Research

JAMA | Original Investigation | CARING FOR THE CRITICALLY ILL PATIENT

Effect of Hydrocortisone on Development of Shock

Among Patients With Severe Sepsis
The HYPRESS Randomized Clinical Trial
Didier Keh, MD; Evelyn Trips; Gernot Marx, MD; Stefan P. Wirtz, MD; Emad Abduljawwad, MD; Sven Bercker, MD; Holger Bogatsch, MD;
Josef Briegel, MD; Christoph Engel, MD; Herwig Gerlach, MD, PhD, MBA; Anton Goldmann, MD; Sven-Olaf Kuhn, MD; Lars Hüter, MD;
Andreas Meier-Hellmann, MD; Axel Nierhaus, MD; Stefan Kluge, MD; Josefa Lehmke, MD; Markus Loeffler, MD; Michael Oppert, MD;
Kerstin Resener, MD; Dirk Schädler, MD; Tobias Schuerholz, MD; Philipp Simon, MD; Norbert Weiler, MD; Andreas Weyland, MD;
Konrad Reinhart, MD; Frank M. Brunkhorst, MD; for the SepNet–Critical Care Trials Group

Editorial page 1769

IMPORTANCE Adjunctive hydrocortisone therapy is suggested by the Surviving Sepsis Supplemental content
Campaign in refractory septic shock only. The efficacy of hydrocortisone in patients with
severe sepsis without shock remains controversial. CME Quiz at
jamanetworkcme.com and
CME Questions page 1822
OBJECTIVE To determine whether hydrocortisone therapy in patients with severe sepsis
prevents the development of septic shock.

DESIGN, SETTING, AND PARTICIPANTS Double-blind, randomized clinical trial conducted from
January 13, 2009, to August 27, 2013, with a follow-up of 180 days until February 23, 2014.
The trial was performed in 34 intermediate or intensive care units of university and
community hospitals in Germany, and it included 380 adult patients with severe sepsis who
were not in septic shock.

INTERVENTIONS Patients were randomly allocated 1:1 either to receive a continuous infusion
of 200 mg of hydrocortisone for 5 days followed by dose tapering until day 11 (n = 190)
or to receive placebo (n = 190).

MAIN OUTCOMES AND MEASURES The primary outcome was development of septic shock
within 14 days. Secondary outcomes were time until septic shock, mortality in the intensive
care unit or hospital, survival up to 180 days, and assessment of secondary infections,
weaning failure, muscle weakness, and hyperglycemia (blood glucose level >150 mg/dL
[to convert to millimoles per liter, multiply by 0.0555]).

RESULTS The intention-to-treat population consisted of 353 patients (64.9% male; mean
[SD] age, 65.0 [14.4] years). Septic shock occurred in 36 of 170 patients (21.2%) in the
hydrocortisone group and 39 of 170 patients (22.9%) in the placebo group (difference,
−1.8%; 95% CI, −10.7% to 7.2%; P = .70). No significant differences were observed between
the hydrocortisone and placebo groups for time until septic shock; mortality in the intensive
care unit or in the hospital; or mortality at 28 days (15 of 171 patients [8.8%] vs 14 of 170
patients [8.2%], respectively; difference, 0.5%; 95% CI, −5.6% to 6.7%; P = .86), 90 days Author Affiliations: Author
(34 of 171 patients [19.9%] vs 28 of 168 patients [16.7%]; difference, 3.2%; 95% CI, −5.1% to affiliations are listed at the end of this
11.4%; P = .44), and 180 days (45 of 168 patients [26.8%] vs 37 of 167 patients [22.2%], article.
respectively; difference, 4.6%; 95% CI, −4.6% to 13.7%; P = .32). In the hydrocortisone vs Group Information: TheSepNet–Critical
placebo groups, 21.5% vs 16.9% had secondary infections, 8.6% vs 8.5% had weaning failure, Care Trials Group members are listed
in eAppendix 1 in Supplement 1.
30.7% vs 23.8% had muscle weakness, and 90.9% vs 81.5% had hyperglycemia.
Corresponding Author: DidierKeh,MD,
Department of Anesthesiology
CONCLUSIONS AND RELEVANCE Among adults with severe sepsis not in septic shock, use of and Intensive Care Medicine,
hydrocortisone compared with placebo did not reduce the risk of septic shock within 14 days. Campus Virchow-Klinikum
and Campus Charité Mitte,
These findings do not support the use of hydrocortisone in these patients.
Charité–Universitätsmedizin Berlin,
Augustenburger Platz 1, 13353 Berlin,
TRIAL REGISTRATION clinicaltrials.gov Identifier: NCT00670254 Germany ([email protected]).
Section Editor: Derek C. Angus, MD,
JAMA. 2016;316(17):1775-1785. doi:10.1001/jama.2016.14799 MPH, Associate Editor, JAMA
Published online October 3, 2016. ([email protected]).

(Reprinted) 1775

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Research Original Investigation Hydrocortisone and Septic Shock in Patients With Severe Sepsis

D
espite decades of study and debate, the role of
adjunctive “low-dose” hydrocortisone treatment Key Points
(200-300 mg/d) in patients with severe sepsis and sep-
Question Does adjunctive early hydrocortisone therapy prevent
tic shock remains controversial.1 The current recommenda- the development of septic shock in patients with severe sepsis
tion for hydrocortisone use is mainly based on 2 randomized who are not in shock?
clinical trials (RCTs).2 In the study by Annane et al,3 hydro-
Findings In this randomized clinical trial that included 380 adults,
cortisone improved survival and reversal of septic shock in pa-
occurrence of septic shock was not significantly different between
tients with relative adrenal insufficiency. In the CORTICUS patients who received hydrocortisone or placebo (21.2% vs 22.9%,
study,4 septic shock was reversed more quickly but mortality respectively).
was not significantly reduced. The higher risk of mortality and
Meaning Administration of hydrocortisone did not prevent the
septic shock severity in the study by Annane and colleagues
development of shock in patients with severe sepsis.
resulted in more restrictive recommendations for hydrocor-
tisone use only in patients with inadequate response to fluid
and vasopressor resuscitation.2 However, septic shock rever- tients, patient-authorized representatives, or legal represen-
sal in the CORTICUS study was reported to be significantly ac- tatives. Patients were enrolled if they met all inclusion criteria
celerated by the administration of hydrocortisone irrespec- (for details, see eAppendix 2 in Supplement 1): (1) provided in-
tive of the adrenal response to corticotropin. An international formed consent; (2) had evidence of infection; (3) had evi-
consensus statement recommended replacing the terms rela- dence of a systemic response to infection, defined as at least 2
tive or absolute adrenal insufficiency, which reflect only adre- systemic inflammatory response syndrome criteria12; and
nal cortisol release, by the critical illness–related corticoste- (4) had evidence of organ dysfunction present for not longer
roid insufficiency (CIRCI) concept.5 than 48 hours. The main exclusion criterion was septic shock.
Although meta-analyses report controversial results on mor- Other exclusion criteria were being younger than 18 years, hav-
tality reduction by administration of corticosteroids,6,7 there is ing known hypersensitivity to hydrocortisone or mannitol
consistency regarding shock reversal irrespective of disease se- (placebo), or having a history of glucocorticoid medication with
verity or the presence of CIRCI.5-7 In a smaller RCT in patients indication for continuation of therapy or other indications for
with community-acquired pneumonia (CAP), hydrocortisone treatment with glucocorticoids. Patients were not excluded for
significantly improved survival and prevented progression to using etomidate within 72 hours before enrollment, using a
shock.8 Furthermore, 2 recent RCTs9,10 and a meta-analysis11 re- short course of glucocorticoids within 72 hours before enroll-
vealed positive effects of steroids in patients with CAP. Assum- ment, or using topical or inhaled glucocorticoids.
ing that severe sepsis and septic shock reflect a disease con-
tinuum, it was hypothesized that early hydrocortisone Definitions
administration might prevent shock development owing to the Septic shock was defined as sepsis-induced hypotension de-
attenuation of an exaggerated inflammatory response. To our spite adequate volume status for longer than 4 hours (ie, mean
knowledge, this is the first RCT investigating the effects of hy- arterial pressure <65 mm Hg, systolic arterial pressure
drocortisone to prevent progression to shock in patients with <90 mm Hg, or the use of vasopressors to keep mean arterial
severe sepsis presenting without shock. pressure ≥65 mm Hg or systolic arterial pressure ≥90 mm Hg).
Patients who had a transient need for vasopressors during ini-
tial resuscitation but were not hypotensive and did not use va-
sopressors for at least 2 hours were eligible for enrollment when
Methods septic shock was not present at the time of randomization. Ad-
Study Design equate volume status was defined as a central venous pressure
The Hydrocortisone for Prevention of Septic Shock (HYPRESS) of 8 mm Hg or greater (≥12 mm Hg in ventilated patients) and a
study is an investigator-initiated, multicenter, placebo- central venous oxygen saturation greater than 70%. For fluid re-
controlled, double-blind RCT supported by the German Fed- placement, patients were to receive at least 500 to 1000 mL of
eral Ministry of Education and Research. The study was con- crystalloids or 300 to 500 mL of colloids over 30 minutes. The
ducted in cooperation with the German Sepsis Competence use of hydroxyethyl starch preparations was discouraged ow-
Network (SepNet) and the Clinical Trial Centre Leipzig, which ing to possible harmful effects on kidney function.13,14 Use of
provided internet-based randomization and data capture, as- vasopressors was defined as therapy with dopamine at a dos-
surance of accuracy and completeness of data, biostatistical age of at least 5 μg/kg/min or with any dose of epinephrine, nor-
analysis, and pharmacovigilance. Monitoring of sites was per- epinephrine, vasopressin, or other vasopressors. For further
formed on defined intervals. The protocol was approved by the details and definitions, see eAppendix 2 in Supplement 1.
responsible ethics committees of all 34 participating sites. The
trial protocol is available in Supplement 2. Randomization
Randomization was stratified by participating center and sex.
Study Patients It was performed with an internet-based computerized ran-
Patients were screened in intermediate care units or intensive domization tool that uses a modified version of the Pocock
care units (ICUs) of university and community hospitals for eli- minimization algorithm15 with a random component to gen-
gibility, and written informed consent was obtained from pa- erate balanced 1:1 randomization in the strata at any time. All

1776 JAMA November 1, 2016 Volume 316, Number 17 (Reprinted) jama.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Hydrocortisone and Septic Shock in Patients With Severe Sepsis Original Investigation Research

patients, study personnel, sponsors, medical staff, and nurs- Data Acquisition, Cortisol Measurement,
ing staff were blinded regarding the allocation of study medi- and Patient Treatment
cation throughout the entire study period. Disease severity was assessed by the SOFA score,20 Simpli-
fied Acute Physiology Score II (ranging from 0-163),21 Acute
Study Medication Physiology and Chronic Health Evaluation II (ranging from
The study medication (hydrocortisone and placebo) was pro- 0-71),22 and Simplified Acute Physiology Score 3 (ranging from
duced and released by BAG Health Care GmbH. The medica- 0-217) (with higher scores indicating greater severity on all
tion was delivered in boxes, each containing 17 brown glass vi- instruments).23 Serum cortisol concentration was batch ana-
als for 1 patient. Each vial contained 100 mg of lyophilized lyzed by mass spectrometry in a reference laboratory in blood
hydrocortisone hydrogen succinate or the same amount of ly- samples stored at −80°C and taken before and 60 minutes af-
ophilized mannitol as placebo, which was indistinguishable ter administration of 250 μg of corticotropin at baseline. All
from hydrocortisone. The medication was administered as an patients had to be treated according to guidelines of the
intravenous bolus of 50 mg, followed by a 24-hour continu- German Sepsis Society13 (eAppendix 2 in Supplement 1).
ous infusion of 200 mg on 5 days, 100 mg on days 6 and 7,
50 mg on days 8 and 9, and 25 mg on days 10 and 11. Hydro- Statistical Analysis
cortisone dose corresponded to those used in the 2 major RCTs The study was planned to detect an absolute difference of 15%
performed earlier,3,4 which had shown significant effects on in the proportion of patients with septic shock within 14 days
septic shock resolution. A continuous infusion was preferred with a significance level of .05 and a power of 0.8. The differ-
to avoid possible undulation of the blood cortisol concentra- ence of 15% was postulated to be a meaningful difference that
tions by bolus administration, which had been reported to com- could change clinical practice, and was supported by similar
plicate blood glucose control.16 A continuous infusion was also differences of hydrocortisone on 7-day septic shock resolu-
recommended in the recent Surviving Sepsis Campaign tion in patients with septic shock,24 assuming that hydrocor-
guidelines.2 To reduce possible hemodynamic and immuno- tisone might be as effective in preventing septic shock as in
logical rebound effects,17 hydrocortisone was tapered over sev- resolving it.
eral days as in the CORTICUS study.4 Study medication was dis- Assuming 40% of patients in the placebo group had septic
continued for safety reasons or when patients were discharged shock,8,25-27 169 evaluable patients per arm were required. Ac-
from the ICU or reached the primary end point. counting for an expected dropout rate of about 10%, 190 pa-
tients per arm (380 in total) had to be included. The statistical
End Points analysis was conducted consistent with the intention-to-
The primary end point was the occurrence of septic shock treat (ITT) principle. The primary end point was assessed by χ2
within 14 days, which was assessed daily until day 14, or dis- test; heterogeneity between centers with more than 10 re-
charge from the ICU. Secondary end points were time until cruited patients was assessed by I2. Secondary end points were
development of septic shock or death (whichever came first), analyzed by χ2 test, Fisher exact test, t test, Mann-Whitney
mortality in the ICU and hospital, vital status at 28, 90, and U test, or log-rank test, as appropriate. All reported P values are
180 days, duration of stay in the ICU and hospital, organ dys- 2-sided, and P < .05 was regarded as statistically significant. Sec-
functions (Sequential Organ Failure Assessment [SOFA] ondary analyses were not adjusted for multiple testing be-
score, ranging from 0-24 with higher values indicating cause these statistical comparisons were performed with ex-
greater severity), duration of mechanical ventilation, and ploratory rather than confirmatory intention. Planned subgroup
renal replacement therapy. In the subgroup of patients analyses included the ITT population and the per-protocol (PP)
who underwent a corticotropin test at baseline, the occur- population, administration of study medication for at least 48
rence of septic shock, mortality, length of stay (LOS) in the hours, medical and surgical patients, and patients with pneu-
ICU or hospital, mechanical ventilation, organ dysfunctions monia. Post hoc subgroup analyses were performed for CIRCI,
(SOFA score), renal replacement therapy, and secondary delirium, and CAP. A multivariable logistic regression model was
infection were evaluated. Critical illness–related corticoste- performed to investigate adjusted treatment effects. Two in-
roid insufficiency was defined as an increase of cortisol of terim analyses were performed after recruitment of one-third
9 μg/dL or less (to convert to nanomoles per liter, multiply by and two-thirds of the planned sample size. Statistical analyses
27.588) 1 hour after stimulation with 250 μg of corticotropin were performed using SAS version 9.2 (SAS Institute Inc),
(Synacthen). Frequency of delirium was assessed daily until R version 3.1.0 (R Foundation), and SPSS version 22.0 (IBM Corp)
ICU discharge by the Richmond Agitation-Sedation Scale18,19 statistical software. For details of the statistical analyses, see
(ranging from −5 [unarousable; no response to voice or physi- eAppendix 2 in Supplement 1.
cal stimulation] to 4 [combative; overly combative or violent;
immediate danger to staff]) to quantify the level of sedation
and by the Confusion Assessment Method for the ICU for
detection of delirium. Adverse events were assessed until day
Results
28, with special emphasis on muscle weakness, weaning fail- Patient Population
ure, secondary infection, and gastrointestinal bleeding. All From January 13, 2009, to August 27, 2013, 9953 patients with
events not typically associated with the course of disease had severe sepsis or septic shock were screened at 34 study sites
to be reported (eAppendix 2 in Supplement 1). for eligibility. A total of 380 patients were randomized to receive

jama.com (Reprinted) JAMA November 1, 2016 Volume 316, Number 17 1777

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Research Original Investigation Hydrocortisone and Septic Shock in Patients With Severe Sepsis

Figure 1. CONSORT Flow Diagram

9953 Patients assessed for eligibility

9573 Excludeda
1847 Severe sepsis >48 h
632 No informed consent
8004 Did not meet eligibility criteriaa
6073 Preexisting septic shock
881 Preexisting glucocorticoid therapy
524 Current systemic glucocorticoid therapy
516 Moribund
303 Had do-not-resuscitate order
205 Concomitant or previous participation
in another interventional trial
101 Inclusion criteria not met
54 Immunosuppression
28 Aged <18 y
31 Pregnant or breastfeeding women
6 Relationship to investigator (eg, relatives,
colleagues, staff)
6 Known hypersensitivity to study
medication
324 Other reasons
137 Logistic
187 Other

380 Randomized

190 Randomized to receive placebo 190 Randomized to receive hydrocortisone

190 Received intervention as 190 Received intervention as
randomized randomized

9 Discontinued intervention 15 Discontinued intervention

2 Serious adverse event 3 Serious adverse event
4 Withdrew informed consent 4 Withdrew informed consent
2 Investigator’s decision 1 Refused informed consent
1 Other reason 3 Investigator’s decision
9 Lost to follow-up 4 Other reason
6 Lost to follow-up by day 28 9 Lost to follow-up
2 Lost to follow-up between 6 Lost to follow-up by day 28
days 29 and 90 3 Lost to follow-up between
1 Lost to follow-up between days 91 and 180
days 91 and 180

176 Included in primary intention-to- 177 Included in primary intention-to-

treat analysis treat analysis
170 Included in primary end 170 Included in primary end
point analysis point analysis
6 Excluded from primary end 7 Excluded from primary end
point analysis point analysis
4 Withdrew informed consent 6 Withdrew informed consent
before day 15 before day 15 a
Multiple reasons possible.
2 Missing values during follow-up 1 Missing values during follow-up b
until day 14 until day 14 The informed consent process
14 Excluded 13 Excluded could not be completed as defined
9 Finding in informed consent processb 7 Finding in informed consent processb in the protocol.
3 Informed consent refused 3 Informed consent refused c
Septic shock occurred soon after
2 Septic shock at inclusion 3 Did not receive study medicationc
randomization but before
administration of study medication.

hydrocortisone (n = 190) or placebo (n = 190). The median time the placebo group, 5 in the hydrocortisone group) received an
from screening to enrollment was 12.5 hours (interquartile increased dose (quotient of applied dose and expected dose
range [IQR], 6-21 hours) in the placebo group and 14 hours (IQR, >1.2). These patients were excluded from the PP analysis.
6.5-23 hours) in the hydrocortisone group (P = .49). The ITT Follow-up was conducted until February 23, 2014.
population excluded 27 patients and included 353 patients
(64.9% male; mean [SD] age, 65.0 [14.4] years) (Figure 1). The Baseline Characteristics and Treatment During Study
PP population consisted of 322 patients, and the safety analy- Treatment arms were comparable regarding age, type of ad-
sis set included 375 patients. Six patients (2 in the placebo mission to the ICU, severity of disease or organ dysfunction,
group, 4 in the hydrocortisone group) received a reduced dose use of glucocorticoids or etomidate within 72 hours before ran-
(<80% of total dose according to protocol). Ten patients (5 in domization, initial treatment and vital signs within 6 hours

1778 JAMA November 1, 2016 Volume 316, Number 17 (Reprinted) jama.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Hydrocortisone and Septic Shock in Patients With Severe Sepsis Original Investigation Research

Table 1. Baseline Characteristicsa

Placebo Hydrocortisone Total
Characteristic (n = 176) (n = 177) (N = 353)
Male, No./total No. (%) 111/176 (63.1) 118/177 (66.7) 229/353 (64.9)
Age, mean (SD), y 64.6 (14.6) 65.5 (14.2) 65.0 (14.4)
Type of admission, No./total No. (%)
Surgery, elective 42/176 (23.9) 27/176 (15.3) 69/352 (19.6)
Surgery, emergency 32/176 (18.2) 44/176 (25.0) 76/352 (21.6)
Nonsurgery, elective 4/176 (2.3) 5/176 (2.8) 9/352 (2.6)
Nonsurgery, emergency 98/176 (55.7) 100/176 (56.8) 198/352 (56.3)
SOFA score, mean (SD)b,c 6.2 (2.4) 6.4 (2.6) 6.3 (2.5)
APACHE II score, mean (SD)b,d 18.5 (6.0) 19.5 (6.9) 19.0 (6.5)
SAPS II score, mean (SD)b,e 52.2 (9.9) 56.1 (13.3) 54.1 (11.8)
SAPS 3 score, mean (SD)b,f 58.4 (11.0) 58.5 (11.9) 58.4 (11.4)
SIRS criteria, No./total No. (%)
Temperature ≤36°C or ≥38°C 141/176 (80.1) 129/177 (72.9) 270/353 (76.5)
Heart rate ≥90 beats/min 164/176 (93.2) 158/177 (89.3) 322/353 (91.2)
Tachypnea, hypocapnia, 157/176 (89.2) 145/176 (82.4) 302/352 (85.8)
or mechanical ventilation
Leukocytosis, leukopenia, or left shift 132/176 (75.0) 131/177 (74.0) 263/353 (74.5)
Organ dysfunction, No./total No. (%)
Central nervous system 47/176 (26.7) 41/175 (23.4) 88/351 (25.1)
Coagulation 26/176 (14.8) 35/177 (19.8) 61/353 (17.3)
Pulmonary 119/175 (68.0) 117/177 (66.1) 236/352 (67.0)
Renal 73/176 (41.5) 70/177 (39.5) 143/353 (40.5)
Microcirculatory 53/176 (30.1) 61/176 (34.7) 114/352 (32.4)
Source of infection, No./total No. (%)
Community 83/176 (47.2) 82/177 (46.3) 165/353 (46.7)
Nosocomial, ICU 52/176 (29.5) 41/177 (23.2) 93/353 (26.3)
Nosocomial, ward 41/176 (23.3) 54/177 (30.5) 95/353 (26.9)
Focus of primary infection,
No./total No. (%)
Known focus 146/176 (83.0) 149/176 (84.7) 295/352 (83.8)
Pneumonia 78/146 (53.4) 56/149 (37.6) 134/295 (45.4)
Respiratory tract, other 10/146 (6.8) 8/149 (5.4) 18/295 (6.1)
Thoracic 2/146 (1.4) 9/149 (6.0) 11/295 (3.7)
Gastrointestinal 7/146 (4.8) 10/149 (6.7) 17/295 (5.8)
Abbreviations: APACHE II, Acute
Intra-abdominal 24/146 (16.4) 37/149 (24.8) 61/295 (20.7)
Physiology and Chronic Health
Primary bacteremia 2/146 (1.4) 2/149 (1.3) 4/295 (1.4) Evaluation II; ICU, intensive care unit;
Bones or soft tissue 14/146 (9.6) 15/149 (10.1) 29/295 (9.8) SAPS II, Simplified Acute Physiology
Score II; SAPS 3, Simplified Acute
Surgical wound 3/146 (2.1) 4/149 (2.7) 7/295 (2.4)
Physiology Score 3; SIRS, systemic
Urogenital 21/146 (14.4) 24/149 (16.1) 45/295 (15.3) inflammatory response syndrome;
Catheter 3/146 (2.1) 6/149 (4.0) 9/295 (3.1) SOFA, Sequential Organ Failure
Assessment.
Medication within 72 h before
a
randomization For a list of concomitant diseases,
Intravenous glucocorticoids, 6/176 (3.4) 3/177 (1.7) 9/353 (2.5) see eTable 1 in Supplement 1.
No./total No. (%) b
Higher scores indicate greater
Hydrocortisone equivalent, 600 (392-1000) 200 (200-400) 400 (200-1000) disease severity.
median (range), mg c
Possible scores range from 0 to 24.
Etomidate, No. (%) d
Possible scores range from 0 to 71.
No./total No. (%) 11/176 (6.3) 12/176 (6.8) 23/352 (6.5) e
Possible scores range from 0 to 163.
Mean (SD), mg 33.0 (13.8) 23.8 (10.4) 28.2 (12.8) f
Possible scores range from 0 to 217.

after diagnosis of severe sepsis, administration of study medi- Primary End Point
cation, antibiotic therapy, and treatment characteristics dur- There was no significant difference in the proportion of sep-
ing the study. Pneumonia was slightly more frequent in pa- tic shock after 14 days between patients who received hydro-
tients who received placebo (Table 1 and eTables 1, 2, and 3 in cortisone or placebo in the ITT or PP population (Table 2).
Supplement 1). In the ITT population, septic shock occurred in 36 of 170

jama.com (Reprinted) JAMA November 1, 2016 Volume 316, Number 17 1779

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Research Original Investigation Hydrocortisone and Septic Shock in Patients With Severe Sepsis

Table 2. Primary and Secondary End Pointsa

Placebo Hydrocortisone Total
End Point (n = 176) (n = 177) (N = 353) P Value
Primary
Septic shock, No./total
No. (%) [95% CI]
ITT population 39/170 (22.9) 36/170 (21.2) 75/340 (22.1) .70
[17.2-30.0] [15.6-28.1] [17.9-26.9]
PP population 33/156 (21.2) 29/155 (18.7) 62/311 (19.9) .59
[15.4-28.4] [13.3-25.7] [15.8-24.8]
Secondary
Mortality, No./total
No. (%) [95% CI]
28 d 14/170 (8.2) 15/171 (8.8) 29/341 (8.5) .86
[5.0-13.4] [5.4-14.0] [6.0-12.0]
90 d 28/168 (16.7) 34/171 (19.9) 62/339 (18.3) .44
[11.8-23.0] [14.6-26.5] [14.5-22.8]
180 d 37/167 (22.2) 45/168 (26.8) 82/335 (24.5) .32
[16.5-29.0] [20.7-34.0] [20.2-29.4]
ICU 14/172 (8.1) 13/171 (7.6) 27/343 (7.9) .85
[4.9-13.2] [4.5-12.6] [5.5-11.2]
Hospital 22/172 (12.8) 23/171 (13.5) 45/343 (13.1) .86
[8.6-18.6] [9.1-19.4] [10.0-17.1] Abbreviations: ICU, intensive care
LOS, median (IQR), d unit; IQR, interquartile range;
ITT, intention-to-treat; LOS, length of
ICU 9 (6-17) 8 (5-15) 8 (5-16) .23 stay; MV, mechanical ventilation;
Hospital 25 (16-40) 26 (16-46) 26 (16-43) .36 PP, per-protocol; RRT, renal
Mechanical ventilation, 103/172 (59.9) 91/171 (53.2) 194/343 (56.6) .21 replacement therapy; SOFA,
No./total No. (%) [52.4-66.9] [45.8-60.5] [51.3-61.7] Sequential Organ Failure Assessment.
[95% CI] a
For rows including number/total
MV-free time, 5 (2-7) 4 (2-7) 4 (2-7) .34 number, the total number refers to
median (IQR), d the number of patients with valid
RRT, No./total 21/172 (12.2) 21/171 (12.3) 42/343 (12.2) .98 data. Data are missing in up to 5% in
No. (%) [95%CI] [8.1-17.9] [8.2-18.0] [9.2-16.1] the ITT population, except for SOFA
RRT-free time, 7 (4-14) 6 (4-12) 7 (4-13) .35 (data missing in 25%) and delirium
median (IQR), d (data missing in 43%).
SOFA score until day 14, 5.0 (3.5-6.8) 4.7 (3.5-6.5) 4.8 (3.5-6.6) .69 b
The SOFA score for each patient
median (IQR)b
was calculated by the sum of daily
Delirium, No./total 25/102 (24.5) 11/98 (11.2) 36/200 (18.0) .01 SOFA scores divided by the
No. (%) [95% CI] [17.2-33.7] [6.4-19.0] [13.3-23.9]
observation time.

−2.4%; 95% CI, −11.5% to 6.6%; P = .59). In addition, no sig-

Figure 2. Time to Septic Shock
nificant difference between the groups was observed regard-
0.25 ing time to septic shock development (Figure 2), or in time to
Placebo septic shock in those patients who developed septic shock
0.20 (eFigure 1 in Supplement 1). Subgroup analysis of medical or
Cumulative Probability

surgical patients, patients with pneumonia, or study medica-

of Septic Shock

Hydrocortisone
0.15
tion treatment for at least 48 hours did not reveal a benefit for
shock prevention (eFigure 2 in Supplement 1). To exclude a cen-
0.10
ter effect, 11 sites that recruited at least 10 patients (n = 279)
0.05
were analyzed; there was no heterogeneity for the primary end
Log-rank P = .69 point (I2 = 0%; P = .74).
0
0 2 4 6 8 10 12 14
Time After Randomization, d
Secondary End Points
No. at risk
There were no significant differences in 28-day, 90-day, 180-
Placebo 176 161 139 136 134 131 130 128 day, ICU, or hospital all-cause mortality; LOS in the ICU or hos-
Hydrocortisone 177 163 146 142 138 138 134 130
pital; ventilation- or renal replacement–free days; or median
SOFA score until day 14 between patients treated with pla-
Tick marks on curves indicate censored data.
cebo or hydrocortisone (Table 2 and eFigure 3 in Supplement
1). At 28 days, mortality occurred in 15 of 171 patients (8.8%)
patients (21.2%) in the hydrocortisone group vs 39 of 170 pa- in the hydrocortisone group and 14 of 170 patients (8.2%) in
tients (22.9%) in the placebo group (difference, −1.8%; 95% CI, the placebo group (difference, 0.5%; 95% CI, −5.6% to 6.7%;
−10.7% to 7.2%; P = .70). In the PP population, septic shock oc- P = .86). At 90 days, mortality occurred in 34 of 171 patients
curred in 29 of 155 patients (18.7%) in the hydrocortisone group (19.9%) in the hydrocortisone group vs 28 of 168 patients
vs 33 of 156 patients (21.2%) in the placebo group (difference, (16.7%) in the placebo group (difference, 3.2%; 95% CI, −5.1%

1780 JAMA November 1, 2016 Volume 316, Number 17 (Reprinted) jama.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Hydrocortisone and Septic Shock in Patients With Severe Sepsis Original Investigation Research

Table 3. Adverse Events in Safety Analysis Set

Placebo Hydrocortisone Total
Adverse Event (n = 189) (n = 186) (N = 375) P Valuea
Secondary infections, No. (%) 32 (16.9) 40 (21.5) 72 (19.2) .26
MRC Scale for Muscle Strength score 151 (79.9) 150 (80.6) 301 (80.3) .86
available, No. (%)
b
Muscle weakness, No. (%) 36 (23.8) 46 (30.7) 82 (27.2) .18
Respiratory, No. (%) 24 (12.7) 24 (12.9) 48 (12.8) .95
Weaning failure 16 (8.5) 16 (8.6) 32 (8.5) .96
Respiratory failure 7 (3.7) 3 (1.6) 10 (2.7) .34
Other 9 (4.8) 6 (3.2) 15 (4.0) .45
Cardiovascular, No. (%) 19 (10.1) 17 (9.1) 36 (9.6) .76
Arterial hypertension 1 (0.5) 5 (2.7) 6 (1.6) .12
Other 18 (9.5) 14 (7.5) 32 (8.5) .49
Abdominal, No. (%) 6 (3.2) 7 (3.8) 13 (3.5) .76
Gastrointestinal bleeding 2 (1.1) 3 (1.6) 5 (1.3) .68
Gastrointestinal ulcer 1 (0.5) 0 1 (0.3) .99
Other 4 (2.1) 7 (3.8) 11 (2.9) .35
Impaired wound healing, No. (%) 3 (1.6) 5 (2.7) 8 (2.1) .50
Central nervous system, No. (%) 9 (4.8) 8 (4.3) 17 (4.5) .83
Abbreviations: IQR, interquartile
Stroke, TIA, or convulsion 5 (2.6) 2 (1.1) 7 (1.9) .45
range; MRC, Medical Research
Delirium 4 (2.1) 5 (2.7) 9 (2.4) .75 Council; TIA, transient ischemic
Other 0 1 (0.5) 1 (0.3) .50 attack.
Hypernatremia, No. (%)c 10 (5.3) 10 (5.4) 20 (5.3) .97 SI conversion factors: To convert
sodium to millimoles per liter,
Maximum sodium concentration, 141 (6) 141 (5) 141 (6) .29
mean (SD), mEq/L multiply by 1.0; glucose to millimoles
per liter, multiply by 0.0555.
Sodium concentration during study 140 (6) 141 (5) 141 (6) .15
a
medication administration, Calculated by χ2 test, Fisher exact
mean (SD), mEq/L test, Mann-Whitney U test, or t test,
Hyperglycemia, No. (%)d 154 (81.5) 169 (90.9) 323 (86.1) .009 as appropriate.
b
Maximum glucose concentration, median 160 (134-196) 164 (145-204) 161 (140-201) .04 The MRC Scale for Muscle Strength
(IQR), mg/dL scores range from 0 to 60; a score
Hyperglycemia during study medication 145 (76.7) 164 (88.2) 309 (82.4) .004 less than 48 indicates muscle
administration, No. (%) weakness.
Maximum glucose concentration during study 157 (133-198) 170 (147-208) 163 (141-201) .006 c
Defined as a sodium concentration
medication administration, greater than 155 mEq/L.
median (IQR), mg/dL
d
Defined as a glucose concentration
Other, No. (%) 18 (9.5) 12 (6.5) 30 (8.0) .27
greater than 150 mg/dL.

to 11.4%; P = .44). At 180 days, mortality occurred in 45 of 168 veloped severe hypertension during hydrocortisone admin-
patients (26.8%) in the hydrocortisone group vs 37 of 167 pa- istration, which required antihypertensive therapy. Both pa-
tients (22.2%) in the placebo group (difference, 4.6%; 95% CI, tients recovered without sequelae. Secondary infections,
−4.6% to 13.7%; P = .32). Analysis of the PP population re- weaning failure, muscle weakness, hypernatremia, or other ad-
vealed no significant differences for secondary end points be- verse events were not significantly different between treat-
tween the treatment arms. A post hoc analysis of 54 patients ment groups (Table 3).
with CAP did not reveal significant differences for the pri-
mary or secondary end points between patients treated with Critical Illness–Related Corticosteroid Insufficiency
hydrocortisone (n = 24) or placebo (n = 30). Cortisol data from corticotropin tests were available from 206
of 353 patients (58.4%), of whom 69 (33.5%) had CIRCI. Base-
Adverse Effects line characteristics were comparable between patients with and
There were more episodes of hyperglycemia (blood glucose without CIRCI (eTable 4 in Supplement 1). In the placebo group,
level >150 mg/dL [to convert to millimoles per liter, multiply septic shock occurred in 10 of 37 patients with CIRCI (27.0%)
by 0.0555]) in the hydrocortisone group (169 of 186 patients and in 9 of 66 patients without CIRCI (13.6%) (difference, 13.4%;
[90.9%]) than in the placebo group (154 of 189 patients [81.5%]) 95% CI, −2.1% to 30.5%; P = .09). In a multivariate analysis,
(difference, 9.4%; 95% CI, 2.4% to 16.4%; P = .009) (Table 3). SOFA score (per SOFA point: odds ratio = 1.26; 95% CI, 1.07-
The total amount of administered insulin was not signifi- 1.49; P = .007) and CIRCI (odds ratio = 2.58; 95% CI, 1.13-5.91;
cantly different between the hydrocortisone and placebo P = .03) at baseline, but not age or sex, were independently
groups (safety set analysis: mean [SD], 264.6 [312.2] vs 212.2 prognostic for development of septic shock (eTable 5 in
[246.8] IU, respectively; difference, 52.4 IU; 95% CI, −21.8 to Supplement 1). In this subpopulation of 206 patients, there was
126.7 IU; P = .17) (eTable 3 in Supplement 1). Two patients de- no significant difference regarding the primary or secondary

jama.com (Reprinted) JAMA November 1, 2016 Volume 316, Number 17 1781

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Research Original Investigation Hydrocortisone and Septic Shock in Patients With Severe Sepsis

end points between patients with or without CIRCI who re- this beneficial effect of hydrocortisone could outweigh
ceived hydrocortisone or placebo (eTable 6 in Supplement 1). adverse effects such as hyperglycemia, which can be readily
treated by administration of insulin. Vasopressor-dependent
Delirium (volume-resistant) septic shock in patients with sepsis is
Delirium was assessed by the Richmond Agitation-Sedation still the key indication for suggesting hydrocortisone in the
Scale and the Confusion Assessment Method for the ICU in current international guidelines of the Surviving Sepsis
286 of 353 patients (81.0%). The median number of assess- Campaign.2 Hence, it seems plausible to take this as a pri-
ments per patient was 7 (IQR, 4-12) in the placebo group and mary end point if prevention of septic shock is tested. We
6.5 (IQR, 4-10.5) in the hydrocortisone group (difference, 1.61 assumed a rate of septic shock of 40% in patients with severe
assessments; 95% CI, −0.36 to 3.58 assessments; P = .21). sepsis based on previous studies8,25-27 and data from the Sep-
Twenty-six patients were excluded from analysis owing to Net Clinical Trials Group. However, owing to different study
low Richmond Agitation-Sedation Scale score or incomplete populations, effect sizes, and septic shock criteria, only a
data. In the remaining 260 patients, delirium was less fre- rough estimation was possible (see limitations described
quent in patients who received hydrocortisone than in those later). According to recommendations of the German Sepsis
who received placebo (11 of 130 patients [8.5%] vs 25 of 130 Society to transfer patients with severe sepsis to an interme-
patients [19.2%], respectively; difference, −10.8%; 95% CI, diate care unit or ICU, only patients treated in these units
−19.2% to −2.3%; P = .01). The results remained significant were enrolled in HYPRESS. This was different from a recent
after exclusion of another 60 patients (28 from the placebo retrospective cohort study of patients with sepsis and similar
group, 32 from the hydrocortisone group) who were diag- lactate values but lower hospital mortality (7.9% mortality in
nosed by the investigator to have no delirium but had at least the study by Liu et al29 vs 13.1% mortality in this study), who
1 incomplete delirium assessment or had only 1 baseline were treated predominantly in the general ward.29
assessment (n = 6 in the hydrocortisone group) (with However, the current study did not show a protective ef-
delirium occurring in 11 of 98 patients [11.2%] in the hydro- fect on shock development, time to septic shock, mortality, or
cortisone group vs 25 of 102 patients [24.5%] in the placebo LOS in the ICU or hospital. The results are in contrast to a small
group; difference, −13.3%; 95% CI, −23.7% to −2.6%; P = .01) study in 46 patients with severe CAP, in which hydrocorti-
(Table 2). sone administration was associated with improved survival and
significantly lower rates of septic shock. This study was pre-
maturely stopped after an interim analysis owing to improve-
ment of oxygenation and hospital mortality.8 The presented
Discussion data did not support that hydrocortisone was more effective
In this RCT, low-dose hydrocortisone did not prevent the evo- with regard to the primary or secondary end points in pa-
lution from severe sepsis to septic shock. There were no sig- tients with pneumonia or in the subgroup of patients with CAP.
nificant differences between treatment groups with regard to However, recent larger RCTs of steroids in patients with CAP
mortality or LOS in the ICU or hospital, or mortality up to 180 did not reveal reduction of mortality but did show some ben-
days. Patients treated with hydrocortisone had a signifi- eficial effects on other outcomes. In 304 non-ICU patients, a
cantly higher risk of developing hyperglycemia and a lower risk 4-day treatment with 5 mg of dexamethasone reduced me-
of developing delirium in a post hoc analysis. dian hospital LOS by 1 day.30 Another RCT enrolled 120 pa-
The rationale for this study was based on the notion that tients with severe CAP and a high inflammatory response who
severe sepsis and septic shock reflect stages of a disease con- received methylprednisolone at a dosage of 0.5 mg/kg twice
tinuum with increasing mortality.28 Occurrence of septic daily or placebo for 5 days.10 Primary end points were para-
shock was chosen as the primary outcome variable for the meters of early (<72 hours) or late (72-120 hours) treatment fail-
following reasons. First, hydrocortisone in septic shock ure. Treatment failure rates were lower in patients who re-
showed conflicting results in terms of mortality, but con- ceived steroids than in those who received placebo (13% vs 32%,
sistent hemodynamic effects and faster septic shock res- respectively; P = .02), with late radiographic progression as the
olution.6,24 Second, administration of hydrocortisone was only significantly different parameter; notably, late septic shock
associated with immunomodulatory effects including was rare and less frequent in patients who received steroids
reduced inducible nitric oxide formation—a key mediator of than in those who received placebo (0 vs 4 patients, respec-
septic shock pathophysiology.17 Thus, it was hypothesized tively). In the largest RCT to date, with 785 patients with CAP,
that attenuation of inflammation in early severe sepsis could patients received either 50 mg of prednisone or placebo for 7
prevent progression to shock. Third, this hypothesis was sup- days.9 The primary end point was time to clinical stability for
ported by data of Confalonieri et al 8 showing that septic at least 24 hours. Treatment with prednisone shortened time
shock was prevented by hydrocortisone in patients with CAP. to clinical stability by 1 to 4 days irrespective of severity of CAP,
Fourth, because occurrence of septic shock is associated and it shortened time to hospital discharge and duration of an-
with an increased risk of mortality,28 it was hypothesized tibiotic therapy by 1 day. Thus, there is some evidence that
that prevention of septic shock leading to organ dysfunction moderate doses of steroids may be more effective in patients
could also affect secondary outcome variables such as LOS with CAP than other causes of sepsis.11
in the ICU or mortality. If prolonged administration of hydro- Another interesting finding in this study was that despite
cortisone was able to prevent septic shock development, a higher risk of septic shock in patients with CIRCI, there was

1782 JAMA November 1, 2016 Volume 316, Number 17 (Reprinted) jama.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Hydrocortisone and Septic Shock in Patients With Severe Sepsis Original Investigation Research

no detectable difference in septic shock development be- illness and provide some evidence of protective effects of pro-
tween treatment groups; however, sample size and event num- longed administration of low-dose hydrocortisone in early se-
bers were small. The results are in line with results of the vere sepsis.
CORTICUS trial, in which effects of hydrocortisone on septic There are limitations of the study to be addressed. First,
shock resolution were independent from CIRCI. Moreover, in inclusion in the trial was possible only after informed con-
a CORTICUS subgroup analysis, etomidate administration was sent could be obtained, so that patients who developed sep-
associated with a significantly higher risk for CIRCI and 28- tic shock early may have been missed. Second, data from cor-
day mortality. However, hydrocortisone treatment did not re- ticotropin tests were available in a subgroup of 206 patients
duce mortality in patients with CIRCI who received etomidate.31 only, because adrenal function assessment was not an abso-
Whether combined biomarkers may be more suitable to de- lute inclusion criterion and therefore was not performed at all
tect eligible patients needs further investigation. A subgroup study sites or in all patients at a single site. Analyses of CIRCI
analysis of patients with CAP showed most effects in patients were post hoc and should be regarded as hypothesis generat-
who had low basal cortisol levels combined with high proin- ing only. Third, mortality in the study population was more
flammatory cytokine profiles.32 comparable to the mortality reported in recent studies with
The results from this study do not support an increased CAP 9,10,30 than to the mortality (approximately 30%) re-
risk for secondary infections. Reasonable concerns were am- ported in the study by Confalonieri et al8; thus, it cannot be
plified by results of the CORTICUS trial, which showed higher excluded that hydrocortisone would have been more effec-
numbers of secondary infections including new sepsis and sep- tive in patients with a higher risk of death. Fourth, data on de-
tic shock in patients who received hydrocortisone.4 A Bayesian lirium should be interpreted with caution because patients had
analysis, mainly based on CORTICUS results, of patients with to be excluded owing to incomplete data sets, only 1 daily as-
septic shock reported an increased risk of infections.33 How- sessment was performed in most patients, and interrater re-
ever, other studies in patients with septic shock,3 acute respi- liability was not assessed. Analyses were performed post hoc
ratory distress syndrome,34 trauma,35 or CAP8-10,30 and a and results should therefore be regarded as hypothesis gen-
meta-analysis7 did not report any increased risk. erating. Fifth, secondary analyses were not adjusted for mul-
An unexpected finding was that delirium developed less tiple testing, as these statistical comparisons were performed
frequently in patients treated with hydrocortisone. Indeed, sev- with exploratory rather than confirmatory intention. Adjust-
eral observational studies reported an association between high ment for clustering within site was not performed because site
serum cortisol concentration and delirium after noncardiac36 was a stratification factor for randomization. Sixth, the ob-
and cardiac 37 surgery and severe sepsis or septic shock.38 There served rate of septic shock in the placebo group (23%; 95% CI,
is compelling evidence that systemic inflammation plays an 17%-30%) was lower than originally presumed for sample size
important role in the pathogenesis of delirium, but it remains calculation (40%). The point estimate of the difference of sep-
controversial whether high cortisol concentration mirrors an tic shock occurrences was −1.8% (ie, close to 0) with a 95% CI
activation of the stress response to cope with inflammation, of −10.7% to 7.2%. Hence, the trial gives no indication of a clini-
or whether delirium is caused by high cortisol concentration. cally relevant major benefit of hydrocortisone treatment, which
A prospective cohort study in 330 mechanically ventilated pa- it was designed to detect.
tients with acute lung injury reported a significant dose-
independent association between systemic corticosteroid ad-
ministration and development of delirium.39 However, in an
RCT with 737 patients undergoing cardiac surgery who re-
Conclusions
ceived dexamethasone, 1 mg/kg, or placebo at anesthesia in- Among adults with severe sepsis not in septic shock, the use
duction did not reveal a significant difference of postopera- of hydrocortisone compared with placebo did not reduce the
tive delirium between treatment groups.40 Thus, the results risk of septic shock within 14 days. These findings do not sup-
question the concept of cortisol-induced delirium in critical port the use of hydrocortisone in these patients.

ARTICLE INFORMATION Anesthesiology, Klinikum der Ludwig-Maximilians- Cardiology and Intensive Care Medicine, Vivantes
Published Online: October 3, 2016. Universität, München, Germany (Briegel); Institute Humboldt Klinikum, Berlin, Germany (Lehmke);
doi:10.1001/jama.2016.14799 for Medical Informatics, Statistics and Department of Emergency and Intensive Care
Epidemiology, Leipzig, Germany (Engel, Loeffler); Medicine, Klinikum Ernst von Bergmann, Potsdam,
Author Affiliations: Department of Department of Anesthesia, Intensive Care Medicine Germany (Oppert); Department of Intensive Care
Anesthesiology and Intensive Care Medicine, and Pain Management, Vivantes-Klinikum Medicine, HELIOS Hospital Berlin-Buch, Berlin,
Charité–Universitätsmedizin Berlin, Berlin, Neukölln, Berlin, Germany (Gerlach); Department Germany (Resener); Department of Anesthesiology
Germany (Keh, Goldmann); Clinical Trial Centre of Anesthesiology and Intensive Care Medicine, and Intensive Care Medicine, University Medical
Leipzig, Leipzig, Germany (Trips, Bogatsch); Ernst Moritz Arndt University, Greifswald, Germany Center Schleswig Holstein, Campus Kiel, Kiel,
Department of Intensive Care and Intermediate (Kuhn); Department of Anesthesia and Intensive Germany (Schädler, Weiler); Department of
Care, University Hospital RWTH Aachen, Aachen, Care, Zentralklinik Bad Berka, Bad Berka, Germany Anesthesiology, Intensive Care Medicine,
Germany (Marx, Schuerholz); Department of (Hüter); Department of Anesthesia, Intensive Care Emergency Medicine and Pain Management,
Intensive Care Medicine, HELIOS Hospital Bad Medicine and Pain Management, HELIOS Hospital Klinikum Oldenburg Medical Campus Carl von
Saarow, Bad Saarow, Germany (Wirtz, Erfurt, Erfurt, Germany (Meier-Hellmann); Ossietzky Universität, Oldenburg, Germany
Abduljawwad); Department of Anesthesiology and Department of Intensive Care Medicine, University (Weyland); Department of Anesthesiology and
Intensive Care Medicine, University of Leipzig, Medical Center Hamburg-Eppendorf, Hamburg, Intensive Care Medicine, Jena University Hospital,
Leipzig, Germany (Bercker, Simon); Department of Germany (Nierhaus, Kluge); Department of Jena, Germany (Reinhart); Center for Clinical

jama.com (Reprinted) JAMA November 1, 2016 Volume 316, Number 17 1783

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Research Original Investigation Hydrocortisone and Septic Shock in Patients With Severe Sepsis

Studies, Center for Sepsis Control and Care, Jena of Leipzig, Leipzig, Germany, contributed to the 10. Torres A, Sibila O, Ferrer M, et al. Effect of
University Hospital, Jena, Germany (Brunkhorst). study: Holger Bogatsch, MD, Ana Lucía Martín corticosteroids on treatment failure among
Author Contributions: Drs Keh and Loeffler had Montañez, Christiane Schönherr, and Anke Schöler hospitalized patients with severe
full access to all of the data in the study and take (project and data management); Dagmar Fiedler, community-acquired pneumonia and high
responsibility for the integrity of the data and the Tobias Kurz, Monika Rohwedder, Anja Schneider, inflammatory response: a randomized clinical trial.
accuracy of the data analysis. Dr Keh and Ms Trips Angelika Siegmund, and Daniela Gayda JAMA. 2015;313(7):677-686.
contributed equally to the work. (monitoring); Evelyn Trips, Christoph Engel, MD, 11. Siemieniuk RA, Meade MO, Alonso-Coello P,
Concept and design: Keh, Trips, Briegel, Engel, Holger Bogatsch, MD, Markus Loeffler, MD et al. Corticosteroid therapy for patients
Gerlach, Loeffler, Oppert, Reinhart. (biometry); Madlen Dörschmann, Thekla Haage, hospitalized with community-acquired pneumonia:
Acquisition, analysis, or interpretation of data: Keh, Bianca Scholze, Anja Schneider, and Holger a systematic review and meta-analysis. Ann Intern
Trips, Marx, Wirtz, Abduljawwad, Bercker, Bogatsch, MD (pharmacovigilance); and Thomas Med. 2015;163(7):519-528.
Bogatsch, Engel, Gerlach, Goldmann, Kuhn, Hüter, Junge, Matthias Collier, and Andre Rothe
(informatics); the institutions received payment 12. Bone RC, Balk RA, Cerra FB, et al; ACCP/SCCM
Meier-Hellmann, Nierhaus, Kluge, Lehmke, Loeffler, Consensus Conference Committee, American
Oppert, Resener, Schädler, Schuerholz, Simon, from Charité–Universitätsmedizin Berlin by grant
01KG0701 from the German Federal Ministry of College of Chest Physicians/Society of Critical Care
Weiler, Weyland, Reinhart. Medicine. Definitions for sepsis and organ failure
Drafting of the manuscript: Keh, Marx, Bogatsch, Education and Research. Anne Gössinger,
Department of Anesthesiology and Intensive Care and guidelines for the use of innovative therapies in
Engel, Gerlach, Hüter, Loeffler. sepsis. Chest. 1992;101(6):1644-1655.
Critical revision of the manuscript for important Medicine, Charité–Universitätsmedizin Berlin,
intellectual content: Keh, Trips, Marx, Wirtz, Berlin, Germany, assisted in designing and 13. Reinhart K, Brunkhorst F, Bone H, et al;
Abduljawwad, Bercker, Briegel, Engel, Gerlach, coordinating the trial; she received payment from Deutsche Sepsis-Gesellschaft e.V. Diagnosis and
Goldmann, Kuhn, Meier-Hellmann, Nierhaus, Kluge, grant 01KG0701 from the German Federal Ministry therapy of sepsis: guidelines of the German Sepsis
Lehmke, Loeffler, Oppert, Resener, Schädler, of Education and Research. Michael Vogeser, MD, Society Inc. and the German Interdisciplinary
Schuerholz, Simon, Weiler, Weyland, Reinhart. PhD, Department of Laboratory Medicine, Klinikum Society for Intensive and Emergency Medicine [in
Statistical analysis: Trips, Bogatsch, Engel, Gerlach, der Ludwig-Maximilians-Universität, München, German]. Anaesthesist. 2006;55(suppl 1):43-56.
Loeffler. Germany, measured cortisol; he received no 14. Brunkhorst FM, Engel C, Bloos F, et al; German
Obtained funding: Keh, Reinhart. compensation. Competence Network Sepsis (SepNet). Intensive
Administrative, technical, or material support: Keh, insulin therapy and pentastarch resuscitation in
Marx, Wirtz, Bercker, Bogatsch, Engel, Goldmann, REFERENCES severe sepsis. N Engl J Med. 2008;358(2):125-139.
Kluge, Loeffler, Oppert, Resener, Simon, Weiler, 1. Patel GP, Balk RA. Systemic steroids in severe 15. Pocock SJ. Clinical Trials: A Practical Approach.
Weyland. sepsis and septic shock. Am J Respir Crit Care Med. New York, NY: John Wiley & Sons; 1983.
Study supervision: Keh, Briegel, Engel, Gerlach, 2012;185(2):133-139.
Kuhn, Oppert, Weiler. 16. Loisa P, Parviainen I, Tenhunen J, Hovilehto S,
2. Dellinger RP, Levy MM, Rhodes A, et al; Surviving Ruokonen E. Effect of mode of hydrocortisone
Conflict of Interest Disclosures: All authors have Sepsis Campaign Guidelines Committee Including administration on glycemic control in patients with
completed and submitted the ICMJE Form for the Pediatric Subgroup. Surviving Sepsis Campaign: septic shock: a prospective randomized trial. Crit
Disclosure of Potential Conflicts of Interest. Dr Marx international guidelines for management of severe Care. 2007;11(1):R21.
reported receiving a research grant and honoraria sepsis and septic shock, 2012. Intensive Care Med.
for lecturing and consulting from B. Braun 2013;39(2):165-228. 17. Keh D, Boehnke T, Weber-Cartens S, et al.
Melsungen and receiving a research grant and Immunologic and hemodynamic effects of
3. Annane D, Sébille V, Charpentier C, et al. Effect “low-dose” hydrocortisone in septic shock:
honoraria for consulting from Adrenomed. Dr Kuhn of treatment with low doses of hydrocortisone and
reported receiving personal fees from Dräger a double-blind, randomized, placebo-controlled,
fludrocortisone on mortality in patients with septic crossover study. Am J Respir Crit Care Med. 2003;
Medical Germany. Dr Schuerholz reported receiving shock. JAMA. 2002;288(7):862-871.
personal fees from Bayer HealthCare for consulting; 167(4):512-520.
serving on an advisory board and receiving personal 4. Sprung CL, Annane D, Keh D, et al; CORTICUS 18. Ely EW, Truman B, Shintani A, et al. Monitoring
fees for lecturing and consulting from Astellas Study Group. Hydrocortisone therapy for patients sedation status over time in ICU patients: reliability
Pharma; receiving personal fees for lecturing from with septic shock. N Engl J Med. 2008;358(2):111-124. and validity of the Richmond Agitation-Sedation
B. Braun Melsungen; and serving as chief medical 5. Marik PE, Pastores SM, Annane D, et al; Scale (RASS). JAMA. 2003;289(22):2983-2991.
officer for Brandenburg Antiinfektiva GmbH. American College of Critical Care Medicine. 19. Ely EW, Inouye SK, Bernard GR, et al. Delirium
Dr Reinhart reported serving as a paid advisor to Recommendations for the diagnosis and in mechanically ventilated patients: validity and
Adrenomed and holding equity in InflaRx. No other management of corticosteroid insufficiency in reliability of the Confusion Assessment Method for
disclosures were reported. critically ill adult patients: consensus statements the Intensive Care Unit (CAM-ICU). JAMA. 2001;
Funding/Support: This study was supported by from an international task force by the American 286(21):2703-2710.
Charité–Universitätsmedizin Berlin and by grant College of Critical Care Medicine. Crit Care Med.
2008;36(6):1937-1949. 20. Vincent JL, Moreno R, Takala J, et al; Working
01KG0701 from the German Federal Ministry of Group on Sepsis-Related Problems of the European
Education and Research. 6. Sligl WI, Milner DA Jr, Sundar S, Mphatswe W, Society of Intensive Care Medicine. The SOFA
Role of the Funder/Sponsor: The funding Majumdar SR. Safety and efficacy of corticosteroids (Sepsis-related Organ Failure Assessment) score to
agencies had no role in the design and conduct of for the treatment of septic shock: a systematic describe organ dysfunction/failure. Intensive Care
the study; collection, management, analysis, and review and meta-analysis. Clin Infect Dis. 2009;49 Med. 1996;22(7):707-710.
interpretation of the data; preparation, review, or (1):93-101.
21. Le Gall JR, Lemeshow S, Saulnier F.
approval of the manuscript; and decision to submit 7. Annane D, Bellissant E, Bollaert PE, Briegel J, Keh A new Simplified Acute Physiology Score (SAPS II)
the manuscript for publication. D, Kupfer Y. Corticosteroids for treating sepsis. based on a European/North American multicenter
Additional Contributions: Peter Suter, MD, Centre Cochrane Database Syst Rev. 2015;12(12):CD002243. study. JAMA. 1993;270(24):2957-2963.
Médical Universitaire, Geneva, Switzerland, 8. Confalonieri M, Urbino R, Potena A, et al. 22. Knaus WA, Draper EA, Wagner DP,
Stephan Harbarth, MD, Hôpitaux Universitaires Hydrocortisone infusion for severe Zimmerman JE. APACHE II: a severity of disease
Genève, Service de Prévention et Contrôle des community-acquired pneumonia: a preliminary classification system. Crit Care Med. 1985;13(10):
Infections, Geneva, Switzerland, and Klaus-Dieter randomized study. Am J Respir Crit Care Med. 2005; 818-829.
Wernecke, MD, SOSTANA GmbH, Berlin, Germany, 171(3):242-248.
served on the data and safety monitoring board; 23. Moreno RP, Metnitz PG, Almeida E, et al;
9. Blum CA, Nigro N, Briel M, et al. Adjunct SAPS 3 Investigators. SAPS 3—from evaluation of
they received payment from grant 01KG0701 from prednisone therapy for patients with
the German Federal Ministry of Education and the patient to evaluation of the intensive care unit,
community-acquired pneumonia: a multicentre, part 2: development of a prognostic model for
Research. The following individuals from the double-blind, randomised, placebo-controlled trial.
Clinical Trial Centre Leipzig and Institute for Medical hospital mortality at ICU admission. Intensive Care
Lancet. 2015;385(9977):1511-1518. Med. 2005;31(10):1345-1355.
Informatics, Statistics and Epidemiology, University

1784 JAMA November 1, 2016 Volume 316, Number 17 (Reprinted) jama.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018

 Hydrocortisone and Septic Shock in Patients With Severe Sepsis Original Investigation Research

24. Annane D, Bellissant E, Bollaert PE, Briegel J, 30. Meijvis SC, Hardeman H, Remmelts HH, et al. 36. Cerejeira J, Batista P, Nogueira V, Vaz-Serra A,
Keh D, Kupfer Y. Corticosteroids for severe sepsis Dexamethasone and length of hospital stay in Mukaetova-Ladinska EB. The stress response to
and septic shock: a systematic review and patients with community-acquired pneumonia: surgery and postoperative delirium: evidence of
meta-analysis. BMJ. 2004;329(7464):480. a randomised, double-blind, placebo-controlled hypothalamic-pituitary-adrenal axis
25. Bone RC, Fisher CJJ Jr, Clemmer TP, Slotman trial. Lancet. 2011;377(9782):2023-2030. hyperresponsiveness and decreased suppression of
GJ, Metz CA, Balk RA. A controlled clinical trial of 31. Cuthbertson BH, Sprung CL, Annane D, et al. the GH/IGF-1 axis. J Geriatr Psychiatry Neurol. 2013;
high-dose methylprednisolone in the treatment of The effects of etomidate on adrenal responsiveness 26(3):185-194.
severe sepsis and septic shock. N Engl J Med. 1987; and mortality in patients with septic shock. 37. Plaschke K, Fichtenkamm P, Schramm C, et al.
317(11):653-658. Intensive Care Med. 2009;35(11):1868-1876. Early postoperative delirium after open-heart
26. Rangel-Frausto MS, Pittet D, Costigan M, 32. Remmelts HH, Meijvis SC, Heijligenberg R, et al. cardiac surgery is associated with decreased
Hwang T, Davis CS, Wenzel RP. The natural history Biomarkers define the clinical response to bispectral EEG and increased cortisol and
of the systemic inflammatory response syndrome dexamethasone in community-acquired interleukin-6. Intensive Care Med. 2010;36(12):
(SIRS): a prospective study. JAMA. 1995;273(2):117- pneumonia. J Infect. 2012;65(1):25-31. 2081-2089.
123. 33. Kalil AC, Sun J. Low-dose steroids for septic 38. Nguyen DN, Huyghens L, Zhang H,
27. Vincent JL, Angus DC, Artigas A, et al; shock and severe sepsis: the use of Bayesian Schiettecatte J, Smitz J, Vincent JL. Cortisol is an
Recombinant Human Activated Protein C statistics to resolve clinical trial controversies. associated-risk factor of brain dysfunction in
Worldwide Evaluation in Severe Sepsis (PROWESS) Intensive Care Med. 2011;37(3):420-429. patients with severe sepsis and septic shock.
Study Group. Effects of drotrecogin alfa (activated) Biomed Res Int. 2014;2014:712742.
34. Steinberg KP, Hudson LD, Goodman RB, et al;
on organ dysfunction in the PROWESS trial. Crit National Heart, Lung, and Blood Institute Acute 39. Schreiber MP, Colantuoni E, Bienvenu OJ, et al.
Care Med. 2003;31(3):834-840. Respiratory Distress Syndrome (ARDS) Clinical Corticosteroids and transition to delirium in
28. Alberti C, Brun-Buisson C, Goodman SV, et al; Trials Network. Efficacy and safety of patients with acute lung injury. Crit Care Med. 2014;
European Sepsis Group. Influence of systemic corticosteroids for persistent acute respiratory 42(6):1480-1486.
inflammatory response syndrome and sepsis on distress syndrome. N Engl J Med. 2006;354(16): 40. Sauër AM, Slooter AJ, Veldhuijzen DS, van Eijk
outcome of critically ill infected patients. Am J 1671-1684. MM, Devlin JW, van Dijk D. Intraoperative
Respir Crit Care Med. 2003;168(1):77-84. 35. Roquilly A, Mahe PJ, Seguin P, et al. dexamethasone and delirium after cardiac surgery:
29. Liu VX, Morehouse JW, Marelich GP, et al. Hydrocortisone therapy for patients with multiple a randomized clinical trial. Anesth Analg. 2014;119
Multicenter implementation of a treatment bundle trauma: the randomized controlled HYPOLYTE (5):1046-1052.
for patients with sepsis and intermediate lactate study. JAMA. 2011;305(12):1201-1209.
values. Am J Respir Crit Care Med. 2016;193(11):
1264-1270.

jama.com (Reprinted) JAMA November 1, 2016 Volume 316, Number 17 1785

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: on 05/30/2018