of bundles
does not make the institution CRSBI
rate very low. Chlorhexidine patient
cleaning can prevent infections as well.
The reduction of bacterial colonization
of central venous catheters is facilitated
by: silver impregnation, antibacterial
impregnation, tunneling, heparin
coating. The use of the plastic sheath
over the PAC has not been effective at
reducing colonization. Central venous
catheters impregnated with sliver and
other antimicrobials have been shown
in RCTs to reduce the risk of CRBSI.
They only reduce the risk, however, if
the institutions baseline CRBSI is above
2%. Below this risk, they do not
improve outcomes. Routine changes of
CVCs do not lower the risk of infection
and routine changes over a wire tend to
increase the risk of CRBSI.
Peripheral blood and catheter tip
growing the same organism defines
catheter infection; or if the hub is
culture positive 2 hours earlier or if
there is more than 3 x the bacterial
load from the catheter culture as
compared to the peripheral blood.
The most common pathogens from
CRSBIs are coagulase-negative
staphylococci [may be over- called],
staph aureus, enterococcus, gram
negative rods and candida.
h
97
What is the treatment course? Vancomycin should
be, essentially, first line. If the MIC are 1.5 or more,
then some suggest using other agents as there
could be worse outcome. Daptomycin is one such
suggestion. Gram negative coverage is

Notes 2014
also a consideration in
some populations.
Candida treatment is
nebulous. Femoral
catheters in ill patients
should be considered for
candida treatment, or
sepsis on TPN, heme
malignancy, known candida
colonization at multiple
sites. Candida treatment
should include an
echinocandin. Fluconazole
should be considered only
if there has been no azole
in 3 months and the
likelihood of glabrata or
kruseii is considered very
low. [For the board exam,
the answer is never
fluconazole in this
situation].
When deciding to pull a
CVC, which is the least
likely to indicate an infected
catheter? A femoral versus
a peripheral catheter
should not sway your
decision. If the patient has:
positive blood cultures
without a focus of infection,
hemodynamic instability,
erythema at the catheter
metastatic seeding], nonimmunocompromised,
without hardware, a negative TEE, with
fever and BSI resolved within 72 hours,
and the patient is stable without signs
of further infection, then 2 weeks of
antibiotics are OK. If the patient has a
positive catheter tip for S. aureus, but
negative peripheral blood cultures,
then 5-7 days is OK with close follow
up [but this is an unusual situation].
Interestingly, enterococci and GNBs are
treated for only 7-14 days because
these bugs are serum susceptible that
is easily opsonized.
98
site or pathogens that are likely to
cause CRSBIs, then strong consideration
should be given to removing the
catheter.
Long-term catheters requires a higher
threshold for removal e.g. suppurative
phlebitis, endocarditis, BSI for more
than three days into appropriate
antimicrobial coverage, S. aureus,
pseudomonas, fungi or mycobacterium.
If its a short-term catheter, then any of
the above should also prompt removal.
Less virulent, but difficult to eradicate
microbes should also be considered
when removing a catheter.
If a patient has bacteremia with a line
in, there are various decision points
about duration of treatment. For
coagulase negative staph aureus with
an uncomplicated catheter-related BSI,
5-7 days of antibiotics is OK if catheter
removed. If there is a desire to keep
the catheter, then 10-14 days with an
antibiotic lock. Interestingly, if the
catheter is removed and if blood
cultures repeated are negative, then
no antibiotics are needed.
For staph aureus, the catheter should
be removed and then 4-6 weeks of
antibiotics. If the patient is non-diabetic
[increases the risk of
Candida is also two weeks. There should be eye
exams at the end of this 14 days as 3% can have
endophthalmitis. Duration is determined from the
first day with a negative blood culture for all of the
above.
When to perform an echocardiogram? If a shorter
course for staph aureus [two weeks] is planned,
then a TEE is recommended. If there is persistent
staph aureus bacteremia. Wait 5-7 days following
diagnosis of BSI to obtain a TEE to allow the
vegetation to ripen. Repeat the TEE earlier if there
is persistent fever or BSI 72 hours beyond removal
of the catheter.

Notes 2014
What not to do? Dont: culture tips
unless suspected infection, order
qualitative tip cultures [broth], culture
the subcutaneous segment of the CVC
[unless its a PA introducer], under-fill
blood culture bottles [some labs weigh
the culture bottles to see if there is 510 cc of blood sampled], start linezolid
for suspected CRBSI, use thrombolytics
adjunctively, routinely re-culture after
stopping therapy [except dialysis
catheters].
VENTILATOR ASSOCIATED PNEUMONIA
How to reduce the incidence of VAP?
Aseptic technique, ventilator bundles
and ETT coatings [silver or
chlorhexidine coatings]. Changing the
ventilator tubing regularly does not
reduce the risk of VAP.
99
Measures to reduce VAP are plenty.
Sub-glottic
suctioning and oral decontamination
have proof, so too do SBTs and
elevation of bed, but regular changing
of ventilator tubing does not reduce the
risk. The CDC guideline for prevention
of VAP and other complications
includes peptic ulcer disease
prophylaxis, and DVT prophylaxis.
Closed versus open tracheal suctioning
does not alter VAP rates [i.e. a
protected tracheal suction catheter].
The diagnosis of VAP is very difficult. In
JAMA 2007, there was an article on
diagnosis of VAP. The presence or
absence of fever, abnormal WBC or
pulmonary secretions do not alter the
probability of VAP. The combination of a
new infiltrate with at least two of:
fever, WBC, or purulent secretions as
an LR of 2.8. The lack of an infiltrate
lowers the likelihood of VAP with an LR
of 0.35. Less than 50% PMNs on cell
count of secretions makes VAP unlikely
[LR 0.05 to 0.10].
How is VAP treated? You need to treat
both MRSA and pseudomonas.
Antimicrobial therapy in the preceding
90 days, in house for 5 or more days,
high frequency of resistance in the
community, hospitalization for 48 hours
of more in the last 3 months, residence
in a nursing home, home infusion
therapy, dialysis within one month,
home wound care, family member with
MDR bacteria are all risks for MDR
bacteria.
Legionella is a possible pathogen in
VAP as is influenza.
There is no evidence that combination
therapy improves gram negative
pneumonia. In a neutropenic patient
with pseudomonas bacteremia, there is
evidence that two drugs work for gram
negative infection.

Notes 2014
Treat for 8 days, unless
acinetobacter or
pseudomonas, then give
two weeks.
UTIs are caused by a biofilm between
the urethra
and catheter, and by accessing the
drainage port in a non-sterile manner.
SURGICAL SITE INFECTION
Surgical site infections can be reduced
by prophylactic antibiotics within 3060 minutes of incision and for no
longer than 24 hours. Razors should
not be used as they increase the risk
of infection. The surgical site should
be cleaned with chlorhexidine.
Glucose control and normothermia
should be attained.
As of October 2008, medicare stopped
paying for IV catheter infections,
mediastinitis post heart surgery and
catheter-associated UTIs, decubitus
ulcers, fractures or other injuries,
objects left in during surgeries, air
embolism, blood incompatibility, VTE
following orthopedic replacements and
poor glycemic control.
A FUNGUS AMONG US
What is the difference between yeasts

and molds?
Yeast are round, single-celled
organisms [e.g. candida, cryptococcus,
and the endemic mycoses] while molds
are filamentous [e.g. aspergillus,
fusarium and mucormycoses the
difference between aspergillus and
mucor is important because they are
treated differently].
In which fungi is a serum beta d-glucan
test important? It is very good for
most fungi, but NOT mucor. Blood
cultures are probably the best for
candidemia. Galactomannan is positive
for molds [e.g. aspergillus, fusarium]
CATHETER-ASSOCIATED URINARY TRACT
100
INFECTIONS
and not yeast and NOT mucor. It can be positive

with pip-tazo administration.
YEASTS IN THE ICU
In terms of candida, most are albicans, but there
are others. The non-albicans are much less likely
to be susceptible to the azoles [e.g. fluconazole].
Glabrata [increasing resistance] and krusei

Notes 2014
[inherently resistant] are
resistant. Parapsillosis
are skin flora seen in the
NICU.
Mucosal candidiasis of the
tongue or esophagus in
an AIDS patient is purely
a localized disease
because AIDS patients
have maintained mucosal
neutrophils. However, in
cancer patients, or
patients on high dose
steroids, this can be a
source of candidemia.
Candidemia in an AIDS
patient is usually from a
line.
Candida albicans is the
single most common
organism isolated from
the urinary tract in
patients in the ICU. Risk
factors include: advanced
age, DM, antibiotics,
catheter in place, urinary
tract abnormalities. The
clinical significance of
funguria is uncertain.
While most patients with
fungemia have
antecedent funguria, the
risk of funguria
progressing to fungemia
azoles inhibit cell wall synthesis by
reducing
ergosterol. By contrast, the
echinocandins [like caspofungin] also
inhibit cell wall synthesis, but by
inhibiting the production of 1,3 BD
glucan. Candidemia is becoming
increasingly common in the ICU and
knowing when to start prophylactic
fluconazole can be challenging. One
review suggests beginning prophylactic
fluconazole in the presence of one
major criteria [central line, broad
spectrum antibiotics] and 2 minor
was 1.3%. Both systemic fluconazole

and amphotericin B bladder irrigation
aide in clearance of funguria,
however, recurrence is exceptionally
common in both modalities.
101
Candida in the blood is not a

contaminant.
In patients with
candidemia, the patient needs a
retinal exam as treatment is a bit
different.
Candidemia can result in target lesions
within the liver, though candida may
not be seen in the blood. Candidemia
has a high mortality and seeding is
common.
Skin lesions should be biopsied. Always
remove plastic, anywhere. Candida in
a respiratory specimen is never
important. Candida pneumonia is
essentially non-existent.
The treatment has been evolving for
candida from ampho to fluconazole
and now to the echinocandins [the
fungins]. For candidemia in the ICU,
echinocandin is the therapy of choice
initially, and then fluconazole if
susceptible.
Echinocandins are very well tolerated
as is fluconazole.
Consider a patient on prophylactic
fluconazole who then develops
candidemia. Recall that the
criteria [TPN, pancreatitis, glucocorticoids, renal
replacement]. The administration of fluconazole in
this situation seems to reduce the incidence of
candida blood stream infection. However, when a
blood stream infection with candida does arise, it is
typically fluconazole resistant. There can be
significant class-resistance effects within candida,
so adding another azole is not the right answer.
There is uncommonly mixed resistance between
azoles and echinocandins, so picking caspofungin,
anidulafungin or micafungin would be a correct
response here in the patient with documented
candidemia.

Notes 2014
The other yeast you must know for the
ICU is the encapsulated one
[Cryptococcus], pulmonary disease may
not need treatment. Meningitis is
mostly seen in HIV or other
immunosuppressed states.
Cyrptococcal meningitis in patients with
HIV occurs when the CD4 count is really
low [less than 50]. The CSF should be
assayed for CrAg as it is commonly
positive [in the blood too essentially
always positive in the blood]. There is a
poor prognosis if there is AMS, high
opening pressure [above 25 cm H20],
positive blood cultures, low WBC in CSF,
high CSF Ag titers. A positive CSF
culture at 2 weeks predicts treatment
failure. The first CSF can be totally
normal in terms of cell count, etc. Do
not rule out based on a normal initial
CSF.
The treatment is amphotericin plus
flucytosine, then fluconazole. The
clinical results are not as good with
initial therapy with fluconazole. There
should be daily LP. The opening

pressure should
be lowered if more than 25 cm H20 to
less than 20 cm H20. This is especially
important in patients with symptoms.
Do not give steroids, mannitol or
acetazolamide.
102
MOLDS IN THE ICU

What are the molds you need to know?
Aspergillus, mucor and fusarium.
Aspergillus shows, septated, acutebranching hyphae. Cutaneous lesions
may be seen. There may be a crescent
sign on CT scan. This is a late finding,
and it occurs when neutrophils return.
There may rarely be CNS disease in
aspergillus.
Treatment for aspergillus has moved
from ampho to L-ampho to voriconazle
now, i.e. the therapy of choice for
invasive aspergillus is voriconazole as it
may not be more effective, but it is
certainly less toxic.
Mucormycosis or zygomycosis is nonseptated [there are many species].
Lung, skin lesions can be seen in
neutropenic patients. In diabetics,
there can be a rhinonasal form which is
horrible. The treatment is amphotericin
and maybe posaconazle [NOT
voriconazole].
What about Fusarium? If there is mold
found in the blood, it is fusarium. The
description is branching, hyaline and
septated hyphae, which sounds similar
to Aspergillus. Fortunately, they are
treated the same, that is, with
voriconazole. There is a skin lesion that
is round, elevated with central
blackened necrosis and surrounding
erythema. Amphotericin cannot cover
this mold, nor can any of the
echinocandins [including caspofungin].
This disease can be seen in neutropenic
patients and occurs from inhaling

Notes 2014
spores. They can be found
in hospital water.
ANTIMICROBIALS
severe
sepsis when patients get
appropriate
antibiotics early [CCM Kumar study]. In
fact, there is a benefit to getting an
antimicrobial in the first 30 minutes
compared to the second thirty minutes
following the onset of hypotension.
Bear in mind that different antibiotics
have different infusion times and that
some may be effectively given as a
rapid bolus. Daptomycin can be
infused in 2 minutes and merropenem
within 3-5 minutes, while pip-tazo
requires 20-30 minutes. All other drugs
require 30-60 mins. This may also be
important in a patient with a limited
number of IV ports.
GRAM POSITIVE ANTIMICROBIALS
For gram positive septic shock, in
general, the best anti-microbials are
daptomycin and vancomycin. There
may be poorer results with linezolid for
septic shock, but this data is not
terribly strong.
The anti-microbial profile between
community and healthcare associated
MRSA is blurred. It is still quoted that
PVL elaboration is rare in healthcare
MRSA, but common in CA-MRSA.
Drugs that are active against
MRSAemia are: vancomycin, linezolid,
daptomycin [should be considered in
critically ill patients], tigecycline
[considered second-line because of low
blood levels], clindamycin, doxycycline
and septra [for outpatient treatment].
Know that nasal carriage of MRSA [NOT
MSSA] is the greatest risk factor for
predicting staphylococcal bacteremia
for unclear reasons. It is a risk above
h
103
Antibacterials and antivirals will be
covered here. Remember that there is a
mortality benefit in
and beyond APACHE score, malignancy, having
been administered antibiotics and mechanical
ventilation. The risk of MRSA nasal carriage
increases with time in hospital, DM, AIDS, ESRD.
Methicillin is nephrotoxic, so it is no longer used.
Nafcillin is used [but can cause leukopenia],
oxacillin [rarely can cause hepatitis]. A serious
staph aureus infection that is MSSA is always
treated with oxacillin or nafcilln as these are

Notes 2014
preferable to vancomycin.
If the patient is
penicillin allergic you should
consider desensitization,
vancomycin may be used,
linezolid and tigecyl are not
optimal. Daptomycin might
be OK. Combination therapy
for MSSA was previously
considered useful [with
gentamicin] as it reduces
bacteremia by one day, but
there is no difference in
outcome and should not be
used for more than 3-5
days. Rifampin in
combination for MSSA does
not increase bactericidal
activity, but may be useful
with foreign bodies.
Vancomycin is a
glycopeptide antibiotic that
cannot cross the outer cell
membrane of gram
negative bugs. Vancomycin
is bactericidal against
almost all gram positives
[including pen resistant
strep] except enterococci.
Vancomycin has no activity
against gram negatives or
anaerobes.
LMWH. Consider a patient in whom the
platelet
count begins normal, but about one
week of vancomycin therapy results in
a profound drop in platelets to about
12K. Anti-platelet factor 4 Ab is
negative [making HIT rather unlikely].
Vancomycin associated
thrombocytopenia is rare, but usually
results in very, very low platelet counts
and bleeding. This is unlike type II HIT
which results in clotting. This disease is
probably Ab-mediated, so giving more
platelets usually does not help.
Stopping vancomycin usually results in
h
104
Remember that VRE is common and
VRSA is rare [but MICs are increasing
since 2000]. The dose of vancomycin is
2 grams per day in normal renal
function. If there is CNS coverage
needed, maybe 3-4 grams per day are
needed. Lung penetration, is poor. No
drug is proven superior for MRSA.
Understand red man or red neck
syndrome. Patients who receive
vancomycin in doses of more than 500
mg over a time frame of less than 30
minutes can develop this reaction. The
treatment is to infuse the drug over
one hour. It is the result of a nonimmunogenic release of histamine and
may be associated with hypotension,
especially in a patient with underlying
sepsis. In patients who have
experienced this reaction before, pretreatment with an anti-histamine may
reduce recurrence rate. Steroids have
not been shown to reduce red-man
syndrome.
Interestingly, vancomycin can be
inactivated with large doses of heparin.
Other complications of vancomycin
include phlebitis, neutropenia and
thrombocytopenia.
Recognize vancomycin-related
thrombocytopenia in someone
concomitantly on ranitidine and
an increase to normal levels in about one week, but
this can be prolonged in renal failure as vancomycin
hangs around. The use of vancomycin and
gentamicin in synergy is effective against most
staph and enterococci.
What about vancomycin levels? The trough should
be above 10 after the fourth dose, if a complicated
infection, troughs should be 15-20.
What about linezolid? Bone marrow suppression
can occur, especially thrombocytopenia after about
two weeks. With long-term use there can be
peripheral neuropathy and optic neuritis some
thought to use with vitamin B6. There is much

Notes 2014
better penetration into the lungs and
CNS as compared to vancomycin or
daptomycin.
Linezolid might be preferable to
vancomycin with MRSA pneumonia,
skin and soft tissue infection and the
need for toxin inhibition [e.g. like
clindamycin].
Daptomycin is inactivated by
surfactant and
biologically unavailable within the
lungs. Daptomycin is bactericidal
against all gram positive bugs including
VRE. It should be adjusted for renal
failure and there are cases of
eosinophilic pneumonia with
daptomycin being reported.
How long is staph bacteremia treated
for? The answer is whether its
complicated or uncomplicated.
Uncomplicated staph aureus means
that you have no foreign bodies, no
immunosuppression [DM, steroids, etc],
with a
prompt clinical response and a

negative TEE on
day 5-7. Then you can get 14 days.
Otherwise its at least one month of
therapy.
105
Moving to pneumococcus. Penicillin

resistant pneumococci is occurring [MIC
above 2]. Clinical failures are rare if
penicillin or ceftriaxone used but with
meningitis, there are failures and
failures are catastrophic, so
vancomycin should be used.
GRAM NEGATIVE ANTIMICROBIALS
What about difficult gram negative
infections? There is no good data that
combination gram negative coverage
improves outcomes. There is some
data that doing so harms the kidneys.
There are emerging bugs with
resistance to carbapenems and
quinolones, likely from excessive use
of gram negative coverage.
When do you use cefepime? It has good
activity against gram negatives
including pseudomonas [like
ceftazidime] but it has the gram
positive coverage of ceftriaxone. It has
enhanced activity against MSSA, but
not enterococcus and weak anaerobic
coverage.
Carbapenems are frequently used to
treat gram negative infections.
However, they do have drawbacks.
Enterococcus fecium has a fair
resistance to the carbapenems. None
of the carbapenems cover MRSA, they
cover E. fecalis [but not fecium]. They
all have excellent anaerobic coverage.
There is some data that merropenem is
less seizurogenic than imipenem, but
that data is a bit weak.
What about quinolones? Levaquin and
moxifloxacin are considered
respiratory quinolones because they
have better activity against

Notes 2014
pneumococcus. Cipro has
some activity against
pneumococcus, but not as
good. None have good
activity against
enterococci, moxifloxacin
threshold, which may be additive with
carbapenem. Quinolones can all cause
delirium. Prolonged QTc can be a
problem.
What about aminoglycosides? Litigation
can be a problem with these drugs.
Once daily dosing is less nephrotoxic.
Peak killing is concentration, not time
above MIC. They can be used for a
short course for bad gram negative
infections.
What about Polymyxins? They are
active against pseudomonas and
acinetobacter and other gram negative
rods. They are nephrotoxic and
neurotoxic, but are a last choice. They
are essentially first line therapy in New
York City for healthcare-associated
pneumonia from certain long-term care
facilities.
Tetracyclines and glycylcyclines.
Tigecycline has good activity against
many organisms including MRSA. It is
not good against proteus, providencia
and pseudomonas. It also has fairly poor
blood concentration so should be last
choice in sepsis or septic shock.
TREATMENT OF ANAEROBES
Anaerobes should be treated with
either flagyl or the beta-lactamlactamase inhibitors or carbapenems.
Dont be fooled by cefepime, its
anaerobic coverage is pretty poor.
Moxifloxacin also has mild anaerobic
coverage, probably not enough to
warrant use in the ICU for this purpose.
TREATMENT OF VIRAL ICU INFECTION
has some activity against anaerobes

while Cipro and Levaquin are active
against pseudomonas. They should
never be used to cover meningitis.
Quinolones lower seizure
106
The viruses in the ICU are typically limited to bad

influenza, CMV and HSV encephalitis & tracheabronchitis.
Influenza, discussed above, tends to be resistant
to amantadine and rimantadine. Zanamivir is
given by inhalation, oseltamivir is only oral.
Peramivir is IV. The drugs that end in mivir are
the neuraminidase inhibitors.

Notes 2014
Ganciclovir [first-line, but
marrow toxic] &
foscarnet [second-line, but
can cause seizure, renal
failure and low mag] are
used to treat CMV. In the
transplant population
[discussed more in section
7], CMV manifests as
pneumonia and colitis. In
AIDS, CMV manifests as
retinitis and colitis, but not
pneumonia.
HSV Encephalitis presents
with abnormal behavior,
thought and speech because
it is not just the meninges
inflamed. The drug of choice
is acyclovir without steroids.
It is, essentially, the only
treatable cause of
encephalitis once listeria is
ruled out.
The MRI shows temporal
lobe enhancement. The CSF
PCR is always positive for
HSV, but commercial labs
are not so certain. High
dose acyclovir is required
with IV fluids [needleshaped crystals in the
urine]. Zoster in the ICU is
almost always shingles, disseminated

zoster is rare today; prednisone may
reduce post-herpetic neuralgia.
107
Recognize HSV tracheobronchitis of

the critically ill and know its treatment.
Consider a patient who has profound
wheezing [has a history of COPD] that
is not resolving with standard
treatment. Additionally, there are
copious amounts of clear secretions
from the ventilator. A bronchoscopy
reveals patchy, pearl-white lesions in
the trachea and diffuse airway edema.
Additionally, there is a tad-pole shaped
cell with odd-appearing intra-nuclear
abnormalities. They have the triple M
features of being molded, marginated
and multinucleated. The nuclear are
many [multinucleated], they stick
together [molded] and the material
within the nuclei is condensed around
the insides of the nucleus
[marginated]. After two weeks of IV
acyclovir, the patients fever,
wheezing, secretions and inability to
liberate all vanished. HSV
tracheobronchitis has been reported in
not only burned, immunocompromised
patients with
malignancy, but also the
immunocompetent. HSV is commonly
found in the lower respiratory tracts
h
of patients within critical illness and ARDS
[upwards of 80% of patients], however,
one RCT did not find that treating these
patients improved their course. Only
patients with overt, symptomatic HSV
tracheobronchitis should be treated.
6. CRITICAL CARE ENDOCRINOLOGY

OVERVIEW OF HYPERGLYCEMIA & LACTIC ACIDODSIS
PHYSIOLOGICAL CONSIDERATIONS DURING

HYPERGLYCEMIA
of ketones in the serum. Only betahydroxybutyrate and acetoacetate contribute

to the anion gap and acidosis. Only
acetoacetate and acetone are measured as
serum ketones such that the serum
The body releases glucagon, and

adrenergic agents in response to
stress. Hyperosmolar states result in
altered mental status. There may be
hypotension as a consequence of
volume depletion and/or infectious
trigger. There is a powerful osmotic
diuresis in the setting of
hyperglycemia. This leads to
electrolyte loss. In the pregnant
woman there can be normo- glycemic
DKA.
The pH may be normal or high in the
setting of preceding nausea and
vomiting. There must be an increase in
anion gap and positive serum ketones
to meet the diagnosis of DKA. There is
about 5 liters, on average, of isotonic
fluid loss in DKA. In addition, there is
free water loss which can be
calculated.
The corrected sodium must be used
when calculating the free water deficit.
The corrected sodium is 1.6 higher for
every 100 the glucose is above 200.
This is essentially what the serum
sodium would be if you hypothetically,
rapidly corrected the serum glucose to
200. When this is done, the calculated
sodium is usually higher than 140 and
this gives an idea of how much free
water has been lost by the patient.
In a patient with improving anion gap,
but stable or elevated ketone
concentration in the serum, the
important thing is that the gap is being
lowered. Do not fret over a stable level
www.heartlung.org
89
JE
K

Review Notes
ketone level can stay stable as

BHB is converted to acetoacetate
and then to acetone. When
patients initially present, there is
a high titer of BHB. So the serum
ketone titer, with time, can stay
constant as acetone is generated.
The anion gap is the better
reflection of the ketoacidosis.
Serum osmolarity is calculated by twice

the sodium + glucose over 18 + BUN
over 2.8 + ethanol over 4.6. The
effective serum osmolarity should
ignore BUN because BUN moves freely
across the membranes.
Treatment of DKA involves giving a
roughly 10 unit push of insulin followed
by 0.1 U/kg/hour with a target of
lowering the blood glucose by at least
50 per hour. In elderly patients,
consider using 0.05 U/kg/hour [half the
dose for lower muscle mass]. Dont
adjust the insulin based on dropping
sugar, you give sugar at that point
which may require D10 or D50.
When calculating the anion gap,

do NOT use the corrected sodium
[unless you also correct the serum
chloride and bicarbonate which
no one does].
The boards likes to give patients
with both an anion gap metabolic
acidosis & superimposed
metabolic alkalosis such that the
patients pH is normal, but you
must recognize the elevated anion
gap and that there is an elevated
bicarbonate, usually as a result of
chloride loss [vomiting].
lung.org
The rate of insulin can be lowered

when the pH rises to 7.3. Start
potassium when the serum potassium
is in the normal range. If the potassium
is low at onset, start K+ immediately,
before insulin [i.e. if K is less than 3.3].
JE
K

Review Notes

Notes 2014
Bicarbonate should not be routinely
given; it
causes a paradoxical lowering of the
intracellular pH, there is a leftward shift
in the Hb dissociation curve [acidemia
might be OK in critical illness because it
facilitates oxygen offloading]; sodium
bicarbonate will facilitate sodium
overload and hypokalemia may worsen.
Phosphorus may be low and that should
be replete. The pH and the PvCO2 is
very similar to an arterial pH in a
patient not in shock.
The hyperosmolar state [HHS]. Glucose
tends to be higher in HHS, the
osmolarity tends to be higher, the pH is
higher, but potassium tends to be lower
because of the diuresis and the
presence of some insulin. The
treatment [compared to DKA] is more
volume, less [if any] insulin.
Sometimes all these patients need is
volume.
LACTIC ACIDOSIS
Lactic acidosis is often the result of
anaerobic glycolysis and this is type A
lactic acidosis. It is the result of tissue
hypoxia and hypoperfusion.
Improvement in sepsis-induced lactic
acidosis correlates with survival. Is
type A lactate bad? It is not the lactate
per se that is bad, it is the badness
underlying the genesis of the lactate
that is bad [e.g. asystole].
Type B lactic acidosis is when there is
no tissue hypoxia. It is seen in hepatic
disease, congenital disorders and acute
leukemia as well as many type of
drugs. Many of the HIV drugs,
methanol, cyanide, salicylate, ethanol,
beta-agonists, ethylene glycol [as an
artifact].
Does giving bicarbonate change
pressor requirements? In patients given
h
111
sodium bicarbonate resuscitation for acidosis,
there was equal increase in wedge pressure, equal
increase in cardiac output but a greater increase in
pH in those who got bicarbonate [there was no
change in vasopressor requirements] Cooper et
al. Ann Intern Med 1990; 492-498.

Notes 2014
Thyroid storm is
hyperthyroidism in the
presence of significant
cardiac or CNS
manifestations.
Apathetic affect may also be
present, particularly among
the elderly.
Thyroid storm is the most
extreme form of
thyrotoxicosis. Graves
disease is the most
common cause of
hyperthyroidism other
causes include: TMN,
thyroiditis, pituitary tumors,
thyroid cancer, struma
ovarii [ovarian tumor
making thyroid hormone],
drug-induced, HCGinduced.
What is the precipitating
factor? This must be asked
of all the endocrine
emergencies including DKA
or HHS. It is typically some
physiological stress. Thyroid
storm can be triggered by:
surgery,
pregnancy/childbirth,
trauma, or significant acute
stimulating hormone (TSH) level (less
than 0.01
U/mL) would support the diagnosis of
hyperthyroidism, as such very low
levels are not often found in euthyroid
patients who have non- thyroidal
illness. Almost all patients with overt
hyperthyroidism will have low serum
TSH concentrations secondary to
appropriate suppression by the high
serum thyroid hormone concentrations.
A very rare cause of overt
hyperthyroidism in which TSH levels
are high, is TSH-induced
hyperthyroidism, due to either a TSH-
Thyroid Storm
112
illness of any kind. As with myxedema,
the diagnosis of thyroid storm is made
clinically, with treatment undertaken in
anticipation of confirmatory laboratory
tests.
Thyroid storm may appear as
psychosis, cardiomyopathy &
cardiovascular collapse. Pretibial
myexedema is the result of the same
substance that causes exophthalmos.
In elderly patients, new onset afib is
the consequence of thyroid storm 3%
of the time.
In 99% of ICU patients, the TSH level
will be the most helpful thyroid assay.
The assays for TSH used to be very
insensitive, now they are quite good.
The tests for thyrotoxicosis can be
deceptively normal in the critically ill
for multiple mechanisms. There may
be reduced TSH production by the
pituitary which will lower T4 and T3
levels [with raised rT3 levels]. There
can be reduced thyroid binding
globulin and therefore lowered T4
levels measured. The finding of a
normal T4 level in a critically ill patient
with a high clinical suspicion of
thyrotoxicosis should not rule out the
disease. In this setting, the finding of a
very low thyroid-
secreting pituitary adenoma, or a partial resistance

to the usual feedback effect of thyroid hormones on
TSH secretion.
Treatment of thyroid storm occurs in 3 steps: Firstly,
propranolol, 60 to 80 mg q4-6h, is administered to
block the hyperadrenergic manifestations of
thyrotoxicosis. Thyroid hormones are not adrenergic
agonists, they increase receptor density. Propranolol
should be given because it is non-selective, it
crosses the blood-brain barrier, it is old and wellknown and it decreases T4 to T3 conversion. [IV
esmolol can be used instead of propranolol].
Secondly, thyroid hormone synthesis is inhibited by
administering either propylthiouracil, 200 mg q4h,
or methimazole, 20 mg q4-6h. The thyroid gland

Notes 2014
also stores thyroid hormones, so even if
synthesis is totally blocked, the gland
will continue to release stored hormone
over a period of days, weeks and
months. Thus, thirdly, iodine, either
saturated solution potassium iodide or
Lugol solution, is administered to block
thyroid hormone release from the
thyroid gland. Importantly, iodine must
only be administered after thyroid
hormone synthesis has been blocked,
in order to avoid exacerbating the
problem by enhanced thyroid hormone
production. Decreasing conversion of
T4 into T3 is accomplished through the
administration of propylthiouracil,
hydrocortisone, 100 mg q8h, and
propranolol. T3 is the biologically active
thyroid hormone.
113
Passive cooling if hyperpyrexia is
present is done
with APAP as acetylsalicylate increases
free thyroid hormone in the serum
through displacement on plasma
proteins [but this should not preclude
ASA in the setting of a thyroid- induced
MI].
Thyroidectomy may be required if a
patient develops life-threatening
agranulocytosis from propylthiouracil or
methimazole. Finally, as with
myxedema, the patient should be
evaluated and treated for the possibility
of concomitant hypoadrenalism. Iodine
therapy is discontinued and
corticosteroids may be tapered once
severe symptoms have resolved.
Patients with Graves disease should
ultimately undergo thyroid ablation
[surgically or with I-131].
An unusual manifestation of
thyrotoxicosis is thyrotoxic periodic
paralysis with hypokalemia. Consider a
young Asian man who ate a large meal
and then went to sleep. He awoke
paralyzed except for his toes and cranial
nerves.
Interestingly, he had no autonomic
symptoms [normal heart rate, blood
pressure]. The hypokalemia occurs
after the ingestion of a high
carbohydrate meal and in conjunction
with thyrotoxicosis it results in profound
hypokalemia and hypophosphatemia.
Importantly, the hypokalemia is the
result of intra-cellular shift, so
replacement neednt be aggressive.
Once the thyrotoxicosis resolves the
paralysis never returns.
Myxedema Coma
Myexedema coma is the severe form

of hypothyroidism. The most common
cause is auto-immune [burned out
thyroiditis] followed by iatrogenic,

Notes 2014
drug-induced and lastly
iodine deficiency.
These patients are very
slow with course facial
A pericardial effusion may be present,
although
significant cardiac compromise is
uncommon.
Triggers include narcotics, sedative
hypnotics [sometimes only one dose]
as well as other physiological stressors
such as myocardial infarction or
infection and septic shock [masked by
profound hypothyroidism]. Myxedema
coma is often the result of prolonged
noncompliance with thyroid
supplementation in the face of absent
thyroid function, such as following I131 ablation. Drugs that can cause
underlying hypothyroidism include
amiodarone, propylthiouracil, lithium,
and sulfonamides.
When the TSH is very low, patients
do not get myxedema coma as the
thyroid gland can still secrete small
amounts of thyroid hormone.
Therefore, myxedema coma does not
really occur in secondary or tertiary
hypothyroidism, only primary
hypothyroidism. Thus, TSH should be
quite high in the patient with
myxedema coma.
Myxedema coma patients tend to have
a low PaO2, high PaCO2 [from blunted
respiratory responses, they are often
intubated], hyponatremia,
hypoglycemia, elevated CPK, high TSH
with low total and free T4 and low T3.
Also, hyponatremia, a normocytic
normochromic anemia, hyperlipidemia.
Hyponatremia is due to an impairment
in free water excretion and can result in
seizure activity. Hypoglycemia can
occur from hypothyroidism alone or
may be due to concomitant adrenal
insufficiency.
features, periorbital edema and a

very delayed response to questions
asked. They answer in a course, froglike voice.
114
The treatment of myxedema coma should begin

based on clinical suspicion and should not wait for
laboratory confirmation. These patients need to be
in the ICU for respiratory support with passive [not
active] rewarming to prevent excessive drop in
blood pressure. The patients sometimes require
glucose infusions for hypoglycemia, large doses of
IV synthroid and hydrocortisone to avoid Addisonian
crisis.
The treatment of myxedema coma is IV thyroxine
- loading dose of 300 mcg of thyroxine then daily

Notes 2014
administration of doses
ranging from 50 to 100
mcg. As unsuspected
adrenal insufficiency is
frequently also common, all
patients with myxedema
coma should be empirically
treated for possible adrenal
insufficiency with daily
administration of 300 mg of
hydrocortisone.
hypothyroidism. Hence, no thyroid

supplementation is required.
115
The time-course in the critically ill is

that free T3 level goes low quickly and
rT3 rises. There is no role for testing
T3 in the ICU because it will always be
low. If it is not low in the ICU, then
either the test is wrong or the patient
should be on the floor. T4 drops later
in critical illness.
Reverse T3 has no known biological
activity.
Sick Euthyroid
There may be an adaptive

response during critical
illness whereby the body
seeks to conserve energy
by down-regulating TSH
production.
Further, metabolism is
slowed by increased
conversion of T3 to reverse
T3. T4 levels may also be
low, particularly in the
setting of protracted critical
illness, and the TSH level
can also vary, being either
slightly elevated or
decreased. Free T4 levels
are normal, indicating the
absence of clinical
Clinicians should recognize that fingerstick
capillary blood glucose measurements
can be inaccurate in critically ill
patients. Finger-stick glucose
measurements are lower than true
glucose measured from venous blood
in hypotensive patients but at other
times may be found to be higher than
venous blood. Serum electrolytes
should be assessed q2-4h in DKA.
Once intravascular volume has been
restored and the patients glucose has
lowered to the 200 range, glucose and
DKA
Serum glucose level is usually below

800 mg/dL in DKA, whereas in HHS a
glucose level in excess of 1,000 mg/dL
is not uncommon.
Leukocytosis and altered sensorium are
proportionate to the degree of acidemia
and can confuse the clinical picture
regarding infection, especially CNS
infection.
Regular insulin is administered as an IV
bolus of
0.10 to 0.15 U/kg/h, followed by a
continuous IV infusion at 0.10 U/kg/h.
Blood glucose should be lowered by
about 50 mg/dL/h and assessed hourly,
with downward adjustments made in
the insulin drip as blood glucose lowers.
hypotonic saline solution, in the form of dextrose in
0.45 NaCl, should be administered until the DKA has
resolved. This serves to avoid hypoglycemia in the
context of ongoing DKA - this permits the continued
administration of IV insulin to reverse ketogenesis
and replete free water deficit.
Despite initial hyperkalemia, with the
administration of insulin, hypokalemia develops and
should be treated with IV potassium
supplementation. Usually 20 to 30 mEq/L is added
to 0.45 saline solution, as the addition of potassium
to normal saline solution would result in the
administration of hypertonic fluids.

Notes 2014
Hyperosmolar Non-Ketotic Dehydration Syndrome

116
Hypophosphatemia often develops

during treatment of DKA, but it seldom
requires supplementation, which should
be administered only if clinically
significant or severe [less than 1.0 mg/
dL].
After the normalization of the anion gap

has occurred, the patient should receive
subcutaneous regular insulin. The
administration of IV dextrose is
stopped, and IV insulin is discontinued
30 min later. These changes are best
made once the patient has resumed
oral nutrition, otherwise ketogenesis
may resume.
Cerebral edema can occur as a
complication of DKA treatment in
patients under 20 years of age, but the
risk is mitigated if rapid correction of
sodium and water deficits are avoided
and glucose is added to IV fluids once
serum glucose level has dropped to the
low 200 range.
The severity of hyperglycemia is often

quite significant [>1,000 mg/ dL]. The
resultant hyper- osmolality produces
depression of the CNS, which, when
severe, can cause coma. An anion gap
metabolic acidosis from ketogenesis is
typically not present.
Serum sodium is often low in HHS due
to osmotic shifting of water from the
intracellular compartment. As serum
glucose levels tend to be higher in HHS
than in DKA, this effect can be quite
profound. In addition, just as in DKA,
plasma volume is contracted at the
same time, owing to osmotic diuresis
from glucosuria. If, however, the
glucosuria effect predominates
[especially if PO water is restricted],
hypernatremia may be observed. In
either circumstance, the serum sodium
level is altered by hyperglycemia. A
correction factor to determine the
estimated serum sodium when the
glucose level is 200 is:
Na corrected = Na measured x [0.016
(Glucose in mg/dL - 100)]
The corrected sodium is used to
determine free water deficit, which can
serve as a guide to the amount of
volume resuscitation required. The
corrected sodium should not be used as
the marker for sodium correction rate.
Rate of correction should still follow the
measured serum sodium as that is the
sodium level which determines osmotic
water shifts.
The treatment of HHS involves the same
management principles as DKA:
vigorous volume replacement and
sometimes an IV insulin drip.
The amount of normal saline solution
required to restore extracellular fluid
tends to be greater in HHS than in DKA.
Replenishing intra-vascular volume is

Notes 2014
sometimes all that is
needed as it facilitates
renal perfusion, glycosuria
and diminishes the stress response

which maintains hyperglycemia.
117

Notes 2014
Glucose Control in the ICU

Insulin and tight glucose control in
surgical patients improved mortality in
one trial if BG was kept between 80 and
110. Several additional trials showed
worsening mortality in strict glucose
control [from hypoglycemia]. This is
best shown in the NICE-SUGAR trial.
The control group was 150, the
intensive group was at 107. There was
a worsening outcome in those on strict
control; there was an increase in 90-day
mortality in patients treated with tight
glycemic control, [80-100 md/dL]
compared with a less aggressive
approach. Only the subset of patients
with trauma or those being treated with
corticosteroids demonstrated a trend
toward benefit with tight control.
However, it should be recognized that
the control group in NICE- SUGAR had a
mean glucose level around 140 mg/dL.
Thus targeting 150 mg/dL in the ICU
seems warranted.
Hypoglycemia
Hypoglycemia can be caused by drugs,

ethanol, sepsis, hepatic failure, renal
failure, etc.
Hypoglycemia can be a result of the
cessation of TPN [for a multitude of
reasons]. These patients have insulin
on-board and have an abrupt drop in
blood glucose. Note that the adrenergic
symptoms of hypoglycemia can occur
not just based on the absolute value of
glucose, but on the rate of decline. If a
patient does not have an IV for
emergent D50, the patient can be
given IM glucagon. Once treated, the
underlying cause of the hypoglycemia
must be sought or else the low glucose
will reappear.
Blood glucose should be monitored
hourly. A second ampule may be
required within an hour of treatment.
9
h
Patients should also receive a dextrose drip of

either 5% or 10% solution, at a rate appropriate to
the clinical circumstances encountered. Glucagon,
hydrocortisone, or octreotide can be administered if
hypoglycemia is

Notes 2014
profound and refractory to
the above measures,
but it is seldom required.
Adrenal Crisis
Adrenal insufficiency may

be either primary
[Addisons syndrome], that
is, due to insufficient
production of
glucocorticoids and
mineralocorticoids; or
secondary, from
underproduction of ACTH.
In Addisons day, the most
common cause of adrenal
insufficiency was TB.
Today it is idiopathic,
mostly autoimmune.
Altogether, the most
common cause is the
abrupt discontinuation of
corticosteroids.
Causes of primary adrenal
insufficiency include
autoimmune, that is,
Addisons disease; bilateral
adrenal hemorrhage;
abrupt withdrawal of
exogenously administered
corticosteroids, TB; septic shock;
meningococcemia [Waterhouse
Friderichsen syndrome]; metastatic
malignancy; amyloidosis; and drugs
such as etomidate and ketoconazole.
Causes of secondary adrenal

insufficiency include pituitary tumors;
craniopharyngioma; as a postoperative
complication; postpartum hypopituitarism [Sheehans syndrome];
infiltrative diseases such as
hemochromatosis, sarcoidosis,
histiocytosis, or histoplasmosis; TB.
Like thyroid storm, adrenal crisis is
often triggered by physiologic stress
such as trauma, surgery, or acute
medical illness.
Abdominal, flank, lower back, or chest
pain are common in patients with
bilateral adrenal hemorrhage or
infarction, the main risk factors for
which are anticoagulation and
postoperative state.
Hyperpigmentation may be seen in
patients with primary adrenal
insufficiency, but this is uncommon.
Patients with secondary adrenal
insufficiency lack hyperpigmentation,

Notes 2014
dehydration, and hyperkalemia.
Hypotension is
less prominent, whereas hypoglycemia
is more common than in primary
adrenal insufficiency.
In summary: in primary hypoadrenalism
you tend to see the textbook things
hyperpigmentation, hyperkalemia and
hypotension whereas in secondary
hypoadrenalism you tend to see
normal skin tone, normal potassium but
more hypoglycemia.
Again, in terms of electrolytes, the
common findings in adrenal
insufficiency are: hyponatremia,
hyperkalemia, hypoglycemia,
azotemia, hypercalcemia, acidosis,
anemia, neutropenia and eosinophilia.
By the time these patients are seen in
the unit, these findings are uncommon.
The urea from adrenal insufficiency is
from protein catabolism.
In individuals who are not stressed, a
total cortisol level of more than 15 g/dL
is sufficient to rule out adrenal
insufficiency. A level less than 5 mg/dL
constitutes absolute adrenal
insufficiency with 100% specificity but
low sensitivity (36%). A cut-off level of
10 mg/dL is 62% sensitive but only 77%
specific.
The appropriate response of the
adrenal glands in the setting of critical
illness is unknown.
Some authors suggest that a level less
than 25 mg/dL may be insufficient in
critical illness such as sepsis.
Cortisol is protein bound, and total
cortisol levels bear a variable
relationship to free cortisol levels.
Patients who are hypoproteinemic may
have a normal free cortisol level
despite a seemingly insufficient total
cortisol level [remember, it is the free
cortisol which is biologically active].
h
120
In patients who are not septic, a rapid cosyntropin
test can be performed. This is a cortisol level 30
and 60 minutes following the cosyntropin. With
stimulation, the plasma levels go well above 20 in
normal patients. Thus a rise

Notes 2014
in total cortisol level less
than 9 mg/dL or an
absolute level less than
20 mg/dL may be
indicative of RAI.
In the CORTICUS trial
[septic patients], the
cosyntropin stimulation test
was found to be unreliable
when correlated with free
cortisol levels. In addition a
mortality benefit was not
observed in the
corticosteroid group.
Patients receiving
corticosteroids were able to
be weaned off vasopressor
medications 2 days sooner
than the placebo group but
also were found to have a
threefold risk of
subsequent sepsis while in
the ICU.
The standard dose in sepsis
is hydrocortisone, 50 mg IV
q6h for 5 days.
Adrenal crisis is treated
with an initial dose of 200
mg of IV hydrocortisone
followed by 100 mg q6h. IV
spinal injury or Guillain-Barre
syndrome]; the use
of sympathomimetic drugs such as
cocaine, phencyclidine, or
amphetamines; and the ingestion of
tyramine-containing foods in patients
taking monoamine oxidase inhibitors.
Diagnosis is made via plasma levels of
metaneph- rine and normetanephrine
or 24-h urine levels of metanephrines
and catecholamines when the patient
is stable and not critically ill, as the
stress of critical illness can produce
misleading values that may be false
positives. Plasma normetanephrine has
the best distinguishing characteristics.
121
administration of normal saline solution
is important to correct volume
contraction. Hypo- tonic fluids should
not be administered, as they can
worsen hyponatremia.
Mineralocorticoids are not needed when
large doses of hydrocortisone are given.
When this dose gets to less than 100
mg per day. If a patient is completely
adrenally insufficient then
fludrocortisone should be given when
HC doses fall below 100 mg per day.
Pheochromocytoma
A pheochromocytoma is a
catecholamine- secreting tumor of
chromaffin cells; most common in the
adrenal glands, though it may occur
elsewhere in the body. 10% are
malignant, 10% are familial, 10% are
bilateral 10% are multiple and 10% are
extra-medullary.
The classic triad of episodic headache,
sweating, and tachycardia is seldom
present. Episodes dont last more than
a few hours at time [usually 20 minutes
or so].
Mimics of pheochromocytoma in the
ICU are many, including autonomic
dysfunction [e.g.
The administration of tricyclic antidepressants can
also result in falsely elevated results.
Subsequent to a chemical diagnosis, imaging such
as CT scan or Iodine 123-metaiodobenzyl
guanidine scan localize the tumor. MRI can light up
chromaffin tissue quite well.
Patients with known pheochromocytoma for
elective surgery should receive preoperative
management with an alpha-agent, such as
phenoxybenzamine.
BB administration is contraindicated unless prior
alpha- blockade has been accomplished. The CCB
nicardipine can be a useful adjunct. Metyrosine, an
inhibitor of catecholamine synthesis, may also be
used.

Notes 2014
As with the other endocrinopathies,
stress can precipitate a hypertensive
crisis in pheo patients. Undiagnosed
pheochromocytoma patients
presenting with post-operative
hypertensive crisis have a high
mortality these patients are treated
with phentolamine intravenously 2 to 5
mg every 5 min until the target BP is
achieved.
Diabetes Insipidus
When water adsorption by the

collecting tubules of the kidney is
impaired, either from a 1. lack of the
antidiuretic hormone [ADH] also known
as arginine vasopressin [AVP] [ddAVP
or
desmopressin is the synthetic

analogue] or due to
2. the lack of responsiveness of the
collecting tubules. 1. & 2. are central
and nephrogenic DI, respectively.
122
Patients with complete central DI will

have a very dilute urine and a low
urine osmolality. Patients in the
ambulatory setting do not have
hypernatremia because they drink a
lot, they are always drinking water!
Once in the hospital, this intake can be
impaired. These patients will have
large amounts of dilute urine. Urine
osmolality should be less than the
plasma osmolality.
Clinically, this does not distinguish
nephrogenic from neurogenic DI. The
use of a V2 [selective] agonist [ddAVP
desmopressin] can be done to
distinguish central from nephrogenic
DI; further, V2-selectivity does not
cause vasoconstriction [as AVP or
vasopressin does].
Administration of 1 mcg desmopressin
SQ is done. If the urine osmolality
increases by 10- 50%, then DI is partial
central, if it increases more than 50%,
then its complete central DI. If there is
no change, then this is nephrogenic DI.
One unusual form of central DI is
gestational diabetes insipidus.
Consider a patient who presents in her
third trimester with acute cholecystitis
and over a 24 hour period urinates
large amounts of free water and the
serum sodium shoots up 20 points. The
onset of gestational diabetes insipidus
can occur in the third trimester and is
typically mild and self- limited. The
placenta elaborates vasopressinase an
enzyme that degrades vasopressin,
but normally maternal vasopressin
levels increase in response. In this
patient, she was made NPO for surgery
and was unable to regulate her serum

Notes 2014
sodium. Importantly, the
use of vasopressin is
typically less effective
because of the innate
vasopressinase elaborated
123
by the uterus. The synthetic
desmopressin, however, is resistant to
vasopressinase and therefore is the
correct response in addition to giving
free water. Thus
cause of malnutrition in the ICU. What is the

difference between starvation and stress hypermetabolism?

this disease will behave in a manner
similar to central DI.
Nephrogenic DI can be caused by
several drugs, including lithium,
demeclocycline, amphotericin B, and
antiretroviral drugs such as tenofovir
and indinavir. Hence, ADH levels are
elevated in nephrogenic DI but are
diminished or absent in central DI.
Treatment of central DI entails
correcting the free water deficit as well
as prevention of ongoing polyuria
through the administration of
desmopressin 1 or 2 mcg
subcutaneously q12h. Without this,
there can be 20-25 liters of water lost in
24 hours. This can very, very rapidly
result in hypernatremia and shock.
Patients who are hypotensive due to
hypovolemia should receive normal
saline until intravascular volume has
been replenished. Otherwise, hypotonic
fluids may be administered.
Management of nephrogenic DI
desmopressin is not administered. The
discontinuation of any nephrogenic DI
drugs is an important. A thiazide
diuretic leads to mild extracellular fluid
volume depletion, and increased water
reabsorption at the PCT. There is less
water delivered to the DCT and less
urine is produced.
Summary - key points with endocrine
emergencies: the TSH is the best test
of thyroid function in the ICU, the
cosyntropin stim test is the best test of
adrenal function, isotonic fluids are
always given for hypovolemia, closely
titrate IV fluids in DI and use
phentolamine to right high blood
pressure in pheochromocytoma.
NUTRITION IN THE ICU
Altered metabolism that results from a

disease process is the most common
www.heartlung.org
JE
K

Review Notes
STARVATION
Starvation occurs when

nutrient supply cannot
meet nutrient demand.
The body responds to
preserve lean body mass
and this response is
usually carried out by
decreasing energy
expenditure, use of
alternative fuel sources
and reduced protein
wasting. Glycogen is gone
in about 24 hours, glucose
creation goes on for a few
more days by breaking
down protein from amino
acids. Then fatty acids,
ketones and glycerol
become the primary fuel
sources in all but obligate
glucose-utilizing tissues
[brain and RBCs]. In the
fully adapted starved
state there is decreased
protein catabolism and
ureagensis [as compared
to the fed state]. The
starved animal must
preserve muscle mass so
as to obtain a meal in the
near future.
lung.org
CRITICAL ILLNESS HYPER-METABOLISM

This is different from stressed hyper- h
metabolism
this occurs at the expense of the
lean body mass. It is a general
response where energy and substrate
are mobilized to support inflammation
and the immune system. It is often
associated with poor tissue perfusion.
It is seen with high lactate, high
urinary nitrogen excretion. The RQ is
increased compared to the starved
state [see below]. There is a decrease
in insulin-mediated glucose uptake and
increased gluconeogenesis [stresshyperglycemia]. There is more glucose
oxidation [increased non-insulin
mediated glucose uptake]. There is
increased Cori cycle to convert lactate
to glucose.
There is increased total protein
anabolism and catabolism [net increase
in catabolism] with increased muscle
release of amino acids. There is a rapid
drop in lean body mass with less
albumin synthesized and increased
urinary nitrogen losses with increased
ureagenesis. The amino acids that are
released are used as fuel and for
immune system function at wound
sites. They serve and gluconeogenic
substrates and are used in the liver to
create acute phase reactant proteins.
In
JE
K

Review Notes

Notes 2014
other words, the skeletal muscle is
degraded to
become the substrate for processes of
inflammation and tissue repair.
Insulin is very important not only for
glycogen storage but also fat and
protein synthesis and storage. Cortisol,
glucagon and catecholamines are all
important for breaking down these
biomolecules. Injured tissue hormones
are often the cause of the stress-hypermetabolic response aforementioned
[e.g.TNF, IL-1, IL-2, etc.].
STARVATION VERSUS HYPER-METABOLISM
The key differences between starvation
and hyper-metabolism are the
following: energy use is increased in
the stress response and decreased in
the starvation response. The
respiratory quotient, similarly is low
and high, respectively while the
primary fuel is fat and mixed
respectively.
Hyper-metabolism results in much more
synthesis of glucose and a large
breakdown of protein [though there is
some protein synthesis more than
starvation]. Ureagensis is high in hypermetabolism. By contrast, ketone
formation and response to feeding in
starvation is very high. If you feed a
starved patient, the starvation response
will go away. This is not true in hypermetabolism, where the response will
resolve when the underlying problem is
treated.
ROUTE AND TIMING OF NUTRITION IN THE ICU
Nutritional support is important for
malnourished patients and in those in
whom it is likely to occur [e.g. severe
injuries]. Start nutrition support as soon
as need is recognized, but always within
7 days. Early TPN, however, is bad.
h
126
It is important to minimize starvation effects,
prevent specific deficiencies, and treat the
underlying disease.
Randomized trials have not shown a benefit from
approaches that try to counteract the natural
catabolic state that occurs in the acute phase of

Notes 2014
critical illness and some
experts suspect
catabolism during acute
critical illness is in fact
adaptive, or at least not
harmful in and of itself.
Providing enteral nutrition
early to people with critical
illness may reduce their
infection risk, as compared
to delaying enteral nutrition
or not providing any. Enteral
nutrition should be provided
within 48 hours to people
with critical illness who are
not at high risk for bowel
ischemia. This conclusion
comes from a metaanalysis of 15 randomized
trials showing an
approximately 50% reduced
risk for infection among
those receiving early
enteral nutrition. Another
meta-analysis suggested
early enteral nutrition
reduced mortality risk by
50% as well, but fell just
short of statistical
significance.
However, these data are
fairly old and from
heterogeneous trial designs;
better than, early TPN for critically ill.
That being
said, early total parenteral nutrition has
never been shown to increase mortality
from critical illness it increases the
nosocomial infection risk by 4-5%.
What are the other early risks of TPN?
Arrhythmia, hemolysis, rhabdomyolysis
are all acute complications of TPN.
Metastatic calcification is not an acute
complication. The reason for many of
these risks is re-feeding or the rapid
uptake of intra-cellular electrolytes such
as potassium, phosphate and
127
some have argued these findings might
be due to publication bias or other
biases. Also, most of the included
patients in these randomized trials
were surgical [burns, trauma, etc.] and
not medical patients [who have mainly
been studied in observational trials].
As stated above, caloric goals are
theoretical and controversial, and were
developed without outcomes-based
evidence. In patients who were
previously adequately nourished,
providing minimal calories [trophic
feedings] enterally for up to 7 days led
to equivalent outcomes as compared
with more aggressive feeding, in
mechanically ventilated critically ill
patients. A 2011 randomized trial
suggested feeding critically ill patients
below caloric goals might improve
survival.
Many critically ill patients have reduced
gut motility and fail to tolerate enteral
feedings in the amounts calculated to
meet their theoretical caloric needs. For
these patients, there appears to be no
benefit to starting total parenteral
nutrition in the first week after impaired
gut motility occurs, and doing so may
increase the risk for nosocomial
infection. Providing no nutritional
support or dextrose infusions are as
good as, or
magnesium. The provision of glucose leads to
profound uptake of these electrolytes with depletion
in serum levels. Acute depletion of the serum levels
is known as the re-feeding syndrome. These drops
result in the rhabdomyolysis, etc.
During recovery from critical illness, the body
rebuilds muscle and fat (anabolism) and replenishes
other energy stores (fat and glycogen). Nutritional
supplementation should often continue after acute
critical illness to support this process.
DOSE AND CONTENT OF NUTRITION IN THE ICU

Notes 2014
How many calories to provide? The
Harris- Benedict formula can be used to
calculate BEE with a stress factor of at
least 1.2 [just to be a person], severe
illness is a factor of 2.0 [and everything
in between].
Excess
calories
can
cause
hyperglycemia, excess CO2, and
lipogenesis. The basic rule of thumb is
25-30
kcal/kg/day
total
[5
gram/kg/day] total.
How much of this is glucose? It should
be about 60-70% of the calories to be
given or 20 kcal/kg/day. What about
fat? It should be about 15-40% of
calories. Limit fat to 1 gram/kg/day.
Protein catabolism is not suppressed by
the provision of adequate calories,
protein or amino acids. Therefore to
attain nitrogen balance, there should be
provision of protein synthesis with 1.22 gram/kg/day [which is higher than
normal daily
128
protein requirements]. A negative nitrogen
balance is common in critical illness
despite what you give. Rule of thumb
for fat-protein- carbohydrate is 1-2-3.
That is 1g/kg/day of fat, 2 g/kg/day
protein and 3g/kg/day of carbohydrates
in the critically ill patient.
There is not a lot of evidence in terms
of vitamins and minerals in the ICU.
There is some data that increased zinc
and vitamin C may help wound healing.
When you give TPN, vitamin K is not a
part of the bag. It needs to be given
vitamin K once per week in this
situation as it is not light stable.
The types of enteral formulae are
many. Intact formulae contain intact
protein without gluten; they are
lactose free. They are isosmotic with
low residue and fiber. The hydrolyzed
formulae provide protein as peptides
or amino acids, they are low in fat.
Elemental formulae provide protein as
crystalline amino acids and
carbohydrates as mono or
disaccharides. This classification is
based on protein form.
Another means of breaking down
formulae is based on their clinical use.
There are standard formulae for
starved patients who cannot eat. High
protein formulae are for the hypermetabolic, critically ill patients and
they contain 45-60 grams of protein
per 1000 kcal. Calorie dense formulae
contain 2 kcal per mL and they are for
fluid restricted patients and are
relatively low in protein.
There are also organ specific formulae
and immunity enhancing formulae
which are designed to alter immune
function and reduce the inflammatory
response. The organ specific formulae
include those for pulmonary failure
which are used for acute on chronic
respiratory failure. There is 50% less fat

Notes 2014
for less carbon dioxide
production. Probably
avoiding overfeeding is
more important. Hepatic
129
failure formulae contain high levels of
branched chain and low aromatic amino
acids in an effort to

reduce encephalopathy, but probably
dont work all that well. Renal failure
formulae are developed for patients
who are not being dialyzed. Once on
dialysis, they should be on a standard
high protein diet. Renal failure
formulae are low in protein, potassium
and phosphorus. None are of proven
benefit.
was studied in ARDS patients in 1999 and found to

significantly improve outcomes and then this was
repeated again in 2006 and shown to improve
outcomes, but in 2009 the largest trial [an ARDSnet
trial] OMEGA trial casted doubt.
CALORIMETRY & SUMMARY
When a patient is being weaned from

mechanical ventilation who has a high
RQ, and high daily calories should first
have his total calories reduced, before
modifying the content of the feeds.
Protein and carbohydrate swapping will
not significantly change CO2 load to the
lungs.
What about the immunity enhancing
enteral formulae? Arginine is a
nonessential amino acid that is a nitric
oxide precursor and a non-specific
immune stimulant with enhanced
wound healing. Glutamine is
conditionally essential and is fuel for
enterocytes, lymphocytes and
macrophages & reportedly improves
gut barrier. Glutamine has been looked
at in a meta-analysis in 2002 that
showed decreased infectious
complications and mortality in patients
receiving TPN [surgical patients]. Then
a follow up RCT showed that glutamine
supplementation is harmful and should
not be done.
Omega-3 PUFAs are metabolized to
less inflammatory leukotrienes than
the omega-6 PUFAs. These formulae
have been tried to alter the
inflammatory responses of critical
illness.
There was a meta-analysis of 22
randomized trials of 2500 patients and
there was no difference in mortality but
there was a reduction in infectious
complications. Then 4 years later there
was an RCT of 600 ICU patients and
found no difference in anything. Oxepa
www.heartJE
lung.org
K

Review Notes
RQ of 1.0 means that the fuel source is

carbohydrate and this occurs when
excessive carbs are used and when the
RQ is more than 1.0, the fuel source is
fat synthesis and this occurs in overfeeding. Critically ill patients should
have an RQ of 0.8 to 0.95 based off of h
a mixed fuel source.
Indirect calorimetry is not often used,
1 0 0
but
may b
In summary, stress-hypermetabolism
from starvation is an important
physiological distinction. Critically ill
patients require more energy but are
less able to tolerate glucose and require
fat to meet energy requirements.
Critically ill patients require more
protein to achieve nitrogen balance.
Enteral nutrition is preferred.
on the boards. An RQ of
0.6-0.7 means that fat is
the fuel source and occurs
during starvation. 0.80.95 means that the fuel
source is mixed and this is
the normal condition. An
lung.org
JE
K

Review Notes
101
STEM CELL TRANSPLANT
The preference is to use conventional MA

7. CRITICAL
CARE
HEMATOLOGY
stem cell transplant [SCT], unless the
A
&A
S
C
UTOLOGOUS
TRANSPLANTS
LLOGENEIC
TEM
ELL
patient is old or has medical co-morbidities

there may be a higher risk of recurrence in
reduced-intensity [NMA] transplants. In the
late 80s-90s very few over the age of 50 got
transplants, now many
The general approach to an autologous

stem cell transplant is cytoreduce the
malignancy, wipe out marrow, restore
marrow by peripheral blood stem cells
or cryopreserved stem cells.
Autologous transplants are first line
therapy for multiple myeloma as it
prolongs progression free survival [PFS]
and survival and is second line for
relapsed NHL and Hodgkins disease.
Mortality rates are less than 5% now.
Why they die post auto-transplant
mostly relapse of disease, very little
procedural complication [total body
radiation, infection & organ toxicity].
Allogeneic transplants, by contrast,
procure stem cells from a donor, or
allogeneic cord blood.
Allogeneic transplant gives the patient
a whole new immune system and this is
the basis of the graft-versus-tumor and
graft-versus-leukemia effects. The
patient is cytoreduced and
hematopoietic function is returned from
the donor. Allogeneic transplants are
conventionally via myeloablative [MA]
conditioning i.e. eradication of
malignant cells first. Now there are
non-myeloablative [NMA] or minitransplants based on the concept that
graft-versus-leukemia may be all that is
needed to destroy the residual
malignancy. NMA transplants are
designed for patients over 55; reducing
conditioning regimen via minitransplants could decrease the toxicity
of treatment and potentially cure them.
MA is still more frequent, but NMA is
increased [~ 20%].
www.heartlung.org
JE
K

Review Notes
more are, even 10% or so are

above the age of
60. Of all patients going through
allogeneic transplant about 25%
will have an HLA identical donor
[sibling]. Now with the increase in
unrelated donor registry there is
about a 50% chance of finding an
HLA identical unrelated donor. Up
to 60% of Caucasians will have a
fully HLA matched unrelated
donor as Caucasians are less HLA
diverse.
Cord-blood is a rich source of

hematopoietic progenitor cells. They
can be frozen for later use. The cordblood cells are very nave, so they can
be mismatched for 1/6 or 2/6 HLA loci
with less GVHD MHC mismatching. The
rates of GVHD are comparable to
identical siblings and chronic GVHD
may be less likely!
Matched unrelated donors versus cord
blood have superior outcome for
survival at 24 months by about 5-10%
[still total survival around 50% for
leukemia]. Cord blood has significant
morbidity and mortality during
transplant for two big reasons, the first
is that engraftment is delayed about 25
30 days of PMN recovery [normally
about 10 days] so opportunistic
infections are really common in the first
30 days with cord blood transplants;
secondly, cellular immunity in these
patients does not exist in 100 days. So
if CMV reactivates, it may take 3-4
months to have an adaptive immune
response, CMV
The risk of graft versus host

disease is much higher using
matched unrelated donors, but
this has decreased given the use
of high resolution HLA typing, to
levels similar to matched siblings.
UNRELATED CORD BLOOD STEM CELL
TRANSPLANTS
What about the 25-30% who dont
have matches? These patients
could be suitable for unrelated
cord-blood transplantation.
lung.org
JE
K

Review Notes

Notes 2014
pneumonitis, colitis, adenovirus
reactivation, if
EBV reactivates there can be lymphoproliferative disorders; all together
there is a 44% mortality in the first 100
days and about 60% in the first 180
days. There is a lot of work being done
now to develop transplants dual cord
transplants to shorten the
neutropenia phase, with the addition of
viral reactive T cells to restore cellular
immunity.
ALLOGENEIC SCT OUTCOMES
There were 30,000 allogeneic
transplants performed historically and
outcome over two decades comparing
90s to 2000s - there was a significant
reduction in all infections, less liver
toxicity, less veno-occlusive disease,
there is a reduction in GVHD, reduction
in 200 day transplant-related mortality
about a 40% reduction in mortality.
Relapse is much less common in
allogeneic transplants, but GVHD is a
big cause of death following allogeneic
transplant. Infection is another big
cause of death.
SCT DRUG TOXICITIES & INFECTIOUS
OVERVIEW
What kinds of toxicities can occur in terms

drugs used? Calcineurin inhibitors are
used for GVHD prevention and they
cause renal insufficiency, HTN,
hypomagnesemia, gingival
hyperplasia, PRES with seizures. PCP
PPx everyone is on Bactrim, but it
can cause marrow suppression and
hyperkalemia. For CMV reactivation
gancyclovir and valgancyclovir can
cause neutropenia especially after one
week patient can then go to
foscarnet, but this causes renal
insufficiency. CMV pneumonitis which
drug to use? If bone marrow is
h
134
in bone marrow transplant about 18 days, cord
blood 25-30 days maybe longer. This is
when fungal and bacterial infections
occur.
Following neutrophil recovery, the
problem is compromised cellular
immunity usually a consequence of
CSA, which is kept on for at least 6
months and then tapered. About 2 of 3
patients will have their CSAs
completely tapered off as the new
immune system takes hold and
recognizes the host after about 1 year.
RSV is bad December to early April.
RSV is a big problem in transplant
RSV PNA has a 60-70% mortality rate,
there is no effective therapy other than
taper immune suppression. PCP
incidence is common 6 months to one
year. Chronic GVHD can occur 100
days to 2 years. Risk factors for
invasive fungal infections is problems
with PMNs
delayed engraftment [e.g. cord
blood], acute and chronic GVHD [e.g.
on steroids], secondary neutropenia
post-transplant [e.g. drug toxicity like
gancyclovir]. However, death from
invasive fungal infections from acute
GVHD has dropped with new antifungal drugs.
CMV
of
suppressed use foscarnet, if kidneys are injured,
use gancyclovir.
GVHD treatment steroids, cyclosporine [CSA],
sirolimus, monoclonal Abs all of which can cause
opportunistic infections.
Infectious complications conventional peripheral
SCT neutropenia for about 10 days,

Notes 2014
CMV reactivation occurs in
30-50% of SCT patients
defined as detection of
CMV in blood by PCR or
antigenemia without
evidence of CMV disease.
While CMV reactivation
can occur, if there is no
reactivation by 100 days,
CMV is unlikely to
reactivate; usually these
patients are screened with
serum CMV PCRs and
treated preemptively.
pre-transplant. About 20% of the adult
population are CMV sero-negative.
Those not at risk for CMV are when the
patient and donor are both
seronegative for CMV pretransplant.
CMV disease is almost always
reactivation of CMV virus dormant in
recipient cells not de novo infection.
Thus, if the patient is seronegative pretransplant, reactivation is highly
unlikely.
If the donor has T cells that can
recognize CMV, they are more likely to
prevent reactivation in the seropositive
recipient. If the recipient is seropositive,
and the donor is CMV sero-negative,
there is more likely to be reactivation
and chronic CMV as the new immune
system is CMV nave.
CMV reactivation comes 14-100 days
post- transplant. Those getting T
depleted transplants, in patients getting
suppressed from drugs for acute GVHD,
or in patients who received a CMV
negative donor transplant the risk for
reactivation is higher.
The highest risk of CMV reactivation
comes when the patient is positive, but
donor negative, then patient and donor
positive. If patient is negative
135
However, CMV disease is an organspecific disease most commonly
affecting the lung with a very high
mortality rate [50%], CMV also causes
hepatitis and colitis. From reactivation
to disease is 2-3 weeks, unless cordblood, its probably shorter as there is
no immunity as the T cells from the
cord are totally nave.
Those at risk for CMV reactivation the patient
or [less likely] the donor are CMV seropositive
& donor positive the risk is very low and if both are
negative, the risk is, essentially, zero.
Consider a patient with aplastic anemia who
received an allogeneic peripheral blood stem cell
transplant patient was CMV positive, donor
negative. On day 24 the patient is with RLL infiltrate
with BAL growing stentotrophomonas. The patient
gets Bactrim but becomes neutropenic and after 7
days of GMCSF the ANC goes from 100 to 200. CMV
routine shows 2500 copies [more than 250 copies
warrants treatment] and the patient has low grade
fevers. The correct answer is to give foscarnet
because the patient is neutropenic. Gancyclovir will
worsen the Bactrim-associated neutropenia, this is
could lead to graft failure. Foscarnet can cause
renal failure.

Notes 2014
Cidofovir is a salvage
treatment for both
foscarnet and
gancyclovir, the response
rate is only 30% with
cidofovir. IVIg is not a
treatment for CMV
reactivation, only CMV
pneumonitis.
What about non-infectious
complications of SCT?
Veno-occlusive disease or
sinusoidal-obstructive
syndrome occurs in the
first week-3 weeks.
Engraftment syndrome
within 96 hours of PMN
recovery. Idiopathic PNA
syndrome first 120 days.
Acute GVHD up to 100
days, chronic GVHD 100
days to 2 years and
disease relapse up to 2
years, past 5 years
essentially cured.
SINUSOIDAL OBSTRUCTIVE
SYNDROME
Veno-occlusive disease aka
sinusoidal obstruction
syndrome [VOD or SOS], is
now less than 5%.
High dose alkalator
therapies did this. It is a
133, CSA [cyclosporine] is therapeutic.
No
ascites, PAI is very high. Why is the
bilirubin high? The patient is on TPN,
NPO, sepsis, CSA; it is not because of
SOS.
PULMONARY COMPLICATIONS OF SCT
Bone marrow transplant has had a
generally positive effect on outcomes
in adult acute leukemia and multiple
myeloma and may be curative in
lymphoma. However, pulmonary
complications occur in more than half
h
136
clinical syndrome typified by painful
hepatomegaly, elevated bilirubin,
ascites and edema. Sinusoidal
obstruction syndrome [SOS] is rarely
the cause. It occurs within 3 weeks of
conditioning and presents just like Budd
Chiari high dose alkalator therapy
injures venous endothelial cells with
progressive hepatic venous occlusion. It
is a clinical diagnosis. A PAI-1 test is
100% sensitive but very poor
specificity, its like a d-dimer for PE. It
also takes a while to get the blood test
back. Treatment is supportive and
complications are common.
Hepatorenal syndrome requiring CVVHD
is common in SOS. Anticoagulation and
tPA have been investigated and
indeterminate. Difibrotide is a singlestranded oligonucleotide with AT and
fibrinolytic effects on the microvascular
endothelium; it does not improve
outcome. Other treatments include
TIPS, or liver transplant.
Consider a 43 year old male with CLL
who undergoes NMA allogeneic
transplant using cyclophosphamide and
fludarabine from HLA identical sibling.
Both are CMV neg. Nausea and vomiting
ensue with TPN for 2 weeks. Day 8 the
patient gets gram negatives in the
blood. PMNs come back day 13. Day
14 bili is 3.4 with ALP
of transplant patients, resulting in admission to
the ICU in 25 to 50% of patients during the course
of illness.
Autologous bone marrow transplantation and
peripheral stem cell transplantation (AHSCT) are
used with increasing frequency. In this procedure
patients are infused with their own hematopoietic
stem cells after high-dose chemotherapy. AHSCT
has virtually eliminated graft-vs-host disease, and
infection with cytomegalovirus (CMV) and
Toxoplasma gondii are very rare compared to
allogeneic bone marrow transplantation. Although

Notes 2014
overlap is common, respiratory
complications can usually be grouped
temporally.
Most pulmonary problems occurring in
the first month [prior to engraftment]
are not infectious and include reactions
to chemotherapy, diffuse alveolar
hemorrhage [DAH], and ARDS. From
the second month on, even after
resolution of neutropenia, infectious
complications become more common.
While CMV infection still occurs in bone
marrow transplant patients, it is unusual
in AHSCT (< 2%) and usually occurs
after 2 months. The lower incidence of
chronic graft-vs-host disease may be an
important factor in the reduction of late
CMV infection. CMV pneumonia may
present with a patient who has DAH on
bronchoscopy and bilateral diffuse
infiltrates with severe hypoxemia. The
trans-bronchial biopsy reveals enlarged
lung
137
epithelial cells with large intra-nuclear
inclusions
[owls eyes] characteristic of CMV. PCR
of the BAL will reveal high CMV DNA
copies. CMV is known to cause DAH
secondary to DAD. The mortality is
extremely high reaching 60-95%! It is
treated with ganciclovir or foscarnet for
recalcitrant cases. Valgancyclovir does
not treat CMV pneumonia but may be
used as prophylaxis for 3 months posttransplant. EBV does not classically
cause bad pneumonia, nor DAH. It
typically causes a lymphoproliferative
disorder in the post-transplant setting.
Pneumocystis pneumonia is uncommon
in patients who have received adequate
prophylaxis. Aspergillus pneumonia
occurs in the early post -transplant
period, but is associated with profound
neutropenia and nodular or cavitating
infiltrates on chest radiograph. It would
be unusual to develop Aspergillus
pneumonia after the neutrophil count
has normalized.
Bronchiolitis obliterans would be an
unusual late complication in AHSCT as it
is associated with chronic graft-vs-host
disease. The radiograph is either normal
or shows hyperinflation. Pulmonary
function testing usually demonstrates
an obstructive pattern. BOOP or COP is
also an immune sequelae, though
sometimes infectious.
Differentiating organizing pneumonia
[bronchiolitis obliterans with organizing
pneumonia] from brochiolitis
obliterans. In BOOP, the CT scan shows
peripherally based, patchy opacities
with GGO and biopsy reveals
granulation plugs within the airspaces
with preservation of normal lung
architecture. BAL is with lymphocytosis
and a low CD4/CD 8 ratio.
Biopsy in BO [on constrictive
bronchiolitis] demonstrates
peribrochiolar fibrosis that can

Notes 2014
progress to complete
scarring of the lumen.
There is an obstructive
pattern on PFTs.
Fever, rash, dyspnea, hypoxemia. It is a
diagnosis
of exclusion, not fluid overload, not
DAH. Patients need BAL to rule these
things out. Patient then needs steroids,
these patients go from ICU to room air
with treatment. This diagnosis is
missed because the PMN count can rise
very rapidly.
IPS or idiopathic pneumonia is noninfectious pulmonary injury after
conditioning regimen up to 120 days or
so. IPS is a diagnosis of exclusion,
probably immune in origin. 12-15% in
all patients with high historical
mortality. IPS is thought to be a
consequence of: TNF-alpha,
transplanted donor T-cells, cytokines
release into the lungs.
Treatment high dose steroids, and there
is data that etanercept might help. The
median survival was 14 days
historically, now 140 days.
DAH is idiopathic pulmonary
hemorrhage usually in the first three
months. They present with hypoxemia
with bilateral pulmonary alveolar
infiltrates, and rising LDH, often with
hemoptysis. There is a 50% mortality.
All infectious etiologies are ruled out.
Hemosiderin-laden macrophages is a
buzz word. High dose steroids, 2-10
mg/kg once per day tapered over 2-4
weeks. rVIIa has been used recently,
but unknown why, but used in
conjunction. More than one DAH
episode gives a mortality rate of about
90%.
Consider a patient who is CMV
seropositive, and gets allogeneic
peripheral blood SCT from HLA identical
sibling [CMV negative]. There is full
donor engraftment on transplant day
h
138
Pulmonary
engraftment
syndrome
[PES] occurs within 96 hours of PMN
recovery; it is essentially an ARDS like
picture as PMNs come back online.
14. 60 days post-transplant with the patient is with
DOE, desaturations. There are BL infiltrates, rising
LDH. The patient is on PPx mediations, CMV serum
is negative. PCP PCR on BAL is positive, silver stain
negative, also macrophages are full of hemosiderin.
Positive PCP PCR is commonly positive, with
negative immune-staining.
Bilateral infiltrates, BAL with blood, elevated LDH.
The answer is DAH despite the patient having a
positive [false positive] PCP PCR from the BAL.

Notes 2014
There is no CMV is blood or
lung making
reactivation of CMV much
less likely.
GRAFT-VERSUS-HOST DISEASE
GVHD occurs up to 100 days
post-transplant affecting
the skin [sunburn], GI tract
[upper and lower more
commonly lower with large
diarrhea], and liver
[obstruction or hepatitis
both common].
Chronic GVHD occurs over
100 days skin, GI, lung,
liver, SICCA, fasciitis
syndrome. Chronic skin
GVHD looks like a sunburn
[diagnose via biopsy]. GI
tract GVHD - T cells from
donor attack the recipients
lower or upper GI
epithelium. The differential
is very broad many drugs
can do it, such as CSA,
magnesium, reglan, etc.
Also CMV colitis, C. Diff or
even Beaver Fever. CMV
does not usually cause
large volume diarrhea.
patient gets 3 L of diarrhea with
abdominal pain
on day 19 post-transplant. T-cell
chimerism shows 100% donor
engraftment. CMV PCR low positive
[negative one week prior]. No flex sig
until Monday [patient presents Friday
night].
Very few things give you 3 liters of
diarrhea. The key is get a C. Diff
sample and, when negative, give
steroids early to treat acute GVHD of
the gut; histology can persist for 5
days despite steroids, so dont fret
about biopsy yield with steroids.
139
GVHD grade depends on volume of
diarrhea. Grade I is less than 500 cc per
day, grade III more than 1 litre per day.
There is a 30-50% mortality rate with
high grade GVHD of the gut. The
diagnosis of GI tract GVHD depends
upon clinical criteria: there must be an
engrafted donor immune system, there
must be clinical symptoms, there must
be no other cause [and biopsy is
preferred]. There can be false negative
biopsies as the disease can be patchy.
The GI tract may look totally normal,
dont get thrown off, the histology can
be totally wonky [GVHD] despite a
grossly normal appearance. Acute
GVHD requires very prompt treatment
with high dose steroids. The GI tract
can go from normal to totally denuded
in one week. Consider restarting CSA.
Survival depends on grade of disease,
grade I is 100% survival, Grade IV 4560% survival. Once C. diff is negative,
start steroids in the setting of bigtime
diarrhea in the transplant patient.
Steroid refractory GVHD affects 10-15%
with a historical mortality > 85%.
Consider a patient who gets NMA PBSC
transplant [cyclophosphamide and
fludarabine] from unrelated donor, Treplete allograft. The
DISORDERS OF COAGULATION
Coagulation requires three things: a blood vessel

wall, functioning platelets and adequate
coagulation cascade. 30-40% of all patients in the
ICU will have some sort of bleeding disorder.
Consider the reason why the patient is in the ICU
[e.g. sepsis think DIC]. Also consider where the
bleeding is located, is it mucosal [e.g. petechial
think of platelet problems] or visceral [consider
coagulation defects] is it immediate [platelet
defect] or delayed [coagulation defect]. This is
because platelets act first [primary hemostasis]
and the coagulation cascade acts second
[secondary hemostasis]. Pre-morbid medical

Notes 2014
conditions and drugs are very
important when considering the
differential.
Primary hemostasis - consider looking
at a peripheral smear. There may be
platelet clumping or a decreased
number of platelets. Bleeding time is
an assessment of platelet function,
bleeding time will increase when
platelet counts are less than 100K.
Consider a patient who is given a
GPIIbIIIa inhibitor and has a sudden,
profound drop in platelet count. In
patients receiving these drugs there is
about a 3% risk of true, severe
thrombocytopenia and a 1% risk of
pseudothrombocytopenia. The gold
standard of diagnosis is to look at the
peripheral smear for clumping. If there
is clumping, it is a pseudothrombocytopenia and is the result of
EDTA.
140
Repeat in a citrate tube can prevent
clumping,
but that technique is not fool proof
Secondary hemostasis consider the
PTT and PT, there may be thrombin
inhibitors [e.g. heparin].
APPROACH TO THROMBOCYTOPENIA
Are platelets being destroyed? Are
they not being produced? Are they
being sequestered?
Drugs are common problems that
decrease platelet production [e.g.
ethanol, chemo] or there may be
infiltrate of the bone marrow.
Platelet destruction can be immune or
non- immune in nature. The treatment
is always the underlying cause, and
giving platelets rationally.
Platelets are scarce, and they only last
one week in the blood bank. There are
two flavors: random donor and
pharesis [single donor] units. There is
an order of magnitude more platelets
in the latter type. In TTP and HIT,
platelets are relatively contraindicated.
The dosing of platelets is 1 unit of
random donor platelets per 10 kg of
patient, or 1 unit of single donor
platelets per 90kg of patient. The
majority of ICU patients will not
change their platelet count when
transfused platelets.
IMMUNE-MEDIATED THROMBOCYTOPENIA
This is typified by large platelets in the
periphery. You may see only one large
platelet per HPF which corresponds to
a peripheral count of about 10K. When
immune-mediated, there is often a
very rapid and profound drop in
platelets.

Notes 2014
The treatment of ITP is
steroids, IVIg [a transient
response] if there is an
immediate need for an
invasive procedure. AntiRh[D] can be used, but the
patient must be Rh + and
presence of vancomycin. Bleeding can
occur in
these patients [about 30%] see
section 5.
THROMBOTIC THROMBOCYTOPENIC PURPURA
This is a favorite on the boards. It is a
deficiency
of the vWF-cleaving protease ADAMTS
13. Absence of ADAMTS 13 prevents
cleavage of wWF and the large
multimers stick to platelets and cause
diffuse thrombosis. It presents
classically with fever, anemia,
thrombocytopenia, renal failure and
neurological symptoms. This pentad is
usually not present. The only things
required for diagnosis are low platelets
and MAHA. Fever and renal dysfunction
are present in about 40%, CNS
dysfunction in about 75%.
The treatment of choice is pharesis, the
provision of FFP or both. The pharesis
removes the inhibitor of ADAMTS 13
and FFP replenishes this enzyme
[pharesis + FFP = plasma exchange].
Steroids are given to lower the Ab
production [which may be the cause of
the low ADAMTS 13]. Platelets are
relatively contraindicated.
HEPARIN INDUCED THROMBOCYTOPENIA
Type I HIT is non-immunogenic and
may be direct heparin binding to
platelets. It is common and may occur
in up to 30% of patients receiving
heparin; it occurs in the first two days
of heparin administration. The platelet
count rarely falls below 100K and
clotting does not occur. Platelet counts
h
141
have a spleen. If steroids fail, rituxan is
the next step. There are tPO agonist
for refractory ITP.
Vancomycin-mediated
thrombocytopenia is immune
mediated. Its an Ab only active in the
normalize with continued administration of
heparin.
In 0.3 to 3% of patients treated with UFH, the
more serious, immunogenic form of HIT can occur.
This is an IgG mediated response against heparinPF4 [type II HIT], which can occur with
administration of both UFH, and LMWH [the
shorter the polysaccharide chain, the lower the
incidence of HIT so less common with LMWH].
The platelets usually fall days 5-10 of
administration, but will occur earlier in patients
who have already received heparin. The
treatment is stopping all heparins do so when

Notes 2014
platelets drop by more than 50% OR` to
less than
50k, if there is bleeding or
if there is thrombosis
[clotting occurs in 20-50%
where venous is more
common than arterial].
HIT risk factors greatest
in CV patients and
orthopedic patients who
may have higher PF4
circulating, also ICU more
than floor patients, women
more than men, UFH more
than LMWH, therapeutic
more than prophylactic
doses and the appropriate
clinical scenario.
The diagnosis is based on
clinical judgment; the ELISA
test can be helpful which is
an anti-PF4 antibody assay.
This test is quite sensitive,
but not specific. It is
positive in more than 50%
of cardiac and vascular
surgery patients [less than
one third of these positive
patients will have HIT].
The functional assay is
platelet aggregation a
DIC is a clotting disorder, intra-vascular
clotting
fibrin deposition in the vessels which
leads to vascular thrombosis and organ
failure.
There is depletion of platelets and
coagulation factors. The diagnosis of
DIC is clinical the right clinical
condition, with an elevated PT/PTT, low
platelets, low d-dimer also with low
fibrinogen, high thrombin time, high
FDP and low anti- thrombin. Treat the
underlying condition, transfusion only
for bleeding or invasive procedures, do
not treat numbers.
142
heparin-induced platelet activation
study or a serotonin release assay
which both have high specificities but
low sensitivities.
How about those who have thrombosis?
40% will have a platelet decrease
before a thrombotic event, 30% will
drop their platelets the day of the clot
and 30% of patients will have their
platelets drop following the thrombotic
event. Further, thrombosis can occur
after heparin is stopped [for up to
weeks!].
The treatment of HIT is argatroban &
lepirudin. The latter is renally cleared
and should be avoided in renal
insufficiency; the former is hepatically
cleared. Bivalirudin is used in PCI,
fondaparinux is not approved for HIT, it
is renally excreted with a long half-life
and there are case reports of
fondaparinux HIT.
Some have suggested that anyone with
HIT [clots or not] should be
anticoagulated for weeks, but this is not
generally held. Warfarin cannot be
started until the platelet count is
normalized.
DISSEMINATED INTRAVASCULAR COAGULATION
FRESH FROZEN PLASMA
FFP is frozen, so it takes some time to thaw [it
requires 20 mins to thaw] unless at a large trauma
center. A random donor unit has 200 to 250 mL. A
pharesis unit is 2 units of FFP. Indications are a
documented coagulopathy, massive transfusion,
reversal of warfarin defect and TTP or HUS. The
initial dose is usually about 2 U of FFP or 1
pharesis unit. Note that 5-6 units of platelets
contain about 1 U of FFP [no V or VIII]. If the INR is
less than 2 there will not be much change with
FFP [normal INR of FFP is about 1.7].
CRYOPRECIPITATE

Notes 2014
Cryoprecipitate is a source of
fibrinogen. It is no longer used to treat
von willebrand disease [vWD], now use
factor VIII for vWD. DDAVP only works
in about 30%. Cryoprecipitate is a
second line choice. Use cryoprecipitate
for low fibrinogen [less than 100].
MASSIVE TRANSFUSION AND COAGULOPATHY
What about the coagulopathy of
trauma and massive transfusion? The
etiology is multifactorial. It is secondary
to hemodilution, acidosis, hypothermia
and DIC. This can be seen with massive
GI bleeders as well. The goal is to
maintain perfusion and oxygen delivery.
What is the blood product ratio? On the
surgical boards, it is typically more FFP
to pRBC. There is
143
poor data in the MICU. Is it 1:1, 1:2, 2:5? Recent
data suggests giving near whole blood
during MTP see section 9.
There is no clinical scenario where
factor VII improves outcome. When
should it be used? The patient should
be salvageable, there should be initial
interventions such as surgery and
factor replacements, the acidosis
should be corrected and patients with
thromboembolic conditions should not
receive factor VII.
HEMOPHILIA
If a patient has hemophilia and is
bleeding call a hematologist. Type A is
VIII deficiency, there are factor VIII
concentrates; these patients have very
high requirements when they bleed, it
should be an emergent consult to
hematology. Type B is a IX deficiency.
PCC can be used. Prothrombin complex
concentrate [PCC or KCentra] contains
factors II, VII, IX, and X, as well as the
antithrombotic proteins C and S. aPCC
[or FEBIA] is very similar, however
FEIBA contains activated components of
X, IX, VII and II.
PHARMACOLOGICAL MISADVENTURES
For the treatment of warfarin overdose,
FFP, PCC [KCentra], or aPCC [FEIBA]
can be used. There are reports of
clotting with some of these. PCC does
not have high levels of VII. Vitamin K
takes 24 hours for full effect.
LMWH bleeding is tough protamine
will not be very effective, can give FFP.
Reversal of fondaparinux is with FEIBA.
Reversal of the new anti-thrombin
inhibitors should begin with FEIBA if lifethreatening. Dabigatran can be
dialyzed, but there can be rebound with
cessation of dialysis.
What about anti-platelets? Plavix is
tricky consider platelet transfusion,

Notes 2014
there are retrospective
studies that suggest some
ERYTHROCYTES IN THE ICU

Understand erythrocytosis. Consider a
patient with pulmonary atresia and a
hematocrit of 76% from chronic
hypoxemia. The patients blood
pressure will be higher because of
hyperviscosity. There is no data to
support the use of routine phlebotomy
in adult patients with cyanotic heart
disease to prevent stroke. The data for
that practice comes from patients with
PCRV who are older with
thrombocytosis. In adult patients with
congenital heart disease, phlebotomy
should only be done when there are
symptoms of hyperviscosity. These
patients have a spuriously high PT and
PTT because with the erythrocytosis,
there is less plasma per volume. Thus
when their blood is put in tubes, there
is less plasma relative to anticoagulant.
Special tubes are required for
hematocrits above 55%. While chronic
hypoxemia is the major culprit for
stimulation of red cell mass, the renin
angiotensin axis is also implicated. In
fact, in some patients, the use of ACEI
can lower red cell mass!
Understand the immune-modulatory
effects of pRBCs. The transfusion of
red cells is associated with multiple,
significant immune-modulatory effects
that are NOT related to the red cells
per se. Instead, they are most likely
related to transfused lymphocytes
and leukocytes within the blood. These
transfused white cells have been
shown to circulate within the recipient
for years and even cause graft-versus
host disease. This has been implicated
in immunosuppression and even tumor
growth in some cancers! Cytokines,
plasticizers, viruses and proteins within
the transfused blood are also
implicated.
h
144
clinical improvement with platelets, but
the data is sparse.
The administration of EPO to critically ill patients
has been shown to reduce the need for transfusion
by 0.4 units per patient on average, though in ICUs
that use a restrictive blood use policy [7 g/dL
transfusion trigger] to begin with, this probably
does not make a difference. There is no change in
mortality, and there may be an

Notes 2014
increase in thrombosis risk
in critically ill patients.
There is certainly no
difference in time on the
ventilator.
Understand blood
conservation in the ICU.
Routine use of parenteral
iron has not been shown to
decrease the need for red
cell transfusion in the
intensive care unit. The
use of a lower transfusion
threshold has, so too has
145
been the use of aprotinin the cardiac
surgery, so too has been the use of
recombinant epo and autologous blood
cell transfusion from patients
undergoing elective surgery. Cell
savers have been shown to improve
hemoglobin, but it is unclear if they
reduce transfusion. When matched for
organ dysfunction, patients who
receive blood transfusion have a
higher mortality than those who dont
[Vincent J, Baron J, Reinhart K, et al.
Anemia and blood transfusion in
critically ill patients. JAMA 2002;
288:1499-1507].

Notes 2014
8. CRITICAL CARE NEUROLOGY
The dictum is less is more e.g. daily

sedation interruption with a
spontaneous breathing trial. Enacting
the aforementioned increases ventilator
free days; decreases ICU and hospital
LOS.
Strom in 2010 showed only analgesia
[no sedation] seemed to do even
better. A post-hoc analysis of the Strom
study showed that in the no sedation
group there was less fluid
administered, increased urine output;
and more renal dysfunction in the
sedated group. There was no difference
in delirium or VAP.
DEXMEDETOMIDINE
Dexmedetomidine does not decrease
the rate or depth of breathing. It is a
centrally-acting alpha 2 agonist. It has
analgesic properties and it can be used
in extubated patients. Never bolus the
patient, just start a continuous infusion.
There is less time on the ventilator and
less delirium compared to valium.
There is bradycardia with a loading
dose and decrease in the length of stay
in the ICU. If you need the patient to
participate in their care [e.g.
neurological examination]
dexmeditomidine is a good drug.
Extubations were about two days
earlier, there were fewer infections on
dexmedetomidine. Dexmedetomidine
may have a benefit in septic patients
for unclear reasons. Dexmedetomidine
is a cost-effective treatment [Dasta JF
CCM 2010 497-503].
Dexmedetomidine resulted in less
delirium and fewer ventilator-days than
midazolam [SEDCOM]or lorazepam, but
11
h
did not reduce length of stay in the ICU or

hospital, in 3 randomized trials [JAMA 2009,
2007, 2012].

Notes 2014
SEDATION
& ANALGESIA
Propofol appeared equivalent
to
Precedex in sedation efficacy,
length of stay and ventilatordays in the 2012 JAMA trial.
A 2013 meta-analysis of 6 trials
[including some of the above]
suggested non-benzodiazepines
[dexmedetomidine or propofol]
reduced ICU length of stay by
~1.5 days, ventilator days by
~2, but had no impact on
delirium or short-term mortality
rates, as compared to
benzodiazepines [midazolam or
lorazepam].
OPIOIDS
Opioids are great analgesics, but
there are side- effects [nausea
and vomiting, constipation,
pruritis, sedation, tolerance,
dependence, respiratory
depression].
What are the side-effects of mu
opioids? There is blockade of the
11
propulsive peristalsis by increased

smooth-muscle tone and inhibition of
the coordinated peristalsis required for
propulsion leading to nausea,
vomiting and unrelenting constipation.
Methylnaltrexone cannot cross the

blood-brain barrier, it does prevent
nausea and vomiting.
Methylnaltrexone [NEJM 2008; 2332]
given to treat constipation resulted in a
45% laxation response in 4 hours
compared to 15% in placebo. So there
is a roughly 50% laxation failure with
methylnaltrexone.
There are other causes of constipation
including immobility, decreased oral
intake, low fiber intake, metabolic
imbalances, advanced age, other drugs.
Constipation in the ICU is very common
5-83% and is associated with increased
LOS, increased time on the ventilator. In
patients in the ICU who were
constipated, if no BM by day three,
lactulose versus PEG versus placebo.

Notes 2014
Making bowel movements got patients
out of the ICU earlier.
Remifentanil is not superior to fentanyl
in terms of pain and sedation in the
ICU. There may be negative
immunomodulatory effects of opioids.
Morphine and remifentanyl has been
shown impair the immune response.
PROPOFOL
Propofol compared to benzodiazepines
has been shown to decrease length of
stay and potentially mortality as well
[Wunsch CCM 2009; 3031].
Propofol-related infusion syndrome
[PRIS] can cause rhabdomyolysis, renal
failure, hyperkalemia, and cardiac
abnormalities. PRIS was found in 1.1%
in a large study, it could be present
after small amounts of propofol and for
a short period of time. This data has
been shown in neuro-intensive ICUs
where high doses are given for short
periods of time and this may be a risk
for cardiac arrest. The ECG finding of
PRIS is the RBBB that is convex and
curved ST segment in the right
precordial leads. It is suggested that
trigs be monitored after two days of
use.
ATIVAN
Like propofol, it activates the GABA
receptor and is good for the production
of amnesia. Most benzodiazepines are
cleared by the liver. Ativan is
associated with metabolic acidosis at
high levels [propylene glycol] and it is
an independent risk factor for delirium.
DELIRIUM
What is the epidemiology of delirium?

About 50% in non-vent and 70% in
11
h
ventilated patients. This is acute brain injury =

organ dysfunction.
The vast majority of delirium is a quiet delirium =
hypoactive delirium. Hyperactive delirium is 5% or
less.
The cardinal feature of delirium is inattention! So
the little old lady with her head down unable to

Notes 2014
pay attention = inattentive.
To test this, you can ask a
patient to squeeze your
head every time you hear a
letter. 10 letters, 1 per
second. If they are right 8
of 10, they are attentive.
75% of delirium is missed.
DELIRIUM DEMENTIA LINK?
Delirium raises the risk of
death by 10% per day!
Delirium increases the risk
of critical care associated
brain injury. There is a 5070% risk of
neuropsychological
cognitive impairment like a
dementia. There is evidence
to suggest that the increase
in dementia risk in the
country on the whole is
being made by ICU
experiences. As compared
to those never hospitalized,
those who were hospitalized
and those in the ICU had a
graded increase in
dementia. Most of the ICU
type dementias were non-
11
Alzheimer. Sepsis seems to be a big

risk as well. ARDS too.
Part of the problem might be the

underlying leaky blood-brain barrier,
inflammatory mess that occurs in ICU
patients.
SEDATION: LESS-IS-MORE
In the mid-90s, Ely studied
spontaneous breathing trials [NEJM]
and found that 2 days were shaved off
of mechanical ventilation.
Then Kress studied spontaneous
awakening in 2000 in NEJM and also
found 2 days shaved off of the MV.
Then in 2008 in Lancet they paired
them. SBT and SAT showed benzo use
cut in half, narcotic use cut in half.
They didnt stop narcotics when the
patient was in pain. Narcotics were
stopped, only if the pain was
controlled. In modern ICUs, there were
4 days off ICU stay and reduced
mortality with an ARR of 14%.
BENZODIAZEPINES AND DELIRIUM
What about the type of drugs used?
GABA-ergics have been the most widely
used sedative in the last 20 years
worldwide, though propofol is used

Notes 2014
much more in the US. What is analgosedation?
It means using a narcotic as a single
agent = sedation and pain relief. Strom
in Lancet 2010 called this no-sedation.
Another is to use dexmedetomidine.
Another way is to use an anti- psychotic.
The use of large doses of narcotics in
burn ICUs tend to protect against
delirium even in the setting of pain as
deliriogenic. GABA drugs affect a
portion of the brain that will render a
patient unconscious, but the patient is
not getting sleep, repeat, they are not
getting slow-wave sleep.
The locus ceruleus is lower down and is
where the dexmedetomidine hits. As
above, Dex versus midaz SEDCOM
trial After 24 hours there was much
less delirium in the dex group [less
brain dysfunction], there was also less
time on MV.
ASSESSMENT
Assessing for consciousness = arousal
[RASS, etc] plus content [CAM-ICU].
When using the RASS, it separates
verbal from physical stimuli. So you
begin the scale by talking to someone,
if they dont respond with eye contact,
then they are at least -3 and this is
moderate sedation. If tactile stimulation
is required, this is deep sedation and is
a -4. Responding to neither physical nor
verbal stimulation is unarousable and is
-5. The difference between -1 and -2 is
duration of eye contact for 10 seconds.
If longer than 10 seconds, then the
patient is -1, if less than 10 seconds
then -2. This was described in 2002, it
is kind of an expanded eye response
from the GCS. How does the GCS work?
4 points for eyes and 6 limbs. The most
points for verbal is 5.
BRAIN DEATH
h
150
Irreversible cessation of circulatory and respiratory
functions is death, OR irreversible cessation of all
functions of the entire brain including the brain
stem. 2010 Wijdicks.

Notes 2014
In order to begin brain
death testing, the patient
must have a mechanism of
brain injury that would
result in brain death.
Irreversible loss of the
clinical function of the
brain, including the
brainstem, and there must
be a mechanism that must
be compatible with this.
Otherwise think about
metabolic issues.
MITIGATING FACTORS & BRAIN
DEATH
Mitigating factors include
drug intoxication, cervical
spine injury [motor exam,
apnea test do not work so
you need an ancillary test],
the recent use of
neuromuscular blockade.
Vecuronium with renal
failure can lead to a
prolonged effect of NMJ
blockade, as can
cisatracurium.
Remember that the NMBs
act upon nicotinic receptors,
the muscarinic receptors
VOR is slow towards cold ear and fast
away, this
is NOT seen if the patient is BRAIN
DEAD.
The corneal response tests V afferent
and VII efferent as does facial pain
application. IX and X is gag response.
There should be absent cough with
tracheal suction.
Note that posturing [decerebrate or
decorticate] requires brainstem
function, so if present, there is
brainstem function and the patient is not
brain dead. Further, the following DO
indicate brain activity: seizure & facial
grimacing.
151
[pupils] still work. Nevertheless, post
arrest may result in large pupils
following adrenergic or anti- muscarinic
drugs, so check reflexes then. If there
are reflexes, then there is no NMB. Look
for train of 4, if there are any twitches,
there is no prolonged NMB. Recall that
if there are 4 twitches 0-74% of
nicotinic receptors are blocked, if there
are 3 twitches then 75% block is
present, 2 twitches = 80% block, 1
twitch = 90% block and no twitches =
100% block.
The body temperature must be more
than 36 degrees celsius. Now the
systolic must be more than 100 mmHg
previous guidelines were 90.
DETERMINATION OF BRAIN DEATH
So the WHOLE brain needs to be dead,
both hemispheres and brainstem.
Thus the following really must be
present: pupils dilated and
unresponsive, usually mid-line [CN III],
the oculocephalic reflex [Dolls Eyes]
requires III, IV, VI not performed if
there is concern for spinal cord injury so
instead do the cold-caloric vestibuleocular response. A normal or
present
Importantly, upwards of 50% of brain-dead patients
can display peripheral reflexes in response to
manipulation of the head, and stimulation of the
extremities. Thus, the triple flexion response does
not require brain stem function, nor do deep tendon
reflexes as these are both spinal reflexes. The triple
flexion response is: ankle, knee and hip flex with
pain applied to the toe or inner thigh it is a spinal
reflex, so it may be present with brain death. Finger
jerks, undulating toe signs can occur with brain
death. If there is any concern get an ancillary test!
Apnea testing in patients with preexisting
conditions, such as severe COPD or sleep apnea,
may be inaccurate because these patients may
have an abnormal hypercapnic reflex. The apnea
test occurs when you pre-oxygenate the patient,

Notes 2014
normalize PCO2, unless a chronic
retainer. Apply intra-tracheal oxygen at
6 L per minute. The patient must be
undressed and evaluated for
movement. Check an ABG 8 minutes
and every few minutes thereafter to see
if the PaCO2 is above 60. If the patient
becomes unstable, get an ABG and
reconnect. What does stopping the
apnea test requires? If they are
breathing, if the saturation falls below
90 for organ donors or 85 for non-organ
donors. If there are cardiac arrhythmias,
hypotension despite bolus and pressors
[that is an SBP less than 100] or if the
PaCO2 gets above 60 or 20 above
baseline. The time of death is the time
the PaCO2 reached the
152
target level or when the ancillary test is
officially
reported.
EEG is commonly cited as a means to
confirm brain death but it is subject to
false positive brain death assessments
[patient on barbiturates, heavy
sedatives, or undetectable subcortical
neuronal activity] and false negative
brain death assessments [agonal EEG
activity or ICU artifacts]. Transcranial
Doppler should not be used as 10- 15%
of the time no blood flow will be
reported, but this is likely technical in
nature. Cerebral angiography is the gold
standard, followed by nuclear imaging
followed by MRI and CT angiograms.
Nuclear blood flow is the classic
confirmatory study, but sometimes there
is a tiny bit of flow, and there must be a
delay and repeat study 12-24 hours. You
can do an angiogram, but it is more
time-consuming and requires a dye load.
An EEG is not required unless its a child.
They are difficult to perform.
SUMMARY
So brain death is: no cough or gag, no
extensor posturing, no oculocephalic
response or Dolls eyes, but there can
be triple flexion.
10 states require 2 MDs [CA and NY two
of them], some hospitals still require 2
MDs. Except for organ support for
donors, all therapeutic modalities must
be discontinued and DO NOT require
consent.
STATUS EPILEPTICUS
Status epilepticus is which of the

following? Easily terminated, seizures
more than 2-3 minutes, non- detrimental
if non-convulsive, can sometimes only
be detected by EEG and is most
commonly caused by trauma?

Notes 2014
The answer is that it can
sometimes only be
detected by EEG [e.g. in
stopping spontaneously is essentially
zero. In
animal models neuronal damage begins
to occur at about 30 minutes. Thus,
the classical definition is 30 minutes of
continuous seizure while operationally
it is probably 7-10 minutes of
continuous seizure. Quick termination
decreases the chance of continued
seizure activity. Treat quickly if the
seizure is more than 5 minutes.
Status may be convulsive [easy to see],
non- convulsive [twilight], or partial with
continuous focal neurological
abnormalities such as motor [including
the cranial nerves] or sensory without
impairment in consciousness.
The most common causes of SE are: an
epileptic off of his or her meds or after a
change in medication or dose, ethanol,
then much less likely: infection, trauma,
tumor, stroke, anoxia, metabolic.
Status epilepticus occurs as the first
seizure episode in the majority of
patients. The overall mortality of status
is estimated as less than 30% in adults.
PATHOPHYSIOLOGY
The pathophysiology of status is excitotoxicity; there is also a reduction in the
sensitivity of the GABA receptors in the
brain with prolonged seizures, such that
GABAergic medicines can become less
effective with time. This is particularly
important in the hippocampus.
Early on in status [first 30 minutes],
there is significant motor activity and
adrenergic tone leading to low pH, high
glucose, high lactate, high heart rate,
etc. but as a seizure continues, there
begins to be electromechanical
dissociation as the brain burns out
there is a transition from convulsive to
153
the comatose patient]. After 7-9
minutes, the probability of a seizure
non-convulsive status and vital signs tend to
normalize as well as metabolic derangements. Brain
damage is much more likely and rises exponentially
with time. Non- convulsive status still causes brain
damage, but at

Notes 2014
a slower rate as oxygen is being
depleted less
profoundly.
TREATMENT
The benzodiazepine
receptors become more
refractory with seizure time.
Diazepam is very lipophilic,
it works quickly but comes
off quickly. You want Ativan
because it stays in the blood
longer. Coming off propofol
can result in re- emergent
seizures.
The treatment requires
escalation, initially with
benzodiazepines such as
Ativan 0.04-0.08 mg/kg,
then with phenytoin 15-20
mg/kg [higher doses may
provoke seizure activity and
more than 50 mg/min
administration can have
cardiac toxicity with Qt
issues], and then with
phenobarbital at 20 mg/kg;
is very long acting.
For persistent convulsive
seizures, reach for
patient has a known devastating CNS
condition,
there can be prolonged seizures that
can be treated. Anecdotally there have
been patients with a 3 month-long
seizure - who walked out of the
hospital.
154
pentobarbital. Refractory status is bad,
with very high mortality rate.
Could consider ketamine [NMDA
receptor antagonist] use a general
induction dose of 1-5 mg/kg with
infusion of 1-5 mg/kg/hour.
Valproate has been shown to be very
helpful in status epilepticus. There is no
respiratory depression. Follow ammonia
levels.
Levetiracetam, does not go through
P450s, but it is renally metabolized. It
should be adjusted in renal failure in
terms of its dose, but no need to follow
levels. Lacosamide has emerging
literature to suggest that it does not
work. Topiramate can be used.
What about hypothermia? Status has
been treated with this and seizures
have been shown to stop, but seizures
tend to recur as the patient is
rewarmed. If a patient has a paraneoplastic, or non-neoplastic limbic
encephalitis [e.g. autoimmune]
consider steroids. The mortality of SE is
dependent upon age, seizure duration.
The best prognosis is in those with
known epilepsy, the worst in ethanol
withdrawal. Unless, the
injury hypoperfusion is one such cause and why
the systolic must be kept more than 90 mmHg. In
the trauma literature, a single systolic less than 90
increases morbidity and doubles mortality.
Similarly, hypoxemia is a great cause of secondary
damage with a PaO2 of less than 60 mmHg this
increases poor outcomes from 28% to 71% and
increases mortality as well. So perfuse the patient
and improve oxygen.
INTRACEREBRAL HEMORRHAGE
Primary injury is the area of maximal

neuronal damage, the penumbra is the
area of less injured and potentially
recoverable neuronal tissue.
Hemorrhagic stroke can be seen in the putamen,

thalamus, cerebellum, pons, caudate and are often
related to hypertension, amyloidosis, etc.
Secondary injury follows primary injury

and causes further neuronal damage.
There are multiple causes of secondary
The Monroe-Kellie doctrine there are 3

compartments brain, blood, CSF. An increase in
BASIC PATHOPHYSIOLOGY

Notes 2014
one must be compensated by a
decrease in the others or the ICP will
increase. The pressure is related to
volume by compliance. The
compliance curve works against the
young.
Cerebral auto-regulation is also
important. As the CPP goes up, the
vessels reflexively constrict such that
flow remains the same. The curve is
shifted rightwards with chronic
hypertension.
In the area of ischemia, the autoregulation is lost, such that blood flow
is just pressure dependent there is a
straight line from ischemia to edema.
Remember that in addition to MAP,
155
the PaO2 and PaCO2 also have effects
on
cerebral blood flow. The PaCO2 curve to
blood flow approximates a sigmoid
curve, while the PaO2-cerebral blood
flow graph makes nearly a 90 degree
angle such that at low oxygen tensions,
there is a very sudden and abrupt
vasodilation, and increase in cerebral
blood flow which will increase ICP quite
a bit.
NEUROMONITORING
The CPP is MAP less the ICP. Normal
CPP ranges from 70-100 mmHg,
adequate is 50-60 mmHg. The normal
ICP is about 5 mmHg. The arterial line
should be zeroed at the tragus! When
treating neurological insults, the ICP
should be less than 20 mmHg and CPP
of 50-70 mmHg.
EVDs are commonly used as the ICP
monitoring device. Be careful if you
dont have a bolt to monitor the ICP.
Why? An increase in BP may be a sign
of increased ICP. If you reflexively treat
the BP, you will drastically drop the CPP.
Faced with this situation, a
consideration is to give a bolus of 3%
saline. If the MAP then drops, it
suggests that the ICP was the primary
problem [because the 3% would treat
the elevated ICP and break the stimulus
for MAP elevation]. If the BP continues
to climb in response to a 3% bolus, then
ICP might not be the primary problem.
Pain and agitation should always be
sought and treated first.
Jugular venous oximetry is sometimes
used to assess cerebral perfusion. One
trial pushed up CPP in response to
jugular venous oximetry and this fluid
overloaded patients without a change
in outcome. Pbto2 is available; this is a
direct measure of tissue oxygen
utilization. Goal- directed therapy was
used to push up PbtO2 to more than 25

Notes 2014
mmHg. Initially there was a
mortality benefit, and then
there was no change, and
then there was slightly
higher mortality.
Sometimes the PbtO2 is an early

marker of worsening systemic
oxygenation.
156

MANAGEMENT
How does one approach an ICP more
than 20
mmHg? First, HOB up which drains the

veins. If there is pain, treat with opioids
as well as sedation. What about the
second tier therapy?
Firstly, hyperventilation has a peak
effect in 30 minutes. A PaCO2 of 25-30
can cause significant vasoconstriction
and diminished blood flow.
Patients randomized to hypocapnia had
a worsened outcome. If the patients
remain at a low PaCO2, there will be a
change in bicarbonate to compensate
and there will be rebound vasodilation.
Also air-trapping [as a consequence of
hyperventilation] can increase ICP.
Hyperthermia does have a lot

of known metabolic
derangements.
excitotoxicity, though a small trial showed no

difference between mannitol and HTS.
Shivering is bad; it increases cerebral metabolic
rate. Cooling is not indicated, but fever should be
treated. Each increase in 1 degree Celsius increases
cerebral metabolic rate by 7%.
Osmotic rescue should be done.

Mannitol is 1 gram per kilogram. It acts
within 20-30 minutes. Filtered needles
must be used. It causes an osmotic
gradient from everywhere. It will pull
fluid into the vascular space, improves
blood cell rheology, and maybe a free
radial scavenger.
What is the mannitol end-point? Serum
osmolality of less than 320 but
preferred is the osmolal gap of more
than 10. Watch for osmotic diuresis and
maintain euvolemia. Mannitol can
aggravate vasogenic edema in the area
of injury if given in multiple large doses.
Only use it when you need to.
Hypertonic saline is picking up steam.
3% comes in 250 cc bolus. Run it in as
fast as possible, just like Mannitol.
Remember that a rule of thumb is that
1 mL/kg of 3% saline will increase the
patients serum sodium by roughly 1
mEq/L. 7% saline bolus or 23.4% [30 cc
bolus over 10 minutes] can also be
used. NMDA receptors can be activated
by hypertonic saline which can result in
lung.org
JE
K
11
6

Review Notes
In patients with GCS of 3-7,

hypothermia was unhelpful
in multiple trials and a large
meta- analysis; it may
decrease ICP in some
refractory ICP patients.
Pentobarbital coma may
result in unreactive large
pupils, it makes them
hypothermic and an
isoelectric EEG, such that
blood flow and blood
volume drop in the brain.
Loading dose is 10-20
mg/kg. The toxicity of
pentobarbital is sepsis-like.
They are vasodilated, but
cold, hypotensive, there is
ileus edema and
immunosuppression in
addition to unreactive large
pupils.
hemorrhagic strokes are partially

caused by high blood pressure, and
blood pressure goes up after
intracerebral hemorrhage often to
shocking levels - does this acute
hypertension represent an adaptive
response by the body, pushing blood up
into the brain where its needed? Or
does high blood pressure during an
intracerebral hemorrhage make
everything worse?
American Heart Associations
guidelines target mean arterial
pressure of < 110 mm Hg or BP <
160/90, absent evidence of decreased
cerebral perfusion pressures [MAP <
130 if increased intracranial pressure or
decreased cerebral perfusion pressure
are present].
The INTERACT trial showed rapidly
lowering blood pressure reduced
hematoma growth over 72 hours in
patients with intracerebral
hemorrhage, without apparently
hurting anyone [n=404 patients].
What BP meds to use to

maintain the CPP? Consider
levophed over dopamine if
the patient is hypotensive.
However, these patients
are rarely hypotensive,
especially at the outset. As
www.heartlung.org
The ATACH trial rapidly reduced or

normalized blood pressures, and found
it wasnt harmful [n=60 patients].
JE
K

Review Notes

Notes 2014
NEJM 2013 - INTERACT2 trial was
published 2,839
patients who had just had an
intracerebral hemorrhage [<6 hours].
Rapid blood pressure reduction [sBP
less than 140 mmHg] did no worse and
appeared to do better, with 52%
experiencing death or severe disability,
compared to 56% in the standard care
group. Mortality was identical between
groups at 12%. Interestingly, the
possible benefit seen in the first
INTERACT reduced growth of
hematoma was not found in
INTERACT 2.
159
Give IV tPA if the non-contrast head CT is negative.
The treatment of ischemic stroke is to evaluate for
thrombolysis. The non-contrast head CT is the
fastest. If there has been a longstanding stroke,
there will be edema, and effacement. DWI or
diffusion weighted imaging and PWI perfusion
weighted imaging is the use of
In summary, the blood pressure should

be treated to reduce extension of the
size of the blood. Never use nitrates in
neurological patients as they can
increase ICP; nicardipine or labetolol are
great choices.
Surgical evacuation is generally
unhelpful in ICH, so dont call the
neurosurgeons unless its a cerebellar
stroke. These can be evacuated, the
clot should be removed by the surgeon.
When else to call the neurosurgeon?
When the GCS is 8 or less with elevated
CPP, get an EVD, when there is
hydrocephalus or need for
hemicraniectomy or a cerebellar bleed
more than 3 cm in diameter.
Other things, seizure prophylaxis for
one week with phenytoin, corneal
protection, early enteral nutrition, GI &
VTE prophylaxis.
SEVERE STROKE AND SUBARACHNOID HEMORRHAGE
65 year old woman with severe PNA, at

noon, the patient is normal, at 215, the
patient is aphasic with hemiplegia.
What is the correct therapeutic step?

Notes 2014
gadolinium MRI to look for a
mismatch in the
brain to target patients
who may benefit from
intra-arterial [IA] therapy.
Also, a very large stroke
with a large perfusion
defect will not do well and
these patients will be
observed in the ICU.
MANAGEMENT OF ISCHEMIC
STROKE
3 hours is the reperfusion
window for IV tPA. Too late
will result in a leaky BB
barrier. The faster the
better the outcome. The
initial trial gave tPA at 0.9
mg/kg in patients with an
NIHSS score between 4 and
22. Up to 4.5 hours is OK,
unless over 80, on oral AC
regardless of INR, with a
large NIHSS [more than 25]
or in those with a history of
stroke and diabetes.
Thus, acute focal
neurological change
requires the time of onset
determination! The patient
must be assessed for
hyper-coagulable state, if there are
extra-cranial
carotid or vertebral dissections
[unknown efficacy], but especially for
cardioembolic strokes. The risk of
bleeding with clopidogrel plus ASA is
about the same as warfarin alone for
cardioembolic events.
What about hypertension treatment?
Start within 24-48 hours with oral
medications. How about blood pressure
control? There is a U- shaped curve for
mortality. If they are not a tPA
candidate, if over 220/120, lower by
15% in the first 24 hours. If tPA is in the
160
contra-indications to tPA. Also, check
glucose if it is very high or very low
they can have a stroke mimic.
What about IA therapy? The window is
6 hours due to an occlusion of the
MCA. But start the IV tPA while the lab
gets set up. In a patient who has
recently had surgery, there are
systems that retrieve clots. With clot
extraction, there is a stroke and death
rate of about 10%. Further, here is a
very high risk in posterior circulation,
or in patients with blood glucose above
200.
After a thrombotic-ischemic stroke
what is a reasonable treatment choice
to prevent future stroke? The answer is
anti-platelets. Do not give until 24
hours after finishing tPA. The
combination of ASA and Plavix for
stroke is NOT indicated [for stroke
prevention alone]. Options are: ASA,
ASA + dipyridamole, or Plavix. Either
ASA + dipyridamole or Plavix are
favored over ASA alone, but the data is
far from robust.
What about heparin? There is no role
in acute ischemic stroke. However,
heparin should be used in cerebral
venous thrombosis even with cerebral
hemorrhage! Also, if the patient has a
works, then get SBP less than 185. If tPA given, must
be less than 180 mmHg.
NPO, swallow evaluation, bed rest, no heparin,
warfarin or ASA for 24 hours. Statins for all patients
unless LDL less than 70.
The guidelines for glucose are 140-185 in the first 24
hours. Hyperthermia should be treated as well, with
Tylenol [APAP], but it is unknown if this helps
patients. What about steroids? They are not
effective in cytotoxic edema related to infarct or
bleed. Steroids might help with neo- vascularization
seen in tumors, abscesses, etc.
Is there a role for surgery? In young patients with
severe, diffuse non-penetrating TBI, [or rarely large

Notes 2014
ischemic MCA strokes] hemicraniectomy
patients had fewer ventilator days and
shorter ICU stay, but there was no
difference in LOS. The hemicraniectomy
patients spent less time with an ICP of
less than 20 mmHg [DECRA trial]. BUT,
the primary outcome or outcome score
was worse with surgery [made more
chronically debilitated patients].
DESTINY II trial NEJM 2014 concluded:
Hemicraniectomy increased survival
without severe disability among patients
61 years of age or older with a
malignant middle-cerebral- artery
infarction. The majority of survivors
required assistance with most bodily
needs.
Who should get a hemicraniectomy? It
may be a life-saving procedure, but
they may have life-long disability. The
size of the craniectomy is
161
important, there needs to be at least 13
cm. The
zygoma needs to be removed.
PRES
What about PRES? Hypertensive,
leukoencephalopathy, both grey and
white matter affected. There is visual
blurriness. It is not a PCA stroke. It is a
blood-brain barrier leak usually in the
setting of high blood pressure.
Immunosuppressive drugs can also do
it. So PRES is not in a vascular
distribution.
SUB-ARACHNOID HEMORRHAGE
Worst HA of life, nausea, vomiting. LP
should be done and 1st and 4th spun for
xanthrochromia.
Get a non-contrast scan. Look for early
hydrocephalus. The patient needs
angiography to look for an aneurysm.
They need an urgent neurosurgical
evaluation. Early surgery is
recommended in the first 72 hours
aneurysm clipping. Coils can be used
beyond that especially in older
patients.
Unsecured aneurysms have a 4% rebleed risk on day zero, then 1.5% per
day for the next 13 days or about 25%
in the first two weeks. If the patient
cannot be seen by a surgeon, consider
anti-fibrinolytic therapy.
Blood pressure control: systolic less
than 140 mmHg with labetolol or
nicardipine to prevent re-bleeding and
nimodipine 60 mg po every 4 hours to
prevent vasospasm. Vasospasm is the
SAH issue. The leading cause of death
and disability following subarachnoid
hemorrhage is vasospasm. Clinically
evident vasospasm occurs in 20-30% of
patients with SAH and is thought to
occur in response to spasmogenic
substances

Notes 2014
released by subarachnoid
clot lysis. Vasospasm is
very uncommon in the first
3 days following rupture. Its
incidence tends to peak
Vasospasm can present with focal
neuro deficits
days following a definitive procedure.
The benefit of nimodipine in SAH is that
it will improve the odds of a good
outcome. Nimodipine must be given
orally [60mg q4h PO or NG], IV
nimodipine has been associated with
significant hemodynamic compromise.
The treatment of choice of ruptured
aneurysm is the placement of a clip at
the neck of the aneurysm by an
experienced surgeon. Endovascular
repairs may be beneficial in select
patient groups.
Further management of vasospasm
requires triple-H therapy [hypertension,
hypervolemia and hemodilution];
however, some have questioned the
benefit of this therapy. It is probably
best to keep the pressure head up, and
consider angioplasty, or intra-arterial
nicardipine.
What about the approach
hyponatremia in the SAH patient?
Volume status is key as those with
SIADH are euvolemic and those with
cerebral salt wasting are hypovolemic.
Consider treating with 3% saline with
250cc boluses. Do not fluid restrict SAH
patients as they tend to vasospasm
more.
Lastly, the heart can get stressed out in
patients with SAH, contraction band
necrosis which causes subendocardial
problems from symphathetic discharge.
This can cause cerebral T waves, also
heart failure being takotsubo
cardiomyopathy.
Seizures are not prophylaxed with SAH.
PROGNOSIS IN SAH
162
between day 3 and day 7 or so.
Angiographic vasospasm is quite
common, but only clinically
symptomatic vasospasm portends a
poor prognosis.
Sudden death in about 20%, 58% regained
premorbid level. Hunt and Hess grading 1 is mild
headache with a mortality of 11%, Hunt and Hess
of 5 is coma and this is a 71% mortality this is for
SAH and refers to the symptoms at onset of the
lesion.
SPINAL CORD INJURY
SCI is 10-20% fatal at the scene. Breathing and

coughing require the thoracic intercostal muscles

Notes 2014
and this can result from C7
injuries. These
patients cant cough and
have impaired respiratory
status; for example, a C7
spinal cord injury [acute]
with acute dyspnea and a
normal CXR is likely caused
by parasternal muscle
weakness [board question].
Vital capacity can help
gauge the prognosis and
follow deterioration.
Normal the VC is 45 to 65
mL per kg. Once below 30,
there is poor cough and an
inability to clear secretions,
once below 25 cc/kg sigh is
lost, there is excessive
atelectasis and shunting
and once in the 5-10 cc/kg
range there will be
hypoventilation and
hypercapnia. If intubation
is required, RSI with in-line
spinal immobility is the
preferred choice, a fiberoptic approach could be
used if time is not a
concern. About one third of
163
cervical spinal cord injuries require
intubation in the first 24 hours.
Recognize and treat autonomic
dysreflexia. Consider a patient with a
T5 spinal cord injury from 6 months ago
who presents with severe sweating,
flushing, confusion, and a BP more than
200/130. He has a fully distended
abdomen from a large bladder [in
retention]. It is treated with urinary
catheterization.
Autonomic dysreflexia [AD] is
exceptionally common [20-70% of
chronic SCI patients and 5- 6% of acute
patients] and typically only occurs
when lesions are above the T6 level. A
noxious stimulus causes the insult.
When a cord lesion is below T6, the
splanchnic circulation remains with
normal autonomic control. Dilation of
this vascular bed, offsets the
sympathetic charge from noxious
stimuli so AD is much less common in
lower lesions. Untreated AD results in
pulmonary edema, seizure and death.
Other triggers of AD include fecal
impaction, medical procedures, sexual
stimulation, childbirth, abdominal
distension/gas, and somatic pain.
Once corrected, the hypertension
usually resolves, if not, then antihypertensives may be tried.
Understand the treatment of acute

spinal cord
injury with perfusion goals and steroids.
Consider a patient with a bad C5/6
transection.
12
h

Notes 2014
INTUBATION CONSIDERATIONS
In a patient with multiple sclerosis, elevated ICP and

with a ruptured globe secondary to MVA, which
medication is optimal for RSI?
Primary injury involves trauma, shear

force and contusion. Rarely is the entire
spinal cord transected, even when the
vertebrae are totally sheared.
Secondary injury involves the evil
humors thereafter; it is thought that
steroids may mitigate this risk and
been shown to improve neurological
outcome. Further, while there is little
data to support spinal perfusion goals,
it is suggested keeping MAPs above 85
to 90 mmHg to prevent secondary
ischemic injury.
The NACIS II trial [NACIS I was
controversial] treated patients with a
bolus of methylpred [MP] within 8
hours of injury followed by a 23 hour
MP infusion and there was slight
neurological improvement. If steroids
are given beyond 8 hours of injury,
there may be benefit to extending the
MP infusion dose for 48 hours. A
Cochrane review in 2012 calls for more
RCTs on this topic.
Spinal cord injury patients carry a very
high risk of DVT/PE 72 hours to 14 days
post injury [50-100% of all untreated],
so prophylaxis is paramount.
The level and severity of the spinal
injury does not alter the clot risk. If
there is a contraindication for
anticoagulation, then an IVC filter
should be strongly considered. The use
of either compression stocking or
unfractionated heparin [low dose] is
considered inadequate monotherapy
compared to LMWH. The use of both
UFH and compression stocking may be
equivalent to LMWH for prophylaxis.
NEUROGENIC RESPIRATORY FAILURE
12

Notes 2014
The answer is IV lidocaine
[not IV succinylcholine,
IV etomidate, IT lidocaine or
IV Demerol].
Patients can herniate with
intubation. Comparing IV
lidocaine to IT lidocaine has
shown that the ICP rise can
be completely blunted by IV
lidocaine.
Definitely avoid depolarizing
NMBs in patients with
neurological disorders such
as MS as they express fetal
Ach receptor subunits that
result in an excessive
depolarization with
potassium efflux and
hyperkalemia.
Etomidate can result in
myoclonus which can cause
a ruptured globe to extrude.
The patient had a ruptured
globe, so dont use.
ABNORMAL RESPIRATORY
PATTERNS
What about respiratory
patterns in neurological
insults? The respiratory
pattern first Cheyne
Stokes occurs in bi-hemispheric lesions,

central neurogenic hyperventilation is
typically mid brain, below that is
apneustic breathing [inspiration, long
pause, expiration] from a pontine
lesion, cluster breathing is below that in
the medulla which resembles Biots, and
ataxic breathing is the lowest lesion
and is totally random.
GBS
Guillain-Barre is the most commonly
encountered neurology exam scenario.
GBS or AIDP is cross- reacting
antibodies to the peripheral nerves
causing demyelination.
Immunosuppression is not helpful as
the antibodies have already been
elaborated. Classically, reflexes are
absent in GBS.
Plasma exchanges decreases the time
on MV and improves ambulation by
about 50%. Usually do 5 exchanges
over 10 days. The earlier the better. 4
treatments are better than 2, but 6 not
better than 4 in bedbound patients. If
GBS is mild, patient still ambulatory,
these patients should still get plasma
exchange. IVIg is an alternative, but
there is insufficient evidence to suggest
one over the other [IVIg versus
exchange]. It should

Notes 2014
be started within 2-4 weeks of
symptom onset.
But IV Ig likely has more side-effects,
e.g. hyperviscosity.
Sequential treatment doubles the cost,
but there is not much clinical
improvement. Sequential treatment is
pharesis followed by IVIg [for obvious
reasons] though not recommended.
AIDP can be complicated by autonomic
dysfunction such as sinus tachycardia
some can get severe pacer-dependent
bradycardia. There can be wide
variation in blood pressure.
Differentiate CIDP from AIDP. Chronic
inflammatory demyelinating
polyneuropathy presents with
symmetrical extremity weakness over
a period of months. Typically there is a
slow, compensated respiratory acidosis
such that the bicarbonate is high. There
is no clear association between
infection or immunization and the
development of CIDP. Numbness, pain
and autonomic dysfunction can occur
though motor symptoms dominate.
The LP will reveal, like AIDP, an elevated
protein. These patients commonly do
poorly with 60% having a progressive
course, 40% have a response to
corticosteroids.
Plasmapharesis and IVIg are both
beneficial in this disease, but they are
required chronically as opposed to
AIDP.
CRITICAL ILLNESS POLYNEUROPATHY
Critical illness polyneuropathy [CIPN] is
the most commonly acquired NM
condition in the ICU in up to 50%-70%
with sepsis and MODS. There is flaccid
limbs and respiratory weakness;
reflexes are present in up to 33%
[making GBS less likely on clinical
h
166
grounds]. There is axonal degeneration of motor
and sensory fibers. CPK is normal, and muscle
biopsy reveals axonal degeneration of both motor
and sensory fibers. CIPN but it doesnt manifest
until 7-14 days in the ICU.

Notes 2014
Importantly, in severe
disease, the phrenic nerve
may be involved, which
means that diaphragmatic
injury can occur.
While initial reports stated
that CIPN was mixed motor
and sensory, follow up
study
reveals
that
sensation
CAN
BE
NORMAL.
Delayed offset of
neuromuscular blockade
and frank myopathies [with
elevated CPK] are the key
differentials in the ICU.
Considering patients who
fail to wean from the
ventilator secondary to a
neuromuscular etiology the
most common cause is
CIPN, followed by a failure of
central drive followed by
phrenic nerve palsy and
primary myopathies.
Greater than 50% of those
who survive CIPN have a full
recovery.
MYASTHENIA GRAVIS
after the EGD he develops respiratory
distress
that is marked by profound hypoxemia
and an inability to ventilate. Multiple
extubation attempts are made and
each time he fails with midline vocal
cords noted. He undergoes nerve
stimulation with repetitive fatigue
noted, his edrophonium test was
positive and he had high levels of anticholinesterase antibodies.
Consider a middle aged woman who
had upper respiratory symptoms
followed by prominent bulbar
symptoms, weakness with exertion and
167
A significant proportion of people with
myasthenia gravis experience
myasthenic crisis accompanied by
sudden respiratory failure requiring
mechanical ventilation. With immune
treatment [below], most recover and
are liberated from the ventilator. The
steady decline seen in other
neuromuscular diseases causing
chronic respiratory failure does not
occur.
MGC is defined as myasthenia gravis
complicated by respiratory failure
necessitating ventilatory support - this
happens in 20% of patients, typically
within the first two years of diagnosis.
Any physiological stress can trigger an
MGC - notably infection [most often],
but also aspiration, surgery, trauma,
childbirth, reduced MG medications, or
reduction in other immunosuppressive
medication, botox injections, and
certain medications [especially
fluoroquinolones, aminoglycosides, beta
blockers and CCBs].
Consider a scenario with a patient who
is an elderly, sleepy-appearing man
with 18 months of dysphagia which
prompts an EGD. Shortly
combined hypoxemic and hypercapneic
[uncompensated] respiratory failure.
Hypothyroidism should be considered in the
differential diagnosis, however, with a PaCO2 of
over 100, there should be more prominent
cardiovascular and autonomic symptoms at such an
advanced stage of hypothyroidism. With a pH of
7.29 and PaO2 of 45 mmHg, the most appropriate
management step would be initiation of noninvasive mechanical ventilation. If succinylcholine is
an option it should not be given as this can cause
fatal hyperkalemia in MGC.
Daily FVCs can help monitor respiratory function. A
normal FVC should be above 65 mL/kg, the patient's
was 0.99L or 16 m/kg. Vital capacity < 15-20 mL/kg
or MIP that is less negative than -30 cmH2O have

Notes 2014
been proposed as cutoff values for
mechanical ventilation.
Ach receptor antibodies are about 90%
sensitive for the diagnosis of MG, but
may be only 70% sensitive in those
with only ocular symptoms. In the
latter patients, there are usually other
auto- antibodies that are positive to
muscles specific kinase [MuSK], to
sodium and potassium channels and to
titin, etc.
Plasma exchange or IVIg may be used,
with the former usually providing a
more rapid improvement though trials
have failed to find that one therapy is
superior to the other. Chronic treatment
includes steroids and
168
immunosuppressive therapy, though
steroids
[alone] are NOT typically used in
the acute management of MGC as
steroids make MG transiently
worse.
MG may be paraneoplastic [thymoma]
in 10-15% of cases [surgical resection
is the treatment here] once the patient
is stable & out of the ICU.
BOTULSIM
What about botulinum toxin? It causes
an irreversible inhibition of
acetylcholine release at the synapse.
There can be food-borne, enteric,
wound, inhalation and iatrogenic. Unlike
other toxins, botulinum toxin causes
the same disease after inhalation, oral
ingestion or injection. There is no
natural inhalation of botulism.
Difficulty swallowing begins with
symmetrical cranial nerve dysfunction.
There can be GI dysfunction. There is
no fever. Inhalational botulism is very
hard to detect. Miller-Fisher variant of
GBS can present very similarly to
botulinum inhalation as there are
cranial nerve defects. It is very similar
to MGC, Tick paralysis [no bulbar
symptoms] and paralytic fish
poisoning.
Recognize wound botulism in a skinpopper. C. botulinum infection of a
wound can lead to profound bulbar
weakness and xerostomia.
There is typically hyporeflexia often
most prominent where the infection
begins. Symptoms begin 4-14 days
after the infection. Patients should be
intubated when their FVC reaches 30%
predicted. Diagnosis requires culturing
the organism or isolating the toxin.
The treatment is the anti-toxin.
Ingestion or inhalation is not treated
with antibiotics, but maybe useful for

Notes 2014
wound botulism but that
also need debridement. GI
lavage may be used.
and this is associated with steroids and
NMJ
blockers [steroid-based]. On muscle
biopsy there is an absence of thick
myosin filaments. This is different from
acute nectrotizing myopathy where the
CPK is markedly elevated, there is
myoglobinuria, severe quadriparesis
and the muscle biopsy shows
widespread muscle necrosis. Cachectic
myopathy or disuse atrophy reveals a
type 2 fiber atrophy on biopsy.
There are the three entities of critical
illness myopathy. The clinical features
are flaccid weakness, myoglobinuria
and ophthalmoplegia.
ICU WEAKNESS: SUMMARY
So how are the causes of generalized
weakness in the ICU differentiated?
MGC is proximal, GBS is distal in terms
of limb weakness. In CIPN the weakness
is distal, rarely is there cranial nerve
involvement [unlike GBS]. Reflexes are
commonly absent in GBS and usually
[2/3rds] in CIPN and botulism, in MG
reflexes are normal as is sensation.
CIPN and AIDP have sensory deficits so
CIPN is very similar to AIDP with the
only clinical difference being cranial
nerve involvement. Acute quadriplegic
myopathy is differentiated in that the
sensory defects are rare in myopathy.
GBS typically has an ascending pattern,
but less- commonly can present in a
descending manner with bulbar
involvement. The LP in AIDP will reveal
high protein, but not so in botulism. MG
can present like this, but should
improve with edrophonium testing.
NEUROGENIC PULMONARY EDEMA
MYOPATHIES
169
What about myopathies? Thick filament

myopathy will result in a small increase
in CPK
capillary pressure. Treatment is supportive in
nature with oxygen, PPV, diuretics,
alpha- adrenergic blockers, and
dobutamine.

Notes 2014
Neurogenic pulmonary edema is last
classically within minutes to hours after
neurological injury. To be diagnosed,
these patients should not have received
tons of fluids. NPE can be seen in
catasthrophic neurological injuries. It is
clinically like congestive heart failure.
Sympathetic innervation causes
increased pulmonary venous pressure
with a transient increase in pulmonary
170

Notes 2014
9. CRITICAL CARE
OBSTETRICS &
ENVIRONMENTAL
In general, drugs to avoid in OB ICU

patients are: ACEI, barbiturates,
warfarin, fluoroquinolones; also void
drugs that lower placental blood flow.
PHYSIOLOGICAL CHANGES DURING
PREGNANCY
The physiological changes of the
cardiovascular system include an
increase in: blood volume [plasma >
red cells], cardiac output [SV and heart
rate], oxygen delivery relative to
consumption [but also an increase in
consumption] and a decrease in: blood
pressure [lowest in second trimester]
and systemic vascular resistance.
Filling pressures tend to remain stable.
Respiratory changes include an
increase in: Mve [by 20-40%, mostly
tidal volume with a chronic respiratory
alkalosis], airway edema and a
decrease in: functional residual
capacity and chest wall compliance.
There is no change in the A-a gradient
or vital capacity in the sitting position.
Reflux is worse, so aspiration may be
more likely, especially if supine [i.e.
Mendelson syndrome].
The AST and ALT levels tends to drop,
such that an increase is significant. The
WBC may be elevated in pregnancy
and pregnancy is a thrombotic state as
all clotting factors are increased.
HYPERTENSIVE DISORDERS OF PREGNANCY
The hypertensive disorders are a
common cause of maternal death and
ICU admission. In the ICU preeclampsia is seen. HTN after 20 weeks
gestation [BP more than 140/90] with
12
h
proteinuria is the definition of pre-eclampsia

note that there is no need to have edema to
meet the definition.

Notes 2014
Severe pre-eclampsia occurs
with systemic symptoms and a
OBSTETRICS
higher blood pressure. There is
often renal dysfunction,
pulmonary edema, blurred
vision, AMS, hepatic dysfunction
and low platelets. Eclampsia
occurs when seizures are
present and this can occur up to
2 weeks postpartum.
Severe pre-eclampsia should be
admitted and delivery should be
considered if the fetus is more
than 34 weeks. Otherwise it
depends on the status of the
fetus. The two keys aspects of
management are seizure control
and blood pressure control.
These patients should not get
diuretics! Magnesium has a
mortality benefit; it should be
given IV to prevent seizures. A
diastolic blood pressure above
100-115 is an indication for
12
magnesium. Severe hypertension plus

organ dysfunction, also indicates
magnesium administration. Antihypertensives are indicated; its a
hypertensive emergency. A diastolic
pressure goal is 90 mmHg. The
medication of choice is labetolol,
though hydralazine is equivalent.
Nifedipine and nicardipene are safe.
AVOID nitroprusside, diuretics and ACEI.
Placental blood flow does not have
autoregulation, so if maternal blood
pressure drops, so too does the fetal BP
and flow.
In overlap with severe pre-eclampsia is

the HELLP syndrome; it is perhaps,
more dire. There is hemolysis, elevated
liver enzymes and low platelets [less
than 100-150K]. The lower the
platelets, the worse the outcome. There
is less hypertension and high BP can be
absent in about 20%. This syndrome is
associated with roughly 10% of patients
with pre-eclampsia and

Notes 2014
eclampsia. However, HELLP may also
occur postpartum. HELLP is also
typified by malaise, non- dependent
edema and RUQ pain. The
pathophysiology of HELLP is not totally
known, but it is thought that
vasoconstriction leads to RBC and
platelet destruction. The liver tends to
be most affected and hepatic
hematomas can form. An acute
abdomen, shock, or severe RUQ pain
should prompt a search for a ruptured
sub- capsular hematoma. This is not
the same as acute fatty liver of
pregnancy which can present as acute,
fulminant hepatic failure without
hemolysis [see below]. As well, HELLP
can present with ICH and stroke.
The treatment is urgent delivery and
magnesium sulfate, BP control as
needed. If there is TTP, plasmapharesis
can be done. Dexamethasone has
been tried, but an RCT did not show
benefit.
TTP OF PREGNANCY
TTP in pregnancy can occur and has
some overlap with HELLP. There
should be normal liver enzymes, there
is more mental status change in TTP;
DIC is NOT present in TTP and TTP will
not resolve following delivery. TTP is
treated with pharesis.
AFLP
Acute fatty liver of pregnancy is a
fulminant, uncommon disease in
pregnancy. It is not cholestasis or
pregnancy or the high liver enzymes of
HELLP. There is low albumin, high
ammonia, low glucose, and very high
liver enzymes. The treatment is
termination of the pregnancy.
CARDIOMYOPATHY
12
h
Peripartum cardiomyopathy can occur in the final

month of pregnancy to 5 months postpartum.
The LVEF is less than 45% for diagnosis and there
must be no other cause. The management is
similar to systolic heart failure. The majority of
these cases occur following delivery, so ACEI can

Notes 2014
be
given.
Strong
consideration
should
be
given
to anticoagulation
because there is a high
incidence of thrombosis with
peripartum cardiomyopathy.
The patient should not get
pregnant again because it
can recur and if LVEF does
not spontaneously improve,
mortality can be quite high.
HEMORRHAGE & TRAUMA IN
PREGNANCY
World-wide the most
common cause of maternal
death is hemorrhagic shock.
Painless bleeding may be a
hint at placenta previa,
these patients need
transfusion and operative
delivery.
Placental abruption can
present with painful
bleeding and is associated
with DIC. There may be no
overt bleeding. Ectopic
pregnancy can occur early
in pregnancy. Uterine
rupture, atony and retained
placenta may also cause
profound hemorrhage. The
source of bleeding needs to be

controlled, and this may involve
delivery and hysterectomy.
12
Trauma in pregnancy should begin with

being placed in the left-lateral
decubitus position to improve venous
return. A pregnant woman can lose 2
liters of blood and still have normal
vital signs. The fetus will show
abnormal vital signs much earlier than
the mother. Elevating the right hip can
also move the placenta off of the vena
cava to improve venous return. Rh ve
blood can be given for blood loss.
ACLS in a pregnant woman should also
begin with elevating the right hip, and
displacement of the left breast for the
defibrillation pad. Perimortem c-section
can be done if the fetal age is more
than 24 weeks.
Hemodynamics in the traumatic
pregnancy patient can be difficult.
Recall that there is an increase in
cardiac output [increase in heart rate
and stroke volume] and a decrease in
systemic vascular resistance which
tends to improve oxygen delivery.
Plasma volume increases more than
RBC volume, so there is a relative
anemia.

Notes 2014
The uterine artery is normally
maximally dilated
during pregnancy. Fetal oxygen
delivery is a function of 1. Maternal
oxygen content and 2. Uterine artery
blood flow.
The normal umbilical vein oxygen
tension is in the high 30s, but this
results in an oxygen saturation of 8090% because of the marked left shift of
the fetal hemoglobin dissociation curve.
Because under normal conditions, the
uterine artery is maximally dilated,
acidemia does not facilitate uterine
blood flow, but acidemia may improve
fetal oxygen transfer by right-shifting
the maternal Hb curve. Maternal and
fetal circulations interact via a
CONcurrent exchange mechanism
which is less efficient than a
countercurrent exchange mechanism.
To compensate, the fetal hemoglobin
curve is normally left-shifted relative to
the mothers such that it is highly
saturated in a lower oxygen tension
state. Fetal hemoglobin is also much
less sensitive to maternal pH change.
There will be reduced uterine blood
flow during states of maternal
alkalemia, profound hypotension and
catecholamine release.
Maternal alkalemia also left shifts the
maternal hemoglobin-dissociation
curve which reduces oxygen delivery
to fetal hemoglobin.
The most important aspect of
maintaining fetal oxygen saturation
and content is maternal oxygen
content [i.e. via hemoglobin and
cardiac output] much more than
maternal PaO2.
MECHANICAL VENTILATION & RESPIRATORY
CONSIDERATIONS IN PREGNANCY
Mechanical ventilation should begin
with a smaller endotracheal tube [7.0].
h
175
Maintain a saturation above 94% and a PaCO2 in
the low 30s to replicate the normal compensated
alkalosis of pregnancy, i.e. the pH should be kept
normal.
Permissive hypercapnia effects on the fetus are
unknown.

Notes 2014
Higher airway pressures are
OK because of the
decreased chest wall
compliance. Note that
situations that reduce
maternal minute ventilation
can lead to significant
hypoxemia in the mother.
Maternal FRC is reduced as
a function of reduced ERV
and RV [vital capacity is not
reduced]. There is an
increase in oxygen
consumption and
respiratory work. The
increase in oxygen
consumption and reduced
FRC lowers the mothers
oxygen reserve such that in
the face of hypoventilation
or apnea, hypoxemia may
develop rapidly. Remember,
that in pregnancy there is
an increased tidal volume
and minute ventilation
[stimulated by the increased
carbon dioxide and
progesterone] which leads
to a normal physiological
compensated respiratory
alkalosis in pregnancy.
DVT/PE can occur as pregnancy is a prothrombotic state. They usually occur

postpartum. D-dimer levels are totally
unhelpful in pregnancy. The first step is
usually LE ultrasound. V/Q scans can
be done, usually just a Q scan is done
and if normal, the patient is low risk.
CT angiograms exposed the engorged
breasts to a high radiation dose and
should be avoided. If a pregnant
patient is diagnosed with a DVT or PE,
the treatment with LMWH is at a higher
dose.
There is a higher volume distribution in
pregnant woman. Coumadin is contraindicated.
176
Considering other respiratory issues in
pregnancy, tocolytic associated
pulmonary edema should be
considered. This can occur in response
to beta-agonists such as terbutaline.
The mechanism is not completely
known. The treatment is diuretic and
cessation of the tocolytic. Mechanical
ventilation is not typically needed to
treat this disorder.
Asthma is most likely to get worse
during second trimester. Inhaled agents
are preferred, systemic steroids are
safe.
What about amniotic fluid embolism
[AFE]? The presence of DIC should
really prompt this diagnosis. It can
present in a very similar fashion to
venous air embolism in terms of
sudden cardiovascular collapse, but
AFE presents with DIC, so bleeding can
occur. Tumultuous labor, precipitous
deliveries may be the basis for AFE.
Venous air embolism may present with
a mill wheel murmur. The patient should
be placed in left lateral decubitus to
trap air in the right atrium, but this is a
time-honored tradition and not based
on much evidence. 100% FiO2 should
be administered and consideration for
hyperbaric therapy.
Treatment should be continued for 6 weeks postpartum or 3 months total. Thrombolysis has been
done in pregnant women [pregnancy is a relative
contra-indication to tPA].
TOXICOLOGY
Approaching the overdosed patient requires the

ABCDs. Remember the coma cocktail with
naloxone 0.4 to 2 mg, thiamine 100 mg, 25-50
grams of glucose remember that fingersticks
can be unreliable.
The timing of medication ingestion is important as
well as if the drug formulation is regular or

Notes 2014
sustained released and if the drug is
used acutely or chronically.
The important findings on examination
are vital signs and neurological
examination. Things that speed up the
vital signs include: amphetamine
derivatives, anti-cholinergics while
things that slow down the vital signs
include the sedatives, narcotics and
pro-cholinergics or anti- adrenergics.
GASTROINTESTINAL CATHARSIS
Induced vomiting and cathartics are
no longer used. Gastric lavage is
falling out of favor for most drug
overdoses. It can be considered if a
lethal substance is ingested within
one hour. Very large tubes are
required for lavage and there are
complications of their use.
177
Whole bowel irrigation [intensive
application of
Go-Lytely] is suggested but there is
probably poor clearance and it is
probably unhelpful. It is 1-2 liters of
PEG and it requires NG tube as no one
will drink that much.
Activated charcoal [AC] also has a 1
hour time limit because beyond that
the drug is passed the pylorus and
absorbed. The dose of AC is one gram
per kilogram and should be used
within 1-2 hour of ingestion. It does
not absorb iron or lithium.
If someone is critically ill with a lifethreatening ingestion, give charcoal
and consider lavage. If moderately ill,
charcoal is given [if early ingestion]. If
a benign ingestion, just monitor.
OTHER MEANS OF ELIMINATION
Other means of elimination dont use
forced diuresis. Alkaline diuresis is
used for salicylates and barbiturates.
Hemodialysis and hemoperfusion may
be seen in the ICU. There is most
experience with iHD and it gets toxins
out quickly as compared to CVVH.
Hemoperfusion is a technique whereby
blood is run across an absorbent
substance such as charcoal to remove
toxins that are highly lipid or protein
bound and therefore difficult to dialyze.
If drugs can be removed by dialysis, it
is preferred because of the high
incidence of complications with
hemoperfusion including: leukopenia,
thrombocytopenia, hypoglycemia,
hypocalcemia, and hypotension. For
this reason, hemoperfusion is generally
reserved for the treatment of severe
intoxications not amenable to other
therapies.
Drug intoxications that are amenable
to treatment with hemoperfusion
include: paraquat, tricyclic

Notes 2014
antidepressants,
barbiturates, theophylline,
methaqualone, and
glutethimide.
due to APAP and many of these
patients were
unaware of the amount of APAP that
they were taking. These patients will
have high liver enzymes, high APAP
level [above 10] and sometimes
fulminant failure.
Assess the APAP level 4 hours after the
ingestion, before 4 hours, do not draw
a level. The nomogram is only valid for
single, acute ingestion.
Oral or IV NAC can be given and it is
most effective in the first 8 hours. The
short course of IV NAC may not be
sufficient. The critical, critical factor is
to start the NAC within 8 hours because
hepatic failure is low if NAC started
early. If it is a chronic, multiple or late
ingestion, still give NAC. Charcoal can
be given for APAP OD.
TOXIC ALCOHOLS
All alcohols are degraded by alcohol
dehydrogenase to various acids or other
compounds [e.g. isopropyl alcohol goes
to acetone]. Once the alcohol is
metabolized, the osmolal gap will
shrink and the anion gap will rise if the
byproduct is an anion [e.g. isopropyl
alcohol metabolism does not raise the
anion gap because acetone is not an
anion].
Once the osmolal gap has normalized,
there is little or no benefit to giving
fomepizole or IV ethanol to block
alcohol dehydrogenase because the
anionic horse is out of the
dehydrogenase barn. If the patient is
still quite ill, the patient needs
hemodialysis at this juncture.
TYLENOL OVERDOSE
178
Tylenol [APAP] should always be

considered as a co-ingestion. 40% of
acute hepatic failure may be
The situations where you might see an elevated
osmolal gap without an anion gap would be an early
presentation of ethylene glycol or methanol
ingestion [which is clinically uncommon], if there is
co-ingestion of a toxic alcohol with ethanol as the
ethanol will be metabolized first such that the toxic
alcohol will hang around longer and contribute to
the osmolal gap, or if the ingestion is isopropyl
alcohol for reasons already discussed. A common
board scenario is a toxic alcohol

Notes 2014
ingestion that provides and osmolal gap
with
negative anion gap but
positive ketones. The
answer is isopropyl alcohol
ingestion.
Oxalate crystals are
present in less than one
third of ethylene glycol
ingestions and methanol
presents with visual
disturbances. These toxic
alcohols are treated with
inhibition of alcohol
metabolism and then
hemodialysis to remove
alcohol metabolites and the
alcohols themselves. HD is
required when there are
very high levels of the
alcohol [more than 25
mg/dL], metabolic acidosis,
renal failure or visual
symptoms. There is a high
mortality rate, otherwise.
Ultimately, most of these
patients require a vigorous
dialysis. Folinic acid can be
helpful with methanol
toxicity. The patient should
also get thiamine, folate
and glucose.
179
What about propylene glycol toxicity?
This is seen in the patient with high
dose of Ativan infusion in the ICU
typically more than 3 days and there is
more agitation with an evolving anion
gap and osmolal gap. The patients
lactate will rise and there may be
seizures, arrhythmias and hemolysis.
The patient may be thought of as
getting septic. Midazolam does not
have prophylene glycol. Propylene
glycol is a commonly used diluent in
many ICU medications including IV:
Ativan, valium, phenobarb, pentobarb,
phenytoin, etomidate, esmolol,
nitroglycerin and IV Bactrim as well!
Build-up of propylene glycol causes
water to shift from the intra-cellular
space and this causes a hypertonic
hyponatremia [dilutional]. Propylene
glycol is acted upon by alcohol and
aldehyde dehydrogenase respectively
to produce lactate which causes the
lactic acidosis. Interestingly,
commercially available propylene
glycol is both D and L such that dlactate is produced by propylene glycol
and will accumulate potentially the
cause of the mental status change.
Dialysis is typically not needed to treat
propylene glycol.
disease may have a more pronounced effect from

dilaudid.
BENZODIAZEPINES
NARCOTIC MISADVENTURES
Narcotic overdose is getting very
common. Oxycodone and hydrocodone
are common causes of death. The
antidote is naloxone 0.4-10 mg [per
dose]! It can be given endotracheally
and also injected sublingually! Fentanyl
is not detected on drug screens.
Overdose with narcotic patches often
require a naloxone infusion because the
patch creates a small depot- like effect
in the skin; there will be hours of
narcotic release.
Morphine and Demerol clearance.
Demerol [meperidine] is metabolized to
normeperidine which is 2-3 times as
neurotoxic as meperidine. It can
accumulate in renal failure and cause
seizures. Morphine is metabolized to
morphine 6 glucuronide which also
accumulates in renal failure and is
much longer acting and a potent CNS
depressant.
Comparing fentanyl with morphine
more likely to see chest wall rigidity
with fentanyl [and alfentanyl and
sufentanyl] than morphine. The
mechanism is not known. Morphine has
more cardiovascular effects that
fentanyl including histamine release,
veno and vasodilation, and sino-atrial
node blocking properties [? stimulation
of the vagus nerve]. Both morphine and
fentanyl are metabolized by the liver
and then cleared by the kidney.
Morphine has metabolites that are
opioid and therefore build in renal
failure [not fentanyl]. Both fentanyl and
morphine will be cleared less effectively
in the elderly. Chest wall rigidity seems
to occur at fairly high doses of fentanyl.
Hydromorphone, is the least affected by
renal failure as it does not have active
metabolites and does not accumulate.
Elderly patients and those with liver
www.heartlung.org
JE
K

Review Notes
Overdoses usually result in an altered sensorium.

SYMPATHOLYSIS
Moving to benzodiazepines. Alprazolam
When patients present with bradycardia

dont just think BB and CCB. There may
be digoxin, clonidine, or pro-cholinergic
h
drugs as the cause as well.
The treatment of BB and CCB overdose

is a 2-5 mg bolus of glucagon [an
inotrope] plus an infusion, calcium
chloride one gram, ventricular pacing.
IV catecholamines may or may not be
helpful. Milrinone may work better as it
by- passes the beta-receptor. Insulin
[also an inotrope], eugycemia can be
tried with an insulin infusion and
glucose infusion. Lipid emulsion [intralipid] is being used to counteract antiarrhythmic overdose as well.
1 29
s the
most common and most

toxic in overdose. These
patients are treated with
intubation and support.
Have an exceptionally high
threshold for giving
flumazenil [especially in
chronic benzo use] because
it will result in unremitting
seizure activity. Midazolam
is metabolized in the liver by
3A4 to alphahydroxymidazolam which is
a potent sedative and is
renal-cleared will therefore
become prolonged in the
kidney injured chap.
PESTICIDES & HERBICIDES

Pesticide intoxication
[organophosphates] present with
cholinergic syndromes. Sarin is a nerve
gas that causes a cholinergic syndrome.
It is the hyper-secretory syndrome that
includes urination, lacrimation,
salivation, defecation, etc. The
treatment is atropine [first], because
these patients can die from secretions
in the airway.
Pralidoxime is given next to overcome
the other side-effects; pralidoxime
breaks the bond
Zolpidem is an
imidazopyridine compound
with relative selectivity for
the typeI GABAA
benzodiazepine receptor. It
has no pharmacologically
active metabolites and is
eliminated primarily by renal
excretion.
lung.org
JE
K

Review Notes

Notes 2014
between the organophosphate and the
enzyme
anti-cholinesterase; this allows the
enzyme to return to inactivating
acetylcholine at the synapse.
Paraquat is an herbicide that causes
multi-organ badness including
pulmonary edema, pulmonary
hemorrhage and pulmonary fibrosis
[perhaps mediated by oxygen toxicity].
It is treated with hemoperfusion.
SYMPATHOMIMETICS
Cocaine toxicity is multi-faceted. It is a
sympathomimetic with a transient
surge of blood pressure. Acute coronary
syndromes can occur and the temporal
relationship to the time of cocaine use
can be variable [e.g. days to weeks
after use].
The treatment of cocaine chest pain is
ASA, nitroglycerine, benzodiazepines,
phentolamine, BB, and potentially,
reperfusion. Beta-blockers are
probably OK in cocaine chest pain.
Beta- blockers, retrospectively, may
improve outcome in cocaine users
[Arch Intern Med. 2010; 170 (10):8749].
Intra-cranial bleeds can occur with
cocaine use. Stroke can occur with
complete occlusion of a cerebral vessel.
Cocaine augments vasoconstriction and
platelet aggregation.
Pulmonary edema can occur with
cocaine, it is less common now
because the purity is better and there
is less pulmonary edema. Sometimes
the cocaine is cut with levamisole and
this can cause leukopenia.
Other problems are hyperthermia, renal
failure, bowel infarction and
rhabdomyolysis.
Amphetamines and
methamphetamines have a similar
h
182
toxic profile, but they are longer acting so there can
be chronic cardiomyopathies and pulmonary
hypertension. Charcoal and gastric lavage have a
limited role. Mostly treatment is benzodiazepines.

Notes 2014
Ephedra was banned in 2004, so its
toxicity is less
commonly seen. But new
supplements have
synephrine in them [aka
bitter orange] and there
are reports of vascular
deaths here. Caffeine and
energy drinks have been
associated with cardiac
arrest and
hypercoagulability. It is a
common cause of atrial
fibrillation in the ED.
Hospital acquired overdose
can occur.
INHALATION OF TOLULENE
Recognize the electrolyte
panel of a patient who is
sniffing glue. Tolulene is
metabolized to benzoic
acid and then hippuric
acid. Each molecule of
hippuric acid is buffered
by a bicarbonate such that
bicarb levels can be very
low. The hippurate is
renally excreted and pulls
potassium with it such
that potassium levels can
also be very low. Quite
often with glue sniffing,
given, lavage can be considered.
Hypertonic
saline in refractory cases. The TCAs
block sodium channels on the
myocardium, so giving high sodium can
overcome this toxicity [probably why
bicarbonate works and why hypertonic
saline can also be tried]. If you use a
vasopressor, choose an alpha agonist.
Valproic acid toxicity [VPA] can present
with coma, hypotension,
hyperammonemia that is due to
mitochondrial dysfunction. Check a VPA
level and may need serial levels if it is a
sustained release form. There can be
183
there is a mixed anion gap non-anion
gap metabolic acidosis [e.g. anion gap
is 22, but the venous bicarb is 10].
Recall, assuming a normal AG of 12,
the excess of 10 anions would cause a
predicted venous bicarbonate of 15
[25 10].
However, a measured bicarbonate of
10 means that there is bicarbonate
wasting. With respect to glue-sniffing,
this reflects a tolulene-induced distal
[type I] RTA [inability to acidify the
urine]. In general, the hippurate effect
is greater than the RTA effect.
PSYCHIATRIC MEDICATIONS
SSRIs cause CNS depression, seizures,
arrhythmia, tremor, hyperreflexia.
Remember that linezolid and tramadol
are MAOIs that if used with cocaine can
cause the serotonin syndrome.
The treatment is removing the
precipitating agent, cooling, sedation
with benzodiazepines, NMB [rarely],
and serotonin antagonists [consider
cycloheptadine].
Tricyclic anti-depressants can cause a
wide- complex arrhythmia. SSRIs and
atypicals can do this as well, but less
common. The treatment is sodium
bicarbonate. If early, charcoal can be
severe hemodynamic instability. There are case
reports of treatment with L-carnitine and
hemodialysis. L-carnitine has a bad smell.
Sometimes VPA takes days to wear off.
Gabapentin toxicity is becoming very common
especially in the setting of renal failure. It presents
with AMS and coma.
Lithium has a narrow therapeutic index and toxicity
can occur with acute usage, acute on chronic, or
with chronic usage. Typical symptoms with acute
poisoning include gastrointestinal complaints
initially, followed by central nervous system
symptoms. Although adequate urine output should
be maintained, forced diuresis is not effective and

Notes 2014
diuretics worsen lithium toxicity by
causing salt and water depletion and
increased lithium resorption. Dialysis is
effective for lithium overdoses, but it
should be kept in mind that
redistribution between intracellular and
extracellular departments can result in
a rebound elevation in lithium levels 6
to 8 hours later.
Recognize neuroleptic malignant
syndrome. This may be presented as a
patient who is postoperative and
receiving Haldol for agitation. He
develops a fever to 105 with profound
rigidity [not hyperreflexia] and
hypertension with AMS. You will likely
be told that the CPK is mildly elevated.
NMS is a central hyperthermic
syndrome. NMS may also present with
rhabdomyolysis and autonomic
dysfunction;
184
while the classic trigger is Haldol, all
typical and
atypical antipsychotics can be at fault.
Depot forms of Haldol can result in NMS
for a long time. NMS is idiosyncratic and
typically in response to dopamine
blockade [or the removal of dopamine
agonsits such as Parkinson
medications]. It tends to affect younger
males. Re-challenge does not reliably
reproduce symptoms. Untreated NMS
has a mortality upwards of 30% as it
can go on to produce multi-organ
system failure including hepatic
necrosis, DIC, rhabdo, cardiac
dysfunction with arrhythmia. While this
syndrome is not a disorder of a calcium
channel, dantrolene can be effective in
mitigating the symptoms of NMS by
preventing calcium entry into the
muscular cytoplasm. The rigidity is
thought to be the primary mechanism
of injury, though CNS autonomic
dysfunction also plays a role.
Treatment has also consisted of
administration of dopamine agonists
[e.g. bromocriptine, amantadine and
levodopa/carbidopa are used in that
order]. Death is usually from aspiration.
Malignant hyperthermia is fairly similar
in clinical presentation to NMS, but it is
rarer and occurs within 30 minutes of
exposure to an inhalational anesthetic
or depolarizing muscle relaxant
[nitrous oxide is not a culprit in MH]. It
is a genetic defect in calcium transport
in skeletal muscle. There is variable
inheritance patterns, but MH is typified
by hyperthermia, muscle contraction,
and cardiovascular instability.
Increased PaCO2, hypertension, skin
mottling, and masseter spasm with
muscle rigidity all signal MH especially
a rapidly increasing PaCO2.
Hypothalamic regulation is intact in
MH. MH is like a rapidly progressing
rhabdo with profound rigidity and rapid
cardiovascular collapse. The treatment

Notes 2014
is giving dantrolene and
stopping the inhaled
anesthetics start
propofol. If dantrolene has
185
been administered, do not use calcium
channel blockers as they can interact
to produce fatal hyperkalemia and
cardiovascular collapse.
hypertension, hypertrichosis, gingival hyperplasia,

hypomagnesemia hyperkalemia, photosensitivity,
skin lesions and neurological manifestations [the
latter range from hand tremors to seizures, coma
and death].

ASA
Salicylate intoxication may be seen on
the board exam. It is known as
pseudo-sepsis syndrome. Chronically,
it presents with AMS, fever, acidosis,
pulmonary edema. Its clue on the
boards is the metabolic acidosis with
an exaggerated respiratory alkalosis.
The treatment is urinary alkalanization
and potentially dialysis.
HYPOGLYCEMIA IN THE DIABETIC
Oral hypoglycemic agents are common
board testing overdoses. Glucagon or
octreotide should be given. Octreotide
is a somatostatin analogue that blocks
insulin release. It is given 50-100 mcg q
8 hours SQ. It shortens the
hypoglycemic interval.
CHEMOTHERAPIES
Recognize ATRA syndrome. Consider a
young APML patient who received ATRA
and then developed pulmonary edema
and effusions. A high percentage of
patients develop ATRA within the first
three weeks of therapy with the drug
and the symptoms include unexplained
fever, weight gain, respiratory distress,
interstitial pulmonary infiltrates, and
pleural or pericardial effusions.
In the past, the ATRA has been fatal in
up to 50% of untreated cases. High
dose corticosteroid therapy [10 mg IV
dexamethasone b.i.d. for 3 or more
days] is effective in reversing the
syndrome in a high percentage of
cases. In a recent study, 45 of 172
(26%) patients developed the
syndrome. Of 44 treated with
dexamethasone, only 2 died.
Prednisone prophylaxis during
treatment with ATRA will reduce the
incidence of pulmonary toxicity.
Cyclosporine side-effects include
nephrotoxicity hepatotoxicity,
www.heartlung.org
JE
K

Review Notes
interfere with ADAMSTS13 which then

results in an accumulation of vWF
multimers and platelet aggregation.
Both calcineurin-inhibitors [tacrolimus
and cyclosporine] are well known to
cause TTP-HUS. It is OK to substitute
one calcineurin inhibitor for another in h
this situation. Treatment also includes
plasma-exchange which does multiple
things: replaces normal ADAMSTS13,
removes bad wWF and removes bad
antibodies against ADAMSTS13 if
present.
Ketoconazole markedly inhibits 3A4
such
1 3 2
that th
degradation of cyclosporine
is inhibited and can lead to
profound increases in
cyclosporine levels. Be
aware of the transplant
patient started on
ketoconazole for thrush - a
seizure may result.
SNAKE BITES
There are 20 species of venomous
snakes in the
U.S. with the majority of human
envenomations occurring at the fangs
of croatalidae [pit vipers] which include
rattlers, cottonmouths and
copperheads. Croatilids produce venom
that is a mixture of enzymes and toxic
proteins that cause serious badness
including - DIC, myocardial toxicity,
vasodilatation, myoglobinuria and AKI as well as significant local tissue
destruction.
Recognize calcineurininhibitor [cyclosporine]

induced TTP/HUS. Consider a
patient post lung transplant
two months prior and is on
TMP-SMX as well as
cyclosporine. She presents
with fever, new anemia,
thrombocytopenia, renal
failure and confusion. TTPHUS has multiple causative
agents including infection
from E. Coli 0H157, drugs
[quinine, Plavix, OCPS,
mitomycin], malignancy,
pregnancy, HIV and bone
marrow transplantation.
Some drugs are known to
lung.org
The degree of envenomation can be

judged by the progression of local site
erythema, and - more importantly - the
severity of systemic signs and
symptoms [tachycardia, blood
pressure].
CBC and coags should be obtained
frequently.
JE
K

Review Notes

Notes 2014
Local excision to the fascia of a bite has
been
advocated, though only very early after
the bite; the presence of systemic signs
suggests that tissue dissection is too
late. Antivenom [AV] should be
administered as quickly as possible,
but only if there is systemic toxicity
from the venom; AV binds to and
neutralizes venom already present they do not reverse damage already
done. Benadryl can be administered to
those with minor reaction to the
antivenom itself.
A specific antivenom is always
preferable to the polyvalent antivenom
as it will be lower volume of
administration and there will be less
chance of serum sickness. Calling
ahead to the hospital is critical as these
antivenoms are hard to get.
Never pretest for antivenom allergy as
it will delay treatment, and potentially
precipitate anaphylaxis, sensitize the
patient and fails to predict anaphylaxis.
While the package insert of the antivenom suggests performing a skin test
to assess for an allergic reaction prior
to administering the full dose, this test
is poorly sensitive and specific for
identifying reaction and the venom
should be given regardless.
Nevertheless, in patients with a
positive skin test, corticosteroids, antihistamine and IV fluids should be given.
There is no good data to support the
use of tourniquets, surgical exploration,
antibiotics, warfarin or heparin in these
patients.
DIC is actually uncommon with croatilid
envenomation, so empiric treatment
with factors is not recommended.
There are probably over
2.5 million venomous snakebites in the
world every year and 125,000 deaths.
TRAUMA
BASIC HEMODYNAMICS
188
Class I shock in trauma is less than 750 mL of

blood loss or 15% of blood volume, class II is a
loss of 750 cc to 1.5 L of blood loss [15-30%],
class III shock is 1.5-2L of blood loss [or 30-40%]

Notes 2014
and class IV shock is more
than 2 L of blood loss
or more than 40% of TBV.
There is an increase in
heart rate from class II
shock and above. Blood
pressure tends to be normal
until class III and IV. Urine
output trends heart rate,
that is tends to drop with
class II and above. In other
words, in class I shock,
heart rate, BP and UOP are
all unchanged. Note that a
normal urine output can be
due to ethanol, DI,
hyperglycemic,
hypothermia in the trauma
patient. A 70kg man has
about 5 L of intravascular
volume.
The SVR is mathematically
coupled to the MAP, CVP
and cardiac output. The
SVR is MAP CVP over the
cardiac output multiplied by
80. The normal oxygen
delivery is 600 and the
normal oxygen utilization of
150 which is an O2 ER of
about 25%. When oxygen
important when giving these fluids, but
there is
controversy about warm fluids and
head injury.
In 1994, hypovolemic resuscitation was
tested. 600 patients were randomized.
The difference in treatment was the
pre-hospital and ER setting where one
group received about 900 cc of fluid and
the intervention group got about 100 [in
the ER 1.6L versus 300 cc additional
fluids, respectively]. Then in the OR, the
patients were given fluids equally
[about 7 additional liters].
The group that received less field and
ED fluid had diminished hospital stay
189
delivery is poor, as in hypovolemia, the
extraction ratio increases. The
adrenergic response increases the
mathematically-coupled SVR
calculation.
ATLS
The approach to the trauma patient is
to secure the airway with intubation,
RSI, and neck immobilization. The
patients need to be oxygenated above
90%, the hemorrhage must be stopped.
In the last 10 years [Iraq and
Afghanistan conflicts], tourniquets
should be used they can be lifesaving, quick-clot [kaolinite based
products]. You can lose half of your
circulating blood volume into an
unstable pelvis.
Interventional radiology can stop
bleeding! Appropriate access is
important in trauma, large bore IVs,
intra-osseous approach, cordisintroducers. If you put in two 9 French
cordis introducers, in two minutes, 5
liters of crystalloid can be infused! In
2.5 minutes, you can give 5 units of
blood. Intra-osseous infusers can give
large flow as well. Warmed infusion is
very
and improved mortality. The moral of the story is
to give less fluids until bleeding has stopped.
There is also controversy between LR and NS. NS is
only preferred in head trauma and head trauma
only. Albumin stays around in the body for about 20
days, it stays almost entirely intra-vascular for
many hours. It accounts for 80% of plasma colloid
pressure. Overall the difference between colloids
and crystalloids is small in nearly all trauma
patients. There is no difference between colloid and
crystalloid in terms of pulmonary edema, length of
stay, mortality, & pneumonia.
ANEMIA IN TRAUMATIC SHOCK
The decreased viscosity improves cardiac output
and stroke volume. There is an increase in oxygen

Notes 2014
extraction. In JWs there were no deaths
at Hb above 5 and there is no effect of
transfusion on mortality in patients with
hip fracture. Anemia may not be a bad
thing in trauma. Our transfusion
triggers should be lower than we
probably think. Certainly in the stable
patient [TRICC trial], a Hb of 7 is OK
[even in the actively bleeding GI
patients [NEJM January 2013]].
The risk of blood-borne infection is low
[one in 500,000 for a fatal hemolytic
reaction, HIV], [one in 100,000 for HCV
or HBV], [one in 5,000 for ALI]. Blood is
also an immunosuppressant and was
190
given prior to early renal transplants to
reduce
the immune response.
Blood can be auto-transfused [cellsaver] if large amounts are lost [e.g. in
a massive hemothorax] but is not used
when there is risk of contamination [e.g.
bowel perforation].
Early in the massive transfusion protocol
it was argued for 1:1:1 transfusion
[whole blood], this then dropped to 2:1:1
and 4:1:1 in terms of blood units to
plasma to platelets. The pendulum is
swinging back now to whole blood.
CHEST TRAUMA
Commotio cordis is when a nonpenetrating trauma to the thorax

induces ventricular fibrillation.
Only about 15% of thoracic injuries
require a definitive operation. 85%
require an intervention to improve
circulation, oxygenation, etc.
The main causes of hypoxemia in chest
trauma patients are: hypovolemia,
perfusion of an unventilated lung,
ventilation of an un-perfused lung and
abnormal pleural and airway
relationships.
PULMONARY CONTUSION
Blunt chest injury results in pulmonary
contusion often. There is loss of
alveolar-capillary membrane integrity
and this results in focal pulmonary
edema with an interstitial infiltrate. But
it is not blood in the alveolar space nor
is it blood in the pleural space. There is
a crush and recoil tearing/shearing
effect of the lung. The secondary
shearing effect tends to cause the most
problems.
The treatment of contusion is to give
oxygen, intubate, conservative fluids
and analgesia because of rib fractures.

Notes 2014
RIB FRACTURES
Where the rib fractures are can tell you
about the
injury. The higher ribs [1-3] usually
represent a high kinetic energy
mechanism there may be great vessel
injury here. If the ribs 4-8 are injured
then there may be pulmonary injury
bone spicules can puncture the lung and
injure intercostal muscles. If 9-12 are
injured, think splenic and hepatic injury.
Pain can lead to splinting, atelectasis,
impaired secretion clearance and
pneumonia. Early mobilization is
important as is spinal anesthesia with
opiates. NSIADs can also be helpful. In
fact, in patients with thoracic trauma
and rib fractures, the use of epidural
analgesia reduces days on mechanical
ventilation and the risk of pneumonia.
Mortality in elderly patients with rib
fractures is double that of young
patients with rib fractures.
Flail chest results in paradoxical
respirations such that spontaneous
inspiration sucks the segment inwards,
and exhalation pushes the flail portion
outwards. This leads to impaired V/Q
matching of the respiratory pump.
PNEUMOTHORAX & HEMOTHORAX
Pneumothorax is often the result of a rib
fracture. When the collapse becomes
too big, the treatment is with a chest
tube. Traumatic pneumothoraces
should not be treated with needles or
small catheters because this will not
evacuate blood; there should be suction
when treating traumatic pneumothorax.
Nitrogen can be absorbed from air by
giving high FiO2. Air [nitrogen]
anywhere in the body where it shouldnt
be is treated with 100% oxygen.
Tension pneumothorax is an emergency,
if the patient is blue you are in trouble
because at that point its nearly too late
to sustain life.
191
A persistent pneumothorax should prompt
consideration for a ruptured bronchus. Even if there
are broken ribs and flail chest, persistent
pneumothorax with chest tube in place can certainly
be a consequence of ruptured airways.

Notes 2014
If unrecognized, the patient
may require
resection though they can
usually be repaired.
Hemothorax looks like a
white out of the lung, it will
be lung and without breath
sounds. If 2 L of blood is in
the chest, this is a massive
hemothorax. This should
prompt a look for a
vascular injury. Autotransfusion [cell saver] can
be important.
All of the blood must be
removed to prevent a
fibrothorax.
INJURY OF THE HEART AND
GREAT VESSELS
Myocardial contusion can
also occur. It is a focal
region of myocardial
bruising and the right
ventricle is most commonly
affected. It may present
with atrial arrhythmias. The
diagnosis is difficult. Many
cardiac contusions are
admitted and monitored,
mediastinum away. It is possible that
this is
hemothorax or effusion, but also
torsion of the lung; this occurs when
the lung twists around its vascular
pedicle. This, as you can imagine, can
be a catastrophe with profound
hemodynamic and gas-exchange
abnormalities. Usually de-torsion
results in ischemia reperfusion injury of
the lung with showering of emboli to
the left atrium and systemic circulation.
Patients may have a spontaneous detorsion of the lung, but this may lead to
multiple embolic events to the brain.
but there is generally little

complication.
192
IF the patient has pump failure, there is

some sort of infarction going on and
these patients can get sick quickly.
Traumatic aortic rupture is highly
mortal. It is the most common cause of
death in a fall from great height or very
rapid deceleration; it occurs at the
ligamentum arteriosum. With this
mechanism, a wide mediastinum on
CXR can be diagnostic of aortic rupture.
Contrast CT scan is the diagnostic
procedure of choice. Many of these
patients are being treated with
endovascular grafts now.
INJURY OF THE MEDIASTINUM, DIAPHRAGM &
ABDOMEN
Tracheobronchial injuries are diagnosed
by bronchoscopy. Esophageal rupture
can be diagnosed with endoscopy,
barium, thoracostomy.
Recognize torsion of the lung. Consider a
patient who had been shot in the chest
and had a portion of his left lower lobe
removed in the OR. Then there is
excessive blood in the chest tube and the
entire left lung is opacified with some
shift of the
Diaphragmatic injuries may result in the stomach
entering the thorax. Positive pressure ventilation
can sometimes reduce this injury and thus with
extubation the injury can be made worse.
Abdominal trauma can also complicate chest
trauma as well as pelvic injury. Grey Turners sign
suggests RP bleeding, this could be an aortic
injury or pancreatic injury. Decompress the
stomach with an NG tube, but do not do so with
head injury. The NG tube can get into the brain,
which is problematic.
Liver injuries are complex because of its dual
blood supply and its outflow. The liver can be
totally avulsed which is usually fatal, but
transplants have been done. Pelvic injuries can

Notes 2014
bleed heavily, there can be damage of
the GI and GU tract from pelvic injury.
BURN INJURY
In burn injury, there is smoke

inhalation which is the leading cause of
death in burn injury and predicts death
better than patient age, or extent of
burn.
SMOKE INHALATION
The mortality rate of smoke injury
[isolated] is about 10%. Smoke
inhalation carries with it the thermal
injury, asphyxiation and
toxic/carbonaceous injury of the
airways.
193
The thermal injury tends to be minimal
because
the nasopharynx mitigates temperature
change quite well in both cold and hot
situations. The injury from purely
thermal sources tends to be restricted
to the naso and oropharynx [imagine
swallowing scolding hot soup]. The
exception to this is steam which a great
capacity to carry heat. Upper airway
obstruction from thermal injury is about
20-30% in prevalence. Such patients
are usually tripoding and drooling.
The asphyxiants from smoke inhibit
cellular respiration. For example
carbon monoxide [CO] toxicity.
Weakness, seizures and coma can be a
symptoms of CO. Standard pulse
oximetry cannot make the diagnosis of
carbon monoxide poisoning. Cooximetry must be used, because the
CO level can be toxic despite a normal
saturation on the pulse oximeter. A
normal CO- Hb is less than 5% but can
be less than 10% in smokers and
truckers.
Carbon monoxide level at the time of a
fire is difficult to predict, but is
important. The half-life of carbon
monoxide on air is about 4 hours [250
minutes], on FiO2 of 1.0, this half-life
drops to one hour. If the CO level is
5%, 3 hours following a fire while on
FiO2 of 1.0, the initial exposure may
have been very high [i.e. 40%].
There is no dose-response curve
between CO level and outcome. Less
than 5 minutes of exposure can get CO
levels very high [above 20%].
Hyperbaric oxygen is suggested, but
difficult to perform. At 2 atms, the halflife of CO-Hb is 27 minutes. But when
bringing a patient from 2 atmospheres
back to sea level, gas will expand and
there can be barotrauma. Hyperbaric
oxygen is typically reserved only for

Notes 2014
pure CO poisoning or if there
are profound symptoms,
high CO levels above 20%
and cardiovascular
instability despite 4-6 hours
What about cyanide poisoning? There
are many
toxic compounds in smoke. They come
from burned plastics, PVC [produces
cyanide gas]. CN poisons the electron
transport chain which shuts down the
Krebs cycle secondarily. This results in
anaerobic metabolism and the shunting
of pyruvate to lactate. Cyanide is an
asphyxiant despite high levels of PaO2
and oxygen delivery.
There are no good tests for cyanide
poisoning. You can get a venous blood
gas to look for a high PvO2 because
oxygen wont be consumed [or off
loaded] in the tissues. Look for the
trifecta of: metabolic acidosis, carbon
monoxide poisoning and a high venous
oxygen tension. These together should
highly suggest cyanide poisoning.
The treatment for CN toxicity has
evolved over the last 100 years.
Initially the Lilly Kit induced methemoglobinemia because Met-Hb will
preferentially bind CN and be
degraded in the liver, but this is
inducing another toxic hemoglobin
moiety [in addition to any carbon
monoxide poisoning that may be
involved] such that the patient may
have 20% Met-Hb, 20% carboxy-Hb
and therefore only 60% normal Hb. And
the pulse oximeter may read in the
mid-80s throughout.
The most recent kit is the cyanokit
which causes a brilliant red discoloration
of the skin and urine and this
discoloration can interfere with common
lab tests. The hydroxycobalamin directly
binds to the cyanide.
h
194
of normobaric oxygen. What is the
outcome with hyperbaric therapy?
There is no mortality benefit but at 6-8
weeks, it improves cognitive outcomes.
Lastly, smoke inhalation is mediated by
carbonaceous debris that is not directly related to
toxins or thermal injury within the lung. There is
about 48 hours of honeymoon before the endothelial
cells of the airways become denuded and the
airways fill with cellular debris.
Bronchoscopy is the best method to diagnose smoke
inhalation injury. There can be cobblestoning of the
airway. Soot in the sputum, singed nasal hairs and
facial burns are not as reliable as a bronchoscopy. A
bronchoscopy is

Notes 2014
positive only if there are
positive findings and will
identify twice as many
patients than solely
physical signs and
symptoms. The absence of
positive findings on a
bronchoscopy does not
mean that there was not
smoke inhalation.
Treatment involves
aggressive chest
physiotherapy and airway
clearance techniques. IPPV
and the VDR4 ventilators
may facilitate aboral
secretion movement.
Intermittent bronchoscopy
may also be needed.
BURN RESUSCITATION
What about burn
resuscitation? There is a
reduction in cardiac
performance in burn
patients despite fluid
resuscitation. It is a reaction
to profound cytokine
release as in sepsis. It tends
not to improve despite fluid
resuscitation.
195
The classic formula is the Parkland
formula. It uses 4 mL per kilogram per
% TBSA. Then one half of this is given
at the time of injury [in the first 8 hours]
and the other half during the following
16 hours. Usually Parkland is invoked
when TBSA is more than 20% because
that is the level of burn at which
capillary leak becomes pronounced. The
Parkland formula does not factor in
maintenance fluids for the first 24 hours
[see below].
What are the zones of burn? Necrosis
[dead], stasis [needs perfusion] and
hyperemia [high flow]. Burn
resuscitation is to improve oxygen
delivery to the zone of stasis. Over
resuscitation will increase interstitial
edema and increase the distance
between the capillary and the tissue
mass [i.e. increase the distance for
oxygen diffusion].
What is the preferred fluid in burn
resuscitation? The answer is LR. This will
prevent hyperchloremic metabolic
acidosis. Hypotonic fluid is particularly
bad given capillary leak syndrome.
What is the maintenance fluid for a burn
patient following resuscitation? You have
to take into consideration the
evaporative losses of patients
this occurs months later so succinylcholine is

probably OK very early on, but it is still
discouraged. Regardless, these patients often
have acute hyperkalemia which is a contraindication to succinylcholine.

with large burns. Daily ins and outs in
these patients are totally meaningless.
Basic maintenance fluid [for everyone]
is 1.5L x m [squared] [body surface
area], then you must add to that [25+
%TBSA] x m [squared] x 24 which is the
evaporative losses. So if an average
person requires maintenance of 125
mL/hour, if that person has a 60% burn,
there will be an additional 165 mL/hour
of evaporative losses from the burn.
Note there are also 800-1200 mL loss
per day being on a ventilator and
inhalation injury increases these losses.
ELECTRICAL BURNS
In these patients, if there is a normal
ECG, no LOC and no other indication for
admission, they can be discharged
from the ED.
What is the most common cardiac
dysrhythmia following electrical injury?
The answer is atrial fibrillation.
Following a cardiac contusion the most
common dysrhythmia is sinus
tachycardia. In electrical injury, the
cardiac events will occur early, so
admission does not typically need to
be prolonged. Note that both low and
high voltage current can result in Vfib.
Electrical injury burns patients from the
inside out. Bone results in very high
thermal injury because of its high
resistance. Muscle tissue is often
damaged severely in electrical injury
such that myoglobin and potassium are
released in large amounts. Therefore, in
electrical burns, the treatment is quite
similar to rhabdomyolysis. The pKa of
myoglobin is in the 4 range, so the
urine only needs to be mildly alkaline.
If such a patient requires intubation,
succinylcholine should not be used. In
a burn injury, there is a dramatic
increase in cholinergic receptors, but
lung.org
JE
K

Review Notes
prolong the Qt interval, so ECG

monitoring is important as well as serial
calcium and magnesium levels.
CHEMICAL BURNS
138
HYPO AND HYPERTHERMIA
The anterior hypothalamus is the

h
temperature regulation center of the
brain. It responds to heat by increasing
sweating, vasodilation and decreasing
muscle tone. To warm up, you
vasoconstrict, shiver and increase
muscle tone.
Acids cause coagulative

necrosis while bases give a
liquefactive necrosis.
Coagulative necrosis creates
a barrier such that the acid
cannot penetrate deeply.
Liquefactive necrosis, by
contrast, creates a milieu
whereby the base can
continue to erode and
destroy tissue deeply.
MANIFESTATIONS OF HYPOTHERMIA
There are 700 deaths per year in the
US. They are the urban destitute, but
also the wilderness and sports
enthusiasts. Hypothermia can also
happen in Texas. Hypothermia is a core
temperature less than 35 degrees
celsius. 28-32 is moderate, and severe
is less than 28 degrees [less than 82
farenheit]. These classifications are a
bit different in trauma.
Hydrofluoric acid produces a

profound hypocalcemia. It is
a calcium chelator. Calcium
gluconate can be applied
topically to stem the extent
of the burn injury, but not in
the hand; HF injury to the
hand requires intra-arterial
calcium infusion!
Adrenal insufficiency, pan

hypopituitarism and hypothyroidism all
predispose to hypothermia. Various
diseases and drugs can cause
dysregulation of temperature [e.g.
Parkinsons disease].
The hypocalcemia produced

by hydrofluoric acid can
www.heartlung.org
Patients who are profoundly hypothermic are also

profoundly hypovolemic. Shivering is exhausted
JE
K

Review Notes

Notes 2014
by 32 degrees celsius because the
patient has
used up glycogen. By 28 degrees there
is muscle rigidity and shock, by 24
degrees there is the appearance of
death, with minimal cardiac activity
and by 20 degrees there is an
isoelectric EEG and asystole.
Bradycardia, hypotension, increased
SVR as well as fibrillation of the atria
and ventricles can occur. J waves are
classically seen [aka Osborne wave],
and this is a small positive deflection
immediately following the QRS. It is not
pathognomonic for hypothermia, but it
should raise suspicion. It may only be
in some leads.
All organ systems are affected by
hypothermia. The respiratory
manifestations are that of decreased
Mve [both RR and Vt]. There is also a
cold diuresis which comes from shifting
blood volume to the central organs, so
the kidneys may sense hypervolemia
and cause a diuresis. There is also
confusion, lethargy, ileus and hepatic
dysfunction. There may be an
increased hematocrit, low platelets
[from sequestration], coagulopathy,
but with a normal PT or PTT because
the lab warms the blood. Platelet
function is slowed by low temperature.
Electrolytes can be variable. Glucose
levels tend to rise unless its an
alcoholic. There is acidemia with no
need to correct the ABG for
temperature [remember an ABG is put
on ICE].
MANAGEMENT OF HYPOTHERMIA
If intubation is needed, try not to
nasotracheally intubate these patients
because of bleeding. The
neuromuscular blockers tend not to
work as well in hypothermia. Pulse
checks during ACLS should be longer
than 10 seconds. Bradycardia is not
9
h
198
treated, Vfib should be treated with one
defibrillation and if unsuccessful, the patient should
be rewarmed to more than 30 degrees and then reattempted. Similarly, ACLS drugs will be minimally
effective below 30 degrees celsius.

Notes 2014
Chest compressions and
re-warming are the
focus.
How does one re-warm?
There are no rigid
protocols, but it depends
on the clinical situation.
Passive external
rewarming is essentially
insulation, but it is the
least invasive. It is a warm
environment with
protective clothing. The
patient must be able to
generate heat. It is good
for mild hypothermia [3235 C]. Active rewarming is
the application of external
heat such as lamps,
immersion, bare-hugger,
etc.
These are better for
moderate hypothermia
[28- 32 C] and can raise
the temperature 1.5-2.5
degrees celsius per hour.
Slower re-warming may be
preferred, dont feel a
need to rapidly rewarm the
patient.
There is no one factor or combination
of factors
that portend good or bad outcome in
drowning cases, thus a prognostic
scoring system is unavailable. Dry
drowning is uncommon. At the outset
of drowning, laryngospasm occurs and
prevents water from entering the lungs,
but this reflex abates in over 85% of
people and large volumes of water are
aspirated into the lungs.
Early reports of near drowning
commented upon differences in fluid
and electrolyte shifts based on fresh
water versus salt water drowning, but
later reports have not confirmed this.
9
h
199
Core re-warming is the application of
heat to the core of the body from noninvasive to very invasive methods.
Heated, humidified oxygen, heated IV
fluids, gastric, bladder, rectal massage,
peritoneal lavage, pleural lavage and
endovascular rewarming [case reports]
are all such means. Clinical pearl: you
can microwave a bag of NS for one
minute and shake the bag. The tubing
should be short to minimize heat loss.
Lavaging the stomach can increase the
risk of aspiration. Pleural lavage is
probably the fastest means.
Extracorporeal techniques can be done
such as CVVH with warm water as the
counter current in the filter. These
techniques are reserved for severe
hypothermia.
Complications of rewarming are
pulmonary edema, coagulopathy,
rhabdomyolysis, compartment
syndrome, ATN. There are no good
predictors of death. Continue
resuscitation until 32 degrees celsius.
Individualize termination of
resuscitation. There are case reports of
13.7 degree celsius patients with 4
hour resuscitations still surviving!
NEAR DROWNING
Know that aeromonas species are commonly
inhaled during near drowning in fresh water, but
also in salt water near drowning. 70% of patients
with aeromonas inhalation go on to suffer
bacteremia and there is a high mortality rate.
Aeromonas is susceptible to third generation
cephalosporins and beyond.
Francisella philomiragia has been rarely reported
during salt water near drowning [never
freshwater]; interestingly, klebsiella pneumoniae is
commonly a cause of pneumonia following sea
water inhalation.
Pseduoallescheria bodyii is a ubiquitous fungus that
is found in most dirty, stagnant water beds and
should be considered in any patient with
pneuomonia following drowning with consequent

Notes 2014
neurological abnormalities such as brain
abscess. Intra-ventricular miconazole is
the treatment for P bodyii in the CNS!
The cardiac rhythm of drowning is
bradycardia and electromechanical
dissociation. Vfib is rare unless its cold
out.
MANIFESTATIONS OF HYPERTHERMIA
Classic heat stroke is typically the
elderly with chronic illness, it develops
over days with dehydration. There may
be more drowsiness, altered mental
status and then collapse.
200
Exertional heat stroke occurs sporadically in
athletes and military recruits. There is
milder dehydration, but this type of
heat stroke occurs very quickly. This
does not occur in professional athletes
because they are well taken care of.
Medication and environmental heat can
interact. When humidity is high, the
sweat does not evaporate so the body
cannot lose heat. Sweat that drips is
not doing its job. Patients on anticholinergic agents such as antihistamines, anti- psychotics, etc. or
those with cardiovascular disease and
the inability to vasodilate can all be at
risk for hyperthermia.
Generally, the patients hospital
temperature on presentation does not
match the exposure temperature. The
hallmark is CNS dysfunction such as
AMS, seizures, cerebellar abnormalities.
Patients are tachycardic, hypotensive
and tachypneic.
Hyperthermia results in electrolyte
abnormalities similar to rhabdomyolysis
with renal insufficiency, respiratory
alkalosis, and lactic acidosis. There is
also coagulopathy, hepatic dysfunction
and variable electrolytes.
MANAGEMENT OF HYPERTHERMIA
Evaporative cooling can reverse
vasodilatation and hypotension seen in
many patients, so aggressive fluid
resuscitation is not generally needed
[especially in older, heart failure
patients]. Do not flood the patients,
cooling will increase blood pressure!
Conductive cooling is achieved by
placing ice packs in the groin, axilla,
and neck [near great vessels]; further,
cold IV fluids can be helpful [4 degrees
celsius NS]. But the production of
shivering can make things worse in this
situation.

Notes 2014
Evaporative cooling is warm
water mist with fans. A fan
with spray bottles over a
201
naked patient can rapidly resolve body
temperatures even above 108 degrees.

Notes 2014
The patient should be cooled to 102 degrees. The
absolute temperature does not predict
outcome but rather the exposure and
duration of hyperthermia. The elderly
with lactic acidosis, renal failure and
coma [i.e. classic heat stroke] typically
have the worst outcome.
14
h

2014 CCM Review Notes

Uploaded by

Copyright:

Available Formats

2014 CCM Review Notes

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2014 CCM Review Notes

Uploaded by

Copyright:

Available Formats

CRITICAL

Critical Care Board Study

I compiled these notes while studying for the ABIM

Critical Care Medicine

Critical Care Board Study

The quality of CPR is associated with

comatose, and not in shock. Cooling was

Temperature should be maintained at

Critical Care Board Study

prior to the arrest was a major discriminator

CONGESTIVE HEART FAILURE

About 50% of the mortality with heart failure

Critical Care Board Study Notes 2014

bypass surgery [versus medical therapy] and there

Recognize Takotsubo cardiomyopathy

Critical Care Medicine

of symptoms high dose is double their

exercise tolerance, cardiac performance

What about vasodilators? V-HEFT

Nesiritide is a formulation of BNP which

With myocarditis, patients

ACE inhibitors do everything!

Critical Care Medicine

Critical Care Board Study

Critical Care Board Study

flow VAD, however, is better per

Heart transplant? Hearts are limited

ST-ELEVATION MYOCARDIAL INFARCTI

With STEMI, about 50% of people just

What are the clinical events associated

Critical Care Board Study

Critical Care Board Study

Critical Care Board Study

to performing primary PCI is more than

If someone gets fibrinolysis, what is the

Critical Care Board Study

depressed RV function on TTE. RV

Critical Care Board Study

Other considerations are: VSD which

Critical Care Board Study

In each of the mechanical

NON-ST SEGMENT ELEVATION MYOCARDIAL INFARCTION

Ischemic chest pain may involve an

Consider an NSTEMI patient on dialysis

Critical Care Board Study

Critical Care Board Study

BRADYCARDIAS & PACEMAKERS

Critical Care Board Study

Critical Care Board Study

Critical Care Board Study

SVT is usually 100-260 bpm. It can be

There are 3 mechanisms of tachyarrhythmias for both the atria and

Critical Care Board Study

exacerbations of OVD. Lower doses of

MULTIFOCAL ATRIAL TACHYCARDIA

Critical Care Board Study

Critical Care Board Study

When using a drug or shock to

AF with WPW is badness,