Clinical Medicine & Research

Volume 13, Number 3-4: 156-168

2015 Marshfield Clinic Health System
clinmedres.org Review

Understanding Decision Making in Critical Care

Geoffrey K. Lighthall, MD, PhD and Cristina Vazquez-Guillamet, MD

Background: Human decision making involves the deliberate formulation of hypotheses and plans as
well as the use of subconscious means of judging probability, likely outcome, and proper action.
Rationale: There is a growing recognition that intuitive strategies such as use of heuristics and pattern
recognition described in other industries are applicable to high-acuity environments in medicine.
Despite the applicability of theories of cognition to the intensive care unit, a discussion of decision-
making strategies is currently absent in the critical care literature.
Content: This article provides an overview of known cognitive strategies, as well as a synthesis of their
use in critical care. By understanding the ways by which humans formulate diagnoses and make critical
decisions, we may be able to minimize errors in our own judgments as well as build training activities
around known strengths and limitations of cognition.

Keywords: Critical care; Medical decision making; Heuristics; Intuition: Diagnosis

C ritical care practitioners need to make accurate and

appropriate decisions in a timely manner in situations where
diagnoses as well as the formulation of general impressions
(rapid deterioration, poor respiratory reserve, etc.) that form
there is a high degree of stress and uncertainty and when the basis of action. This paper will review some of the current
patients possess little physiologic reserve. Diagnostic concepts of reasoning and their applicability to critical care.
decisions in acute care disciplines often commit patients to With a better understanding of cognitive processes, we can
therapeutic strategies that can be harmful if not correct. take further steps by designing better error prevention
Diagnoses and related decisions surround our practice at strategies and assessing them experimentally.
every moment; however, the schemes our minds use to make
these judgments are probably the aspect of medical care that Diagnostic Error in Medicine and Critical Care
we understand the least. Diagnostic reasoning involves the Is diagnostic reasoning a patient safety problem in critical
use of both conscious and subconscious processes, the blend care? Studies in the United States, United Kingdom, and
of which depends on the experience of the clinician and the Australia indicate that errors or adverse events occur in the
familiarity of the situation. Factors such as overconfidence, range of 1.1% to 16.6% of all hospital admissions,3-6 with
fatigue, and time pressure can create an overreliance on many showing a greater incidence with age, complexity, and
intuition when there may be insufficient expertise to justify acuity of the patient.3,7-11 Analysis of malpractice claims
its use.1,2 Patient harm can result from both errors of indicates that diagnostic error is both the leading cause of
commission (for example, wrong diagnosis made, improper paid claims and the most costly category.12,13 However
therapy given) and omission (not fully appreciating the extent connecting error types to actual error rates from audits, self-
of a patients illness and failing to act quickly). Thus, reports, and other databases without a known denominator is
diagnostic reasoning involves both the making of formal not possible.

Corresponding Author: Received: April 16, 2015

Geoffrey K. Lighthall MD, PhD, Associate Professor Revised: July 18, 2015
Department of Anesthesia, MC 112A Accepted: August 6, 2015
Department of Veterans Affairs Medical Center
3801 Miranda Avenue doi:10.3121/cmr.2015.1289
Palo Alto, CA 94304
Tel: (650) 493-5000, ext. 66756
Email: [email protected]

156
The frequency of diagnostic error is probably not fully The Hypothetical-Deductive Model (HD)
appreciated in medical practice, especially to the physicians Upon encountering a patient, a hypothesis regarding his
involved. Chart audits found that clinicians report mistakes at condition is generated and then further addressed by physical
about half the frequency reported by the affected patients.14 A examination and diagnostic testing. As information becomes
study on resident teaching conferences found that the available, the hypotheses can be revised or additional
reporting and discussion of error varied markedly between hypotheses generated. Typically, recall of prior experiences
medical and surgical departments; a lack of standards for as well as consideration of disease prevalence figure into
reporting and discussing errors was noted, along with a hypothesis generation. The dominant hypothesis then creates
paucity of meaningful discussion of errors and their sources a framework for subsequent data gathering and analysis, for
in educational activities.15 Residents further admit that errors which confirmation and elimination tend to be major
are sporadically reported to patients (24% of the time), and strategies.24 Confirmation is a search for information that
also infrequently to supervising physicians (50%).16 Self- confirms a hypothesis, while elimination is a strategy used to
detection of error is likely poor, as individuals are notoriously rule out a process, often a competing diagnosis or the worse
unaware of knowledge deficits and generally overrate their scenario for a given constellation of symptoms. Through this
performance of difficult clinical tasks.17,18 process, hypotheses are further refined, and ideally, verified
for plausibility prior to commencing a course of action.
The concentration of errors in critical care is likely higher Understanding salient features of competing diagnoses and
than that reported for inpatients in general. A live observational key differentiating factors can improve the selection and
study in Israel documented a rate of 1.7 errors per patient per interpretation of diagnostic tests and is ideally a key
day.19 In another series, 31% of intensive care unit (ICU) component of hypothesis evaluation.
admissions experienced an iatrogenic complication, wherein
almost half were considered major complications where The scientific foundation of the hypothetical-deductive
human error was a major contributor.9 A prospective incident method is Bayes theorem on conditional probability. The
reporting system in Australia found that adverse events occur theorem states that the probability of a diagnosis (the post-test
in 5% to 25% of patients admitted to the ICU,20 while errors probability) is dependent upon the probability of the diagnosis
occurred in 20% of patients in a single center study in the prior to a test (the pretest probability) and the strength of
United States.21 Finally, a German study collected incident evidence added by a test or maneuver. The likelihood ratio
reports from 216 consecutive admissions and found that 15% calculated from experimental test results is converted to odds
were victims of error, and 73% of the errors were due to to form the strength of evidence. The prevalence of a
human failures.10 Despite what appears to be a persistence of condition in the population is often a good estimate of the
errors and mishaps, the fraction attributable to diagnostic pre-test probability.
error is not as clear. Diagnostic reasoning is not always
explored in error analysis, and there is little standardization of Despite its implicit status as the standard diagnostic method,
categories and definitions across investigations. 12 the hypothetical-deductive method can be inefficient or error
Retrospective analyses of patients who experienced unplanned prone and will not yield the same results for all practitioners.
ICU admissions from regular wards found suboptimal care At its best, the hypothetical-deductive method should promote
present in 32% to 50% of cases. In these patients, non- a dynamic and probabilistic view of potential diagnoses and
recognition of the problem or its severity, lack of supervision, seek information that adjusts the relative probabilities of
inexperience, and inappropriate treatment were key factors competing hypotheses. However, in the ICU we often see the
underlying patient morbidity and mortality.22,23 opposite, as junior trainees become lost in the acquisition and
presentation of data (believing that it will improve diagnostic
What We Know about Decision Making General accuracy) but without using it to generate a testable hypotheses
Models for Cognition in Diagnostic Reasoning or meaningful course of action.25 Some of this unfocused
When one thinks about diagnostic reasoning in medicine, the activity can be attributed to inexperience in medicine and not
strategy that comes to mind is the hypothetical-deductive knowing what evidence to seek when faced with certain
model. This method is typified by case-based discussions led conditions.
by senior physicians where complex patients are analyzed
and distilled down to one or two possible diagnoses and then Intuitive Methods
differentiated by a key test, maneuver, or response to therapy. The nature of illnesses encountered in the emergency
Particular strengths, weaknesses, and evidentiary contexts of department and ICU often necessitates actions in the absence
tests are considered when using this method. In contrast, of a comprehensive battery of data and lack of time for
intuitive methods involve comparing a set of conditions with iterative testing of hypotheses. With experience, the clinician
previously learned patterns or prototypes of disease states; is able to compare the problem at hand with mental templates
use of intuition occurs almost instantly, but requires a great or scripts of disease processes stored away from prior
deal of experience to perform optimally. These two approaches experiences.26,27 As will be described, the nature of what we
will be expanded upon in the sections below. remember is based on an entirely subjective set of experiences
including our own clinical cases, diagnoses that we failed to

CM&R 2015 : 3-4 (December) Lighthall and Vazquez-Guillamet 157

Table 1.Contrast between qualities of analytic and intuitive models of decisions in critical care.
Analytic methods Intuitive methods
Basis Serial hypothesis testing based on probabilities Recognition of clinical patterns
derived from clinical information
Role of probability Explicit; estimates and models can rank probabilities Implied; pattern match leads to single
of multiple options diagnosis with greatest probability
Strengths Mathematical, promise of reproducible results Experience based, specific and rapid

Portability of results: Evidence-based

Pitfalls Results of robust models can be skewed by incorrect Overconfidence may cause one to
estimates of pretest probability and strength of over-estimate experience and under-
evidence estimate uncertainty
Not useful in stressful events, accuracy is dependent
upon effort and time
Role of information Requires more information to minimize uncertainty Uncertainty is accepted and does not
slow down decision making
Role of biases Information gathering may follow biases and Heuristics may be associated with
incorrectly skew the output of supposedly objective biases that skew views toward
mathematical models likelihood and probability
Role of experience Provides workable tools for novices Not reliable for novices
Also benefits experienced providers when clinical Allows experience to improve
patterns are not easily recognized efficiency

make in the past, and images created by stories from our judge whether A may be a manifestation of some other
colleagues and the popular press. Psychologists Daniel process that is much more common than B. Thus, a classic
Kahneman and Amos Tversky28 explored the cognitive basis weakness of representativeness is its neglect of base rate
for such judgments in a series of influential studies. They frequency of the diagnosis considered and failure to
found that rather than extensive algorithmic processing of consistently compare it with more common options.
information, humans making judgments resort to a number of
simplifying shortcuts or heuristics, the use of which is often The original laboratory studies on representativeness found
beyond the awareness of the decision-maker. The psychologists subjects repeatedly making flawed inferences of likelihood
found that each heuristic had specific vulnerabilities that and ignoring information that would have led to more
could be exploited in controlled experiments. accurate predictions. From this work, it was taken that
systematic errors such as these revealed a common fallibility
This work has been expanded upon in a medical context by of human decision making. Heller et al31 took a more detailed
Pat Crockery,25 who found that the pace and volume of look by having pediatric residents estimate probabilities in
decisions made in Emergency Medicine create a high reliance both medical and non-medical scenarios and found that
on pattern recognition and heuristics. Interest in heuristics has estimates of likelihood more accurately reflected underlying
led to research in Internal Medicine and Anesthesiology, probabilities in the non-medical problems and that statistically
where experimental work has verified their use.29,30 Some incorrect judgments were more common in the residents
heuristics likely used in medical decision making are responses to problems based on medical issues. The difference
described below. Heuristic judgments can serve as stand- was generally greater when more senior residents were
alone judgments of causality or serve as an entry point to the compared to interns.31 These findings suggested that heuristics
hypothetical-deductive method by identifying diagnoses might not serve as a general strategy for all decisions but tend
warranting further evaluation. Some elements of intuition- to be used where familiarity (or perceived familiarity) with a
based decisions are contrasted with deductive methods in subject leads to confidence that a shortcut can substitute for
table 1. Below is a description of the most common heuristics the more laborious analytic approach. The implication of
and some of the systemic biases and weaknesses associated Hellers work is that as one gains more experience in a
with their use. It is important to note that these are recurrent subject area, use of heuristics may increase.
patterns more than discrete categories; the reader will find
degrees of overlap between a number of heuristics. Certainly, representativeness seems to be a tool used by
experienced physicians to match the salient features of a
Representativeness is used to judge the likelihood of an event patients presentation to a database of prior experiences. For
A belonging to a condition B, due to similarities between the very experienced, a lack of representativeness probably
the two. The shortcut is based on prior experiences or well- plays an important role in the discovery of new medical
accepted definitions of B. The heuristic makes no attempt to entities, for example when a sporadic cluster of young healthy

158 Decision making in critical care CM&R 2015 : 3-4 (December)

males with pneumocystis pneumonia led to the discovery of that treating the change in mental state as a stand-alone
AIDS. condition may occur to the exclusion of understanding the
primary problem.
The Availability Heuristic is the tendency to assess the
likelihood of an event by the ease with which such events are Confirmation Bias describes the tendency to collect redundant
recalled. As high frequency events will be more easily information that seeks to confirm an existing hypothesis
recalled than the less frequent, availability does have the rather than obtaining studies that differentiate it from
tendency of keeping assumptions of likelihood closer to the competing possibilities. Often the subsequent tests contribute
base rates of various conditions. Middle-aged men presenting little gain in information and do not alter the probability of
to the emergency department with substernal chest pain disease by Bayesian analysis.35 In the presence of multiple lab
should recall numerous instances of coronary artery disease studies and other findings, confirmation bias expresses itself
(CAD). However, availability is dependent on what is as the tendency to filter information in a way that places
memorable, and it has been found that since unusual events, greater importance on confirmatory evidence and less on
famous people, or familiar locations are more memorable, findings that may contradict an existing impression.
they tend to skew impressions of frequency and likelihood.28
For example, a report of a missed diagnosis (say, aortic Sources of this bias may arise from other misassumptions
dissection) may cause emergency physicians to consider the including anchoring (discussed below) or the psychological
same diagnosis in situations where he or she would normally tendency to seek comfort in a decision, rather than admitting
suspect myocardial infarction. Likewise, vivid images of an assumption is incorrect and starting over.25 Patient empathy
adverse events can elevate impressions of risk categories that or other influence from the affective sphere may motivate one
exceed their actual frequency. Following the terrorist attacks to want or not want a certain diagnosis to be present in a
of 2001 in the United States, fear of flying caused more patient and to seek evidence to confirm this belief.
people to take vacations by car; this change in behavior led a
greater number of people to use a statistically more dangerous The longer one is able to hold onto a decision, the more
form of travel.32 difficult it may be to give up.36 Thus, diagnostic information
that trickles in slowly rather than all at once may unconsciously
Unpacking and Support Theory describes the tendency to strengthen an initial belief or hypothesis. Receiving a patient
favor diagnoses and interventions that contain some with a provisional diagnosis from another physician may
description of content over categorical descriptions.33 For likewise incline one to confirm rather than reformulate the
example, a radiologist may be more likely to approve a diagnosis.36 The latter is a huge problem particular to critical
computed tomography (CT) scan of the head if the rationale care, as information is constantly being passed back and forth
is to evaluate for tumor, spontaneous bleed, or abscess, between operative teams, consultants, on and off shift nurses,
rather to exclude central nervous system pathology, even and physicians. We frequently admit patients who come
though the latter category actually encompasses a greater set attached to a long trail of impressions and existing care plans,
of findings. The act of listing these diagnoses seems to both of which have been shown to constrain diagnostic
strengthen the case for obtaining the scan, and in experimental thinking.37,38
settings, enumerating contributing etiologies tends to elevate
the perceived probability of the condition in question. Anchoring and Adjustment, and the Power of Suggestion:
One of the most reproducible findings in experimental
Search Satisficing combines the terms satisfy and suffice, psychology is the influence of an external suggestion on the
and was first used by Economics Nobel Prize winner Herbert content of a decision.39 One form of suggestion is anchoring,
Simon34 in his description of the human tendency to settle for the ability of a supplied numerical value (the anchor) to skew
adequate, but not perfect, solutions. This is sometimes called the estimation of another value. For example, if 100 subjects
discounting, because discovery of one possible explanation are asked whether California has 50 major earthquakes each
tends to discount the possibility of alternate diagnoses.31 For year (all would say false), and then asked to make a more
example, a physician may fail to recognize a second, subtle realistic prediction of the rate, the prediction would be
fracture present on a radiograph if a more obvious primary significantly higher than one made by a group of 100 subjects
fracture was first detected. This heuristic can lead to premature who are asked whether California has a major earthquake
closure, a failure to consider other diagnostic possibilities in every 5 years.
situations where there is still uncertainty.
Adjustment refers to the process of consciously moving the
A common finding we see in the ICU is a tendency to find and estimated value away from the anchor toward one thought to
act upon some recognizable element amongst the sea of be more accurate. Even when presented with absurdly
findings in a complicated patient. For example, delirium may abnormal anchor value, final values continue to reflect the
be recognized as a problem by some, without necessarily bias introduced by the anchor, with adjustments moving the
connecting it to the fact that it is being driven by another more estimate to the edge of the range of uncertainty, but never as
serious problem (sepsis or hypercapnia, for example), and far as one would without the anchor.39,40

CM&R 2015 : 3-4 (December) Lighthall and Vazquez-Guillamet 159

The quantitative demonstration of anchoring in controlled experience.43 From such naturalistic analysis of decisions,
experiments may not reflect the struggles of critical care as Klein developed the model of recognition-primed decision
much as the tendency to anchor to qualitative or named making. This model has two main features: (1) perception of
features of a problem. This is especially problematic in the situational variables and characterizing the situation based
dynamic medical situations where labels such as hypovolemic, on prior experience and (2) mentally simulating how a course
vasoplegic, or depressed left ventricle appropriately of action would play out and simulating additional plans if the
generated in one situation may be incorrectly attached to first does not seem workable.43 The model suggests that in
subsequent episodes of care where they may not be true. making important decisions, humans have managed to
Many postoperative and critically ill patients pass through a incorporate some of the best elements of the cognitive
number of different physiologic states as they deteriorate and traditions described above: an initial use of representativeness
convalesce; maintaining an ability to perceive changes in and pattern recognition, satisficing, or recognizing that a
patients without the constraint of prior treatment decisions is workable solution is the main goal, and not feeling the
an important aspect of cognition in critical care. Recognizing necessity to find the best solution, and some mediation by
the strength of anchoring, a well-regarded author and conscious rational processes (mental simulation) without the
diagnostician remarked, when a case first arrives, I dont paralysis and inactivity brought by over-deliberation.42
want to hear anyone elses diagnosis. I look at the primary Kleins model came from extensive post-event probing of his
data.41 expert subjects, where a recurrent finding was a lack of
awareness on the part of the decision-makers that they were
Omission or Status Quo Bias describes the tendency to take a even making any decisions!
course of inaction or observation over intervention. Typical
intervene or not decisions in critical care include watchful The dual process model of reasoning postulates that human
waiting versus exploratory laparotomy and a decision to decisions involve use of both analytic and non-analytic
minimize fluid administration and accept declining renal processes that operate in parallel, with pattern recognition
function versus infusing fluids and risking an abdominal being the main mode for familiar situations and decisions and
compartment syndrome. Omission may derive from cultural analytic thinking taking over when the situation does not
factors such as the maxim to first do no harm, or aversion match up with prior experience.44,45 Kahneman39 elaborates on
to the perceived legal consequences of being the last physician these parallel processes in a recent book entitled Thinking
to intervene on a patient with a poor outcome.25 Omission Fast and Slow, where he uses System I and System II
bias was addressed in a survey of critical care physicians in respectively, to refer to intuitive and analytic processing of
which three clinical scenarios were presented, and respondents information.39 Accordingly, intuition is frugal, and analytic
were queried about what they would do when presented with thinking is costly in terms of energy and time consumption
options of either maintaining the status quo (a CT scan is and the level of concentration required to be effective. The
already ordered, do you want to cancel it?) or performing a hypothetical-deductive method is an example of an analytic
new action (physician asked whether s/he wanted to obtain a method used in medical practice. A unique characteristic of
CT scan for same patient). Physicians responses showed a System II is its potential ability to monitor the activity of
greater inclination to act on a plan that was already in motion System I.46 Like the recognition-primed decision model, dual
rather than change the course in pursuit of the ideal decision.37 processing has the potential to examine the impact of certain
actions and may protect the decision maker from suboptimal
Multiple Process Models choices. Both models seem to describe different formulations
Despite a longstanding interest in how individuals make of a common mental framework.
difficult decisions and the elucidation of reproducible areas of
weakness, it was not until the late 1980s that the line of A general model for dual process thinking is presented in
investigation moved from controlled laboratory investigations figure 1. While more elaborate depictions have been
of cognitive predispositions to direct observation of offered,44,47 we believe that a model analogous to fast and
experienced practitioners in natural settings. Gary Klein and slow AV nodal conduction pathways captures the essence of
associates42 followed fire chiefs, military commanders, and current thinking. A number of external and internal factors
similarly skilled members of high-risk professions and influences the overall function of these systems, can slow use
investigated the thought processes utilized when making of that particular pathway, and dictate that more processing
tough decisions under difficult conditions such as limited occur via the other. Internal factors include affect, stress,
time, uncertainty, high stakes, vague goals, and unstable fatigue, emotion and level of confidence, while external
conditions. The initial expectation was that key actions factors are uncontrolled variables such as patient acuity,
resulted from comparison of leading hypotheses or options location, availability of resources, and attention given to other
and that some type of normative analysis of risk/benefit thoughts and tasks. Factors such as sleep, substance use,
would yield the best choice. In reality, expert decisions were anger, and emotion compromise the function of deliberate
most commonly reached without this deliberation and were and rational decision making (System II) and create a higher
based on the subconscious ability to recognize and act on reliance on intuitive judgment. This construct gains additional
complex patternsan ability developed through years of support from functional magnetic resonance imaging studies

160 Decision making in critical care CM&R 2015 : 3-4 (December)

showing distinct anatomic loci for both deductive and
intuitive processing48 and selective absence of these faculties
in patients with related brain lesions.44

Decision Making in Critical Care

Despite the overall importance of understanding diagnostic
reasoning in critical care, work in this area has not accompanied
the longer-standing traditions of exploring the ethical,
affective, and legal dimensions of critical care decisions. The
daily variation in personnel, work intensity, and patient acuity
seen in the ICU creates a unique environment where decisions
and their underlying processes can be examined and perfected.
Figure 1. A schematic of dual cognitive processes in clinical
The practice of critical care often involves the need to make
decision making is shown. A situation or problem requiring a
rapid judgments in high-risk and dynamic situations. decision activates both pathways simultaneously; however, a
Recognition of disease patterns and their diversity, as well as path of least resistance is taken. Familiarity with the situation
instituting empiric therapy for unclear conditions is a typical favors the faster non-analytic pathway on the left (System I),
start to a patient encounter. Situations calling for quick action while novelty and complexity call for greater analysis and likely
are almost entirely guided by pattern recognition. Less use of System II. The thicker arrow on System I is used to
commonly, patients present with a variety of findings that are indicate a more direct connection and path of lesser resistance
hard to assemble into a coherent single diagnosis, and than that for System II. Plus and minus signs are shown
problems are worked on individually while more deliberate alongside factors that have been shown to positively or
acquisition and organization of information takes place. In the negatively influence the overall quality and output of each
system. The dashed arrow, whereby System II provides
latter case, System II is being used with a much higher
oversight of tentative conclusions reached by System I,
attentional demand.
represents metacognition.

It is worth emphasizing that diagnostic reasoning in intensive

care is an ongoing process that involves not only the which is in sufficient supply. There should be concern that
identification of the underlying disorder in a patient flawed use of heuristics may predominate the diagnostic
(pneumococcal sepsis) but also the ability to recognize process when novices experience cognitive overload. Urgency
changes in a number of physiologic sub-states (ventricular creates a state of arousal that can facilitate information
dysfunction, fluid mobilization, microvascular thrombosis) processing; however, higher degrees of stress drastically
that will alter outcome if not acted upon properly. reduce the number of external cues that can be properly
perceived and utilized.49,50 Accordingly, actions based on
There is a general sense that diagnostic errors are more likely reasoning or computations may be flawed or impossible to
to arise from occasional faults in the use of System I and that execute with too much stress. Expert practitioners can use
with a proper command of clinical evidence and its intuition to arrive at workable solutions, but with inadequate
significance (use of System II), errors may be prevented. In exposure to prototypic and variants of key cases it is highly
critical care, there may be an unappreciated opportunity to doubtful that novices can function in this manner.
improve diagnostic accuracy by conscious review of both
systems. For example, if an initial intervention creates a Flawed Diagnostic Reasoning: a True Case
period of stability, this may buy time to double-check A 65-year-old man with symptomatic CAD and congestive
assumptions, review primary data sources, and invite heart failure was admitted to the ICU after a routine stenting
consultants and others to share their views. In teaching of the right iliac artery. Postoperative ICU admission was
centers, team rounds and sign-overs provide multiple requested due to a history of stable CAD and congestive heart
opportunities to re-review a case. failure (CHF). His ejection fraction was reported to be 28%;
he had two admissions in the prior year for CHF exacerbations.
The diagnostic process described in the paragraph above is a
general overview of what one would likely see if surveying Eight hours after the procedure, the ICU team was called to
experienced practitioners in critical care. In contrast, trainees the patients bedside for a drop in systolic blood pressure to
may be overwhelmed during fast paced encounters where less than 90 mm Hg, after being in the 120s previously.
there is novelty in both patient problems and stress and Assuming a cardiac etiology, the intern and fellow placed a
urgency. Without adequate experience, decision making may pulmonary artery catheter to facilitate hemodynamic
be flawed on a number of fronts. Newer trainees are likely to monitoring and therapy for presumed heart failure. The blood
make better decisions through comparing and contrasting the pressure was 85/45 mm Hg, central venous and pulmonary
most common conditions that give rise to the patients artery pressures were 4 mm Hg and 25/8 mm Hg respectively;
problem, but this requires time and concentration, neither of the cardiac output (CO) was 3 L/min.

CM&R 2015 : 3-4 (December) Lighthall and Vazquez-Guillamet 161

If using the hypothetical-deductive model in our patient, the incorporated into their thinking and believed to affirm the
resident would have perhaps entertained a wider differential diagnosis. The house staff was apparently blind to evidence
diagnosis of shock in post-catheterization patients and may contradicting heart failure as a primary disturbance. It was
have decided to perform tight heart catheterization as an never addressed whether obtaining the ECG and troponins
empiric means of evaluating hemodynamic collapse. was a diagnostic maneuver arising from confirmation bias or
Accordingly, he would have recognized that the systemic part of a precompiled care routine for suspected CAD. Other
vascular resistance of 1250 was inconsistent with distributive team members seemed to be subjected to either availability or
shock and also noted that the low CO with low filling anchoring, as they failed to recognize contradictory evidence
pressures was inconsistent with cardiogenic shock in a patient when presented. Even when told that the patient was probably
with CHF. With an understanding of the salient features of bleeding from his catheterization site, the presenting intern
different shock states, the resident would have hopefully revised his diagnosis to cardiogenic shock with hypovolemia,
recognized the prototypic findings of hypovolemic shock. further demonstrating the power of anchoring, and the
With some healthy skepticism of catheter pressures, he would absence of analytical thought!
have looked for tests or maneuvers that differentiate
cardiogenic and hypovolemic shock. A change in blood How Can We Improve Our Decision Making?
pressure with passive leg raising, inspection of jugular The discussion of decision making has so far focused on the
venous pressure, or an echocardiographic evaluation of the apparent rift between the rational and methodical approach to
heart would have differentiated between fluid overload and diagnostic reasoning that typifies our teaching (System II)
hypovolemia and established the diagnosis of hypovolemic and the subconscious nature of decision making that
shock. Further evaluation for blood loss would have identified predominates most reasoning under uncertainty (System I).
the proximate cause. Intuition can be a source of flawed judgment and performance
or a means of demonstrating true mastery. Proper use of
In this case, cardiogenic shock provided a highly available System II is dependent upon the users understanding of
and easy way to explain hypotension in a patient with a diseases and their distinguishing features plus the time to plan
history of CAD and CHF. The precautionary ICU admission and reason through a diagnosis, and therefore, cannot be
was based on this premise and probably primed the house regarded as the solution to all diagnostic problems. As there
staff to look for this problem. Assuming this diagnosis, the are no determinants of when ones intuition becomes
resident failed to notice the abundance of refuting evidence. reliableespecially at intermediate levels of experienceit
Subsequently, he pursued confirmation of the diagnosis by seems important to consider means to offset areas of potential
ordering an electrocardiogram (ECG) and cardiac enzymes, weakness in this aspect of diagnostic reasoning. The section
and then ordered an infusion of epinephrine to help with the below will introduce strategies that can be employed in
low cardiac output. critical care units and related educational programs to
improve diagnostic accuracy, promote reflective practice, and
The resident believed the patients history, along with a low reduce harm from critical decisions.
cardiac index, was sufficient grounds to make a diagnosis of
cardiogenic shock. Satisfied with this etiology, an objective MetacognitionThinking about How We Think
consideration of the patients data and basic categories of Metacognition refers to the deliberate monitoring of cognitive
shock did not take place until morning rounds. When processes and their impact. In critical care, at least two
presenting the case during rounds, the on-call resident distinct benefits can emerge from a better understanding of
reasserted his initial hypothesis of cardiogenic shock multiple our thinking. First, by identifying the cognitive strategy used
timeseven after being asked about the low cardiac filling to reach a conclusion, the appropriateness of the strategy in a
pressures. Other trainees were subsequently asked about their particular context can be evaluated. In this way, errors may be
impression of the patients problem, and all believed that the trapped earlier in their evolution, before they cause harm.
story was consistent with cardiogenic shock. Authors in several fields have promoted this view.30,45,51-53
Second, an awareness and appreciation of thought processes
Comment: The diagnostic strategy used in this case is may provide extra motivation to follow up on their accuracy
remarkable for a number of reasons. First may be an in novel and unclear situations. With periodic feedback on
undeveloped appreciation for the prototypic shock states and cognitive performance, the ability to build expertise from
their differentiating factors. Equally apparent is a reliance on experience is bound to be greater.52,54
faulty heuristics. Availability bias is apparent in the initial
diagnostic impression. The intern and resident were not As a metacognitive strategy, Klein42,43 proposes the crystal
experienced in postoperative care, so they took the history of ball technique wherein an individual or group imagines they
heart failure and precautionary admission as evidence for this are looking into a crystal ball that says that their assessment
diagnosis without questioning its veracity. Search satisficing is wrong, then asks; if this is the case, then what would be
and early closure were apparent when even after a broader set the right answer? This is a particularly useful strategy in
of information was present, only the single piece of evidence critical care, as many episodes of care fluctuate between
consistent with cardiogenic shock (the low CO) was periods of action and planning. Once action has led to some

162 Decision making in critical care CM&R 2015 : 3-4 (December)

stability, critical examination of earlier impressions may lead systolic dysfunction is instantly apparent. Discovery of poor
one to arrive at a more accurate idea of the patients problem. cardiac function may lead to termination of the exam prior to
The crystal ball is another way of granting permission to finding that a severely stenotic aortic valve was really the key
ask: why do I not feel right about this decision? or what driver of hemodynamic compromise. An example of a forcing
are some leading diagnoses that we left out? The authors strategy would be to insist that use of echocardiography
found this technique a useful group exercise after new involve an evaluation of all relevant structures rather than
admissions, as team members begin to commit important spot examinations.
decisions such as making the diagnosis, prescribing a high-
risk test, or deciding on an operation. The exercise also sends Similar approaches to debiasing judgments made from other
an important cultural message that constructive dissent is heuristics are presented in table 2. Other generic forcing
important and that team members are expected to speak up strategies have been proposed, such as ruling out the worse
and put the patients interests first and not to take it personally scenario, deliberate consideration of three different diagnoses
when their ideas are challenged. or explanations of a problem, and a forced consideration of
new diagnoses if three interventions fail to create improvement
Groupthink and avoidance of group conflict are forms of (the rule of 3s).30,52 There has been limited experimental
sociologic bias that restrict independent thought and can be evaluation of forcing strategies, none of which have used a
remedied in part with an insistence on hearing divergent native clinical environment. The goal of the forcing strategies
views.55 This problem has not yet been explored in critical was for trainees to expand their differential diagnoses prior to
care; however, the general desire to achieve group consensus making final judgment. Both uncontrolled and controlled
on diagnoses and plans and other aspects of team building in studies have failed to demonstrate clinically or mathematically
the ICU may actually promote conformity rather than significant degrees of improvement following this type of
protection against early closure or other biases. intervention.56-58 One interpretation of the results is that the
experimental subjects (all students) may not necessarily use
A key attribute of the analytic decision-making pathway non-analytic processing as a primary decision strategy, and if
described above and in figure 1 is its ability to monitor the using System II, a ceiling effect may exist where further
results of conclusions reached by intuition. Awareness that a analysis may fail to improve accuracy. Indeed, for both
decision was reached subconsciously, however valid it may control and interventional arms of one CFS study, diagnostic
be, may prompt the cautious clinician to revisit the same accuracy was inversely related to the time spent on the
decision with a different set of tools or to seek a second diagnosis.58 In contrast, a series of studies conducted by
opinion. However, just because more methodical consideration Mamede have shown that structured reflection
of a clinical situation can potentially offset the pitfalls of systematically listing the supportive, contradictory, and
intuitive judgment does not necessarily mean that it will. absent findings of an initial and alternate diagnoses for a
Forces such as stress and fatigue that cause an initial reliance written caseimproved both learning and categorization of
on System I will also impair its oversight. illness scripts in medical students.59 The reflective intervention
was similarly tested in internal medicine residents where it
Flawed Diagnostic Reasoning: an Earlier Escape? The case of has some success in overcoming availability bias and
perceived cardiogenic shock was discussed on morning improving diagnostic accuracy.29
rounds, and correction of misjudgment was accomplished
fairly rapidly. After dispatching a fellow to revise the care Why is This a Good Decision?
plan, team members were asked whether any other impressions Intensive care rounds are filled with situations where a
crossed their mind. None were offered, so the team was led practitioner or trainees need to anticipate and treat problems.
through Kleins crystal ball exercise where they were told that Expert-level guidance in this process is helpful for learning
the crystal ball was telling us that the patient had a different what are expected and unexpected findings in the course of a
problem, and given this fact, what could be the right critical illness and how to act upon findings. For example, a
diagnosis? Quickly, other ideas were offered, with hemorrhage 38.5C fever in a patient who underwent an open repair of an
being the overall consensus. Hemodynamic data was revisited aortic aneurysm should trigger different thoughts and actions
in a manner that compared the patients findings with the depending on whether it is the first postoperative night or 3
prototypic forms of shock. days after the operation. Rounds also present an important
opportunity to explore the basis for anticipated actions. If he
Cognitive Forcing Strategies gets hypotensive, we will start norepinephrine. Statements
Forcing strategies involve the obligate pairing of one action along these lines should be followed with a query into why a
to another. Cognitive forcing strategies (CFS) refer to particular choice was made. For this particular example, we
developing the ability to self-monitor and recognize the use have heard explanations including because this is what worked
of specific heuristics in decision making, and then invoking last night on Mr. X, or I think he is septic, and norepinephrine
predetermined actions to counteract known pitfalls or is the drug of choice, or because all of the clinical data
vulnerabilities.52 For example, a portable echocardiogram is including central venous pressure, mixed venous oxygen
obtained to evaluate hemodynamic instability, and severe saturation, and my physical exam suggest he is vasoplegic. Was

CM&R 2015 : 3-4 (December) Lighthall and Vazquez-Guillamet 163

 Table 2. Heuristics and biases discussed in the text are presented alongside strategies for overcoming the
pitfalls inherent to each.
Heuristic Strategies to overcome systemic errors Reference
Compare disease with prototypes of the condition; be suspicious
when there is not a good match 25
Representativeness Be aware of relevant prevalence data 25,80
Incorporate discussion of common versus variant cases into
educational practices
Availability Seek base rate of a diagnosis
Be aware of external influences in diagnostic process 25
Premature closure Use tests or other exercises to refute diagnosis 25,43
Differentiate working vs. final diagnoses 25
Anchoring Insist on first hearing the story rather than someone elses diagnosis 25,41
Assess all key elements of the case 41
Avoid confirmation and early closure; make point of lab tests to 43
prove other leading diagnoses. Crystal ball exercise
Confirmation Use hypothetical-deductive method to assess value and role of
contemplated tests
Try to disprove your diagnosis, consider conditions of higher
prevalence 25,43
Stick to institutional protocols or standard methods when applicable 36
Be aware of affective, ethical and social influences
Reaffirm role of intensivist as independent consultant; avoid influence
of referring physicians
Be aware that alternate descriptions of the same situation may elicit
Unpacking or support different estimates of likelihood. 33
theory Unpack broad categories and compare alternatives at similar levels of
specificity
Satisficing Perform secondary searches once it appears the answer is at hand.
Complete standard protocols where applicable. 25
Omission or status quo Think of what others would do. Consider existing clinical evidence. 37
bias Trust intuition: if you thought of it, you should probably do it 25
Try to approach each case as a clean slate, as in de-anchoring

it availability? i.e. the previous cardiac surgery patient was accelerated deterioration, where a quick exploration of
given norepinephrine when she became hypotensive. Was it assumptions and key data sources is warranted. Group
anchoring? (the resident heard the patient needed norepinephrine demonstration of metacognition may facilitate development
in the operating room), or was it representativeness or physiologic of this important ability.
reasoning that found that the patients overall physiologic profile
looked like vasoplegia, so norepinephrine was chosen? In this Exhaustive Strategies
example, discussion of the decision strategy would reinforce Given the risks of early closure, could we protect against
that deciding on a vasoactive agent is not an act of pairing a misdiagnosis and achieve a broader perspective by obtaining
disease process and agent, but an act of defining a physiologic a more comprehensive battery of laboratory and diagnostic
picture and adding what is missing. It has been advocated that tests? Studies that have looked at this question suggest that
clinical conferences and other educational activities should be this may be an overly optimistic view.61,62 Why is this the
expanded to include an exploration of decision strategies that are case? First, most can form a diagnostic impression early on in
encountered in the course of caring for a patient.60 an encounter, and adding additional information rarely
changes the diagnosis but prolongs evaluation and is
Instructors should identify critical decisions or junctures in financially costly. Acquiring objective information in the
care, essentially triggers such as new vasoactive drug, higher form of diagnostic tests may reflect confirmation bias more
oxygen requirement, new metabolic abnormality or than a true hypothetical deductive strategy.62,63 Exhaustive

164 Decision making in critical care CM&R 2015 : 3-4 (December)

strategies often lack the guidance of experience or intuition hypothetical-deductive strategy can theoretically overcome
and, hence, the necessary perspective to assemble all findings some of the problems arising from intuitive judgments;
into a sensible picture. However, even experienced ICU however, its use can be overridden with biases just as strong
physicians have exhibited great variability in how they as those it attempts to overcome. In acquiring diagnostic tests,
approach the same patient in terms of information gathering, careful attention to test characteristics and disease probability
interpretation, and eventul decisions.64 will provide the greatest yield.

Diagnostic impressions based on pattern recognition The physical examination may languish as a detached set of
classically neglect the base-rate probabilities of candidate routines in some trainees practice. Showing the interplay
diagnoses. Ironically, an exhaustive strategy that strives to between hypothesis formation, gathering of information, and
eliminate bias through completeness can utilize tests in an its ability to change probabilities may help validate the use of
improper manner. For example, in populations with a low the history and physical as an efficient diagnostic tool, and
prevalence of a disease in question, laboratory tests will yield not an exhaustive and antiquated exercise.70,71 Finally, many
more false positives than true positives.65 Kirch and Schafii66 critical care trainees organize thoughts and data into organ-
reported that in a series of autopsies, the history and physical based categories that serve the interests of completeness but
examination led to the correct diagnosis 60% to 70% of the may oppose reflection upon key problems and findings.
time, in contrast to only 35% of diagnoses being possible with Demanding clarity in data acquisition and synthesis during all
the use of imaging techniques. It is attractive to think that if phases of patient care is likely to strengthen both analytic and
biased thinking led to an erroneous conclusion, means to intuitive modes of thought.
expand the differential diagnosis, and essentially place the
correct answer before our eyes, should improve accuracy. Acknowledge the Use and Value of Intuition
Even with a broader information set, clinicians remain In many situations, the System I process provides the most
attached to the first diagnosis considered (even if wrong)67,68 effective starting point for diagnosis and action. For example,
over diagnoses that were suggested by others.38 an experienced critical care physician may walk into the room
of a patient with variceal bleeding and launch right into
The exhaustive strategy is often tolerated as an initial step in intubation and transfusion of blood products, while a resident
learning to care for medical patients. It is perhaps believed in the same room may be at a loss for the specific criteria
that hypothesis development must wait until further experience supporting these actions. As such, we owe it to our trainees to
is gained. Proper guidance regarding the relevance and use of reconstruct our actions as completely as we can and to show
specific studies is essential in preventing getting lost in the how largely non-analytic judgments result from understanding
numbers. Also, without the proper experience base, one may the significance and trajectory of seeming unrelated signals.
run the risk of falling into the trap of pseudodiagnosticity.
The latter term refers to building a diagnosis on laboratory or It is our belief that some basic distinguishing patterns for
historical finding that may not address the patients primary common critical care and postoperative care problems can be
problem.63 For example, a resident recently was operating taught and can serve as reference points for comparison with
under the belief that an elevated lactate level was diagnostic variants and to contrast with other disease prototypes.26,72
for severe sepsis in a patient who, in fact, had a large Revisitation of the different physiologic patterns of shock states
gastrointestinal bleed. Because obtaining lactate levels is part and their correspondence to different trends and findings on
of the workup for sepsis, the resident misapplied this bedside monitors and physical examination is one example of
association to arrive at an incorrect diagnosis. Thus, there building intuition and pattern recognition around an accepted
seems to be no proof that expanding the information set theoretical framework.73 Support for use of pattern recognition as
induces a more deliberative approach to decision making or an instructional strategy comes from a study involving the
that doing so will overcome the biases of non-analytic interpretation of ECGs by psychology graduate students, wherein
judgment. students were taught different patterns and used these to form
hypotheses for the ECGs principal abnormality. Participants, all
Demonstrate Proper Use of the Hypothetical-Deductive Methods of whom lacked domain specific knowledge, were able to
When ordering diagnostic tests and imaging, it is important produce more accurate results when operating from a combined
for practitioners and trainees to understand and articulate both analytic/non-analytic method than those who were taught a more
the proper use and pitfalls of the studies. Most diagnostic deliberate form of analysis.74,75 This study, along with Mamedes
studies cannot deliver omniscient pronouncements of a use of reflection in diagnosis were designed with a dual
diseases presence or absence. Rather, most are deployed processing framework of reasoning in mind, which may offer a
clinically after cut-off values for positive and negative results more robust approach to understanding the diagnostic process
have been established and have demonstrated some utility in than what Norman criticizes as the didactic presentation of a
modifying pre-test probabilities in specific patient series of cognitive biases in which the instructional goal can
populations.69 Failing to understand these statistical factors in easily be misinterpreted as one of remembering definitions.56
ordering tests may lead to false conclusions, thereby adding
more confusion than clarity to the diagnostic process. Thus a

CM&R 2015 : 3-4 (December) Lighthall and Vazquez-Guillamet 165

Establish the Best Conditions for Advanced Learning carry the potential of creating a safe atmosphere for learning,
Clearly, some situations in critical care call for decisions that wherein challenging situations and adverse outcomes can be
emerge with little deliberation, while others require an explored without evoking the defensive behavior that may
analytic strategy. For intuition to develop properly, Kahneman accompany poor outcomes in real patients.
and Klein46 both agree that the context for learning must be
one of high validity, as defined by generally predictable Review Errors; Learn From Them
elements and the ability to receive feedback on the quality Intensive care admissions can result from errors made in the
and outcome of decisions. Thus, if we expect our trainees to course of patient evaluation and treatment elsewhere. For all
advance in their problem solving and decisional skills, they phases of care, identification of errors offers the opportunity
first must be grounded to some sense of what is typical and to explore decision-making strategies and their relationship to
expected for a particular disease state. Faculty, therefore, may outcomes. To maximize the yield of analyzing near misses
have to exercise greater oversight into how cases are assigned and mistakes, it is important that critical care leaders create an
and see that students and early residents initially take care of atmosphere of inquiry (rather than blame) that allows for
patients that have prototypical presentations of common understanding the cognitive and systematic roots of errors.
problems or routine postoperative courses. To this foundation,
more complex cases or classic variants can be added as Conclusion: Towards a Better Understanding of
experience increases.26 Diagnostic Reasoning in Critical Care
The role of cognition in diagnostic reasoning has been an area
Deliberate exploration of the methods underlying diagnostic of interest in Emergency Medicine for over 15 years and is
reasoning should be explored as already noted. Insistence on gaining attention in anesthesiology and critical care; however,
reviewing primary data is also important for general skill examination of the use and validity of different decision
development, as well as for providing some safeguard against pathways is in its infancy. Studies should be conducted to
early anchoring and closure. While managing acute situations, understand the cognitive processes used by skilled and
the practice of thinking out loud is important for generating a seasoned physicians in critical care and to understand how
shared mental model amongst other team members and often stressors and other variables impact their performance. The
invites others to contribute their thoughts and impressions. latter will help overcome some of the questions raised from
Group leaders should also find means of promoting studies on students and early residents.
constructive questioning of key assumptions and conclusions;
generating the ethic that this is important backup for innate Rather than regarding different strategies as either friend or
errors in cognition rather than being right is absolutely foe, it may be more worthwhile starting with an open-ended
essential. Patient sign-overs may present an additional approach to understand what differentiates high and low
opportunity to revisit diagnostic reasoning. Recently Dhaliwal levels of performance. Exposure of a cohort of trainees to a
and Hauer76 proposed that the re-presentation of cases standard of septic shock and development of a validated
admitted by night shift or other cross-covering physicians scoring tool revealed a wide range of technical and behavioral
become a time to review the diagnostic reasoning of the performances.79 Similar studies can be conducted with
admitting physician. This practice would expose trainees and emphasis on evaluating what aspects of decision making
their supervisors to current thoughts on cognition and underlie variations in performance. Creation of educational
cognitive error. Success of this intervention again is dependent programs around the strategies associated with the best
on a positive view of dissent, and according to the authors, is diagnostic performance will provide trainees with models of
dependent upon the attending physicians ability to model expertise and its nature at an earlier part of their training.
how such commentary can unfold constructively and
professionally.76 The concentration of errors in critical care certainly mirrors
the high burden of illness and number of interventions
Patient simulators offer yet another tool for exploring decision experienced by this population but also generates some
making in critical care.77,78 Standardized scenarios of known questions of whether physician factors synergize with the
problems can be presented to trainees, and their actions can patients states in a manner that amplifies risk and whether
be compared with what instructors consider ideal approaches there are means by which this can be minimized. The face
to the condition. The time and place for different diagnostic validity of the psychological concepts of judgment under
strategies (ie, pattern recognition versus physiologic principles uncertainty to critical care seems high; however, naturalistic
versus probabilistic thinking) can be explored in a controlled studies to back up these assumptions are essential to any
setting. When errors are encountered, the evolution can be effort designed to improve patient safety in the ICU.
elucidated from participants recollections, and unlike the
ICU, true actions can be captured with video recordings and
reviewed. Simulation exercises guarantee that a cohort of
standing practitioners or trainees will be confronted with the
same challenges and opportunities to understand decisions
and decision-making strategies. Simulation facilities also

166 Decision making in critical care CM&R 2015 : 3-4 (December)

References 22. McGloin H, Adam SK, Singer M. Unexpected deaths and
1. Croskerry P, Norman G. Overconfidence in clinical decision referrals to intensive care of patients on general wards. Are
making. Am J Med 2008;121:S24-S29. some cases potentially avoidable? J R Coll Physicians Lond
2. Berner ES, Graber ML. Overconfidence as a cause of diagnostic 1999;33:255-259.
error in medicine. Am J Med 2008;121:S2-S23. 23. McQuillan P, Pilkington S, Allan A, Taylor B, Short A, Morgan
3. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, G, Nielsen M, Barrett D, Smith G, Collins CH. Confidential
Lawthers AG, Newhouse JP, Weiler PC, Hiatt HH. Incidence inquiry into quality of care before admission to intensive
of adverse events and negligence in hospitalized patients: care. BMJ 1998;316:1853-1858.
results of the Harvard Medical Practice Study I. N Engl J 24. Kovacs G, Croskerry P. Clinical decision making: an
Med 1991;324:370-376. emergency medicine perspective. Acad Emerg Med
4. Leape LL, Brennan TA, Laird N, Lawthers AG, Localio AR, 1999;6:947-952.
Barnes BA, Hebert L, Newhouse JP, Weiler PC, Hiatt H. The 25. Croskerry P. Achieving quality in clinical decision making:
nature of adverse events in hospitalized patients. Results of cognitive strategies and detection of bias. Acad Emerg Med
the Harvard Medical Practice Study II. N Engl J Med 2002;9:1184-1204.
1991;324:377-384. 26. Bowen JL. Educational strategies to promote clinical diagnostic
5. Vincent C, Neale G, Woloshynowych M. Adverse events in reasoning. N Engl J Med 2006;355:2217-2225.
British hospitals: preliminary retrospective record review. 27. van Schaik P, Flynn D, van Wersch A, Douglass A, Cann P.
BMJ 2001;322:517-519. Influence of illness script components and medical practice
6. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby on medical decision making. J Exp Psychol Appl
L, Hamilton JD. The Quality in Australian Health Care 2005;11:187-199.
Study. Med J Aust 1995;163:458-471. 28. Tversky A, Kahneman D. Judgment under Uncertainty:
7. Weingart SN, Wilson RM, Gibberd RW, Harrison B. Heuristics and Biases. Science 1974;185:1124-1131.
Epidemiology of medical error. BMJ 2000;320:774-777. 29. Mamede S, van Gog T, van den Berge K, Rikers RM, van
8. Bates DW, Miller EB, Cullen DJ, et al. Patient risk factors for Saase JL, van Guldener C, Schmidt HG. Effect of availability
adverse drug events in hospitalized patients. ADE Prevention bias and reflective reasoning on diagnostic accuracy among
Study Group. Arch Intern Med 1999;159:2553-2560. internal medicine residents. JAMA 2010;304:1198-1203.
9. Giraud T, Dhainaut JF, Vaxelaire JF, Joseph T, Journois D, 30. Stiegler MP, Neelankavil JP, Canales C, Dhillon A. Cognitive
Bleichner G, Sollet JP, Chevret S, Monsallier JF. Iatrogenic errors detected in anaesthesiology: a literature review and
complications in adult intensive care units: a prospective pilot study. Br J Anaesth 2012;108:229-235.
two-center study. Crit Care Med 1993;21:40-51. 31. Heller RF, Saltzstein HD, Caspe WB. Heuristics in medical and
10. Graf J, von den Driesch A, Koch KC, Janssens U. Identification non-medical decision-making. Q J Exp Psychol A
and characterization of errors and incidents in a medical 1992;44:211-235.
intensive care unit. Acta Anaesthesiol Scand 2005;49:930-939. 32. Sargent G. Driving versus flying: the debate is settled. The
11. Thomas EJ, Brennan TA. Incidence and types of preventable New York Observer 1998.
adverse events in elderly patients: population based review of 33. Redelmeier DA, Koehler DJ, Liberman V, Tversky A.
medical records. BMJ 2000;320:741-744. Probability judgement in medicine: discounting unspecified
12. Graber ML. The incidence of diagnostic error in medicine. possibilities. Med Decis Making 1995;15:227-230.
BMJ Qual Saf 2013;22 Suppl 2:ii21-ii27. 34. Simon HA. Rational choice and the structure of the
13. Saber Tehrani AS, Lee H, Mathews SC, Shore A, Makary MA, environment. Psychol Rev 1956;62:129-138.
Pronovost PJ, Newman-Toker DE. 25-Year summary of US 35. Elstein AS. Heuristics and biases: selected errors in clinical
malpractice claims for diagnostic errors 1986-2010: an reasoning. Acad Med 1999;74:791-794.
analysis from the National Practitioner Data Bank. BMJ Qual 36. Pines JM. Profiles in patient safety: confirmation bias in
Saf 2013;22:672-680. emergency medicine. Acad Emerg Med 2006;13:90-94.
14. Weissman JS, Schneider EC, Weingart SN, Epstein AM, David- 37. Aberegg SK, Haponik EF, Terry PB. Omission bias and
Kasdan J, Feibelmann S, Annas CL, Ridley N, Kirle L, Gatsonis decision making in pulmonary and critical care medicine.
C. Comparing patient-reported hospital adverse events with Chest 2005;128:1497-5105.
medical record review: do patients know something that 38. van den Berge K, Mamede S, van Gog T, Romijn JA, van
hospitals do not? Ann Intern Med 2008;149:100-108. Guldener C, van Saase JL, Rikers RM. Accepting diagnostic
15. Pierluissi E, Fischer MA, Campbell AR, Landefeld CS. suggestions by residents: a potential cause of diagnostic error
Discussion of medical errors in morbidity and mortality in medicine. Teach Learn Med 2012;24:149-154.
conferences. JAMA 2003;290:2838-2842. 39. Kahneman D. Thinking Fast and Slow. New York: Farrar,
16. Wu AW, Folkman S, McPhee SJ, Lo B. Do house officers learn Straus and Giroux; 2011.
from their mistakes? JAMA 1991;265:2089-2094. 40. Englich B, Mussweiler T, Strack F. Playing dice with criminal
17. Kruger J, Dunning D. Unskilled and unaware of it: how difficulties sentences: the influence of irrelevant anchors on experts
in recognizing ones own incompetence lead to inflated self- judicial decision making. Pers Soc Psychol Bull
assessments. J Pers Soc Psychol 1999;77:1121-1134. 2006;32:188-200.
18. Spencer FC. Human error in hospitals and industrial accidents: 41. Groopman J. How Doctors Think. New York: Houghton
current concepts. J Am Coll Surg 2000;191:410-418. Mifflin Co.; 2007.
19. Donchin Y, Gopher D, Olin M, Badihi Y, Biesky M, Sprung 42. Klein G. Naturalistic decision making. Hum Factors
CL, Pizov R, Cotev S. A look into the nature and causes of 2008;50:456-460.
human errors in the intensive care unit. Crit Care Med 43. Klein G. Sources of Power, How People Make Decisions.
1995;23:294-300. Cambridge, MA: MIT Press; 1998.
20. Hart GK, Baldwin I, Gutteridge G, Ford J. Adverse incident 44. Croskerry P. Clinical cognition and diagnostic error:
reporting in intensive care. Anaesth Intensive Care applications of a dual process model of reasoning. Adv
1994;22:556-561. Health Sci Educ Theory Pract 2009;14 Suppl 1:27-35.
21. Osmon S, Harris CB, Dunagan WC, Prentice D, Fraser VJ, 45. Elstein AS, Schwartz A. Clinical problem solving and
Kollef MH. Reporting of medical errors: an intensive care diagnostic decision making: selective review of the cognitive
unit experience. Crit Care Med 2004;32:727-733. literature. BMJ 2002;324:729-732.

CM&R 2015 : 3-4 (December) Lighthall and Vazquez-Guillamet 167

46. Kahneman D, Klein G. Conditions for intuitive expertise: a 72. Marcum JA. An integrated model of clinical reasoning: dual-
failure to disagree. Am Psychol 2009;64:515-526. process theory of cognition and metacognition. J Eval Clin
47. Stiegler MP, Tung A. Cognitive processes in anesthesiology Pract 2012;18:954-961.
decision making. Anesthesiology 2014;120:204-217. 73. Lighthall G. Use of physiologic reasoning to diagnose and
48. Epstein S. Integration of the cognitive and the psychodynamic manage shock States. Crit Care Res Pract;2011:105348.
unconscious. Am Psychol 1994;49:709-724. 74. Ark TK, Brooks LR, Eva KW. Giving learners the best of both
49. Ozel F. Time pressure and stress as a factor during emergency worlds: do clinical teachers need to guard against teaching
egress. Safety Science 2001;8:95-107. pattern recognition to novices? Acad Med 2006;81:405-409.
50. Wickens C, Stokes A, Barnett B, Hyman F. The effects of stress 75. Eva KW, Hatala RM, Leblanc VR, Brooks LR. Teaching from
on pilot judgment in a MIDID simulator. In: Svenson O, the clinical reasoning literature: combined reasoning
Maule J, eds. Time pressure and stress in human judgment and strategies help novice diagnosticians overcome misleading
decision making. New York: Plenum Press; 1991. 271-292. information. Med Educ 2007;41:1152-1158.
51. Croskerry P. The cognitive imperative: thinking about how we 76. Dhaliwal G, Hauer KE. The oral patient presentation in the era
think. Acad Emerg Med 2000;7:1223-1231. of night float admissions: credit and critique. JAMA
52. Croskerry P. Cognitive forcing strategies in clinical 2013;310:2247-2248.
decisionmaking. Ann Emerg Med 2003;41:110-120. 77. Lighthall GK, Barr J, Howard SK, Gellar E, Sowb Y, Bertacini
53. Graber M. Metacognitive training to reduce diagnostic errors: E, Gaba D. Use of a fully simulated intensive care unit
ready for prime time? Acad Med 2003;78:781. environment for critical event management training for
54. Flavell J. Metacognition and Cognitive Monitoring: a new area internal medicine residents. Crit Care Med 2003;31:2437-
of cognitive-developmental inquiry. American Psychologist 2443.
1979;34:906-911. 78. Bond WF, Deitrick LM, Arnold DC, Kostenbader M, Barr GC,
55. Janis IL. Groupthink: Psychological studies of policy decisions Kimmel SR, Worrilow CC. Using simulation to instruct
and fiascoes. 2nd ed. Boston: Houghton Mifflin Co.; 1983. emergency medicine residents in cognitive forcing strategies.
56. Norman GR, Eva KW. Diagnostic error and clinical reasoning. Acad Med 2004;79:438-446.
Med Educ 2010;44:94-100. 79. Ottestad E, Boulet JR, Lighthall GK. Evaluating the
57. Sherbino J, Dore KL, Wood TJ, Young ME, Gaissmaier W, management of septic shock using patient simulation. Crit
Kreuger S, Norman GR. The relationship between response Care Med 2007;35:769-775.
time and diagnostic accuracy. Acad Med 2012;87:785-791. 80. Klein G. Streetlights and shawdows: the key to adaptive
58. Sherbino J, Kulasegaram K, Howey E, Norman G. decision making. Cambridge, MA: MIT Press; 2009.
Ineffectiveness of cognitive forcing strategies to reduce
biases in diagnostic reasoning: a controlled trial. CJEM Author Affiliations
2014;16:34-40.
59. Mamede S, van Gog T, Moura AS, de Faria RM, Peixoto JM, Geoffrey K. Lighthall, MD, PhD* and Cristina Vazquez-
Rikers RM, Schmidt HG. Reflection as a strategy to foster Guillamet, MD
medical students acquisition of diagnostic competence. Med
Educ 2012;46:464-472. *
Associate Professor, Department of Anesthesia, Stanford
60. Kassirer JP. Teaching clinical reasoning: case-based and University School of Medicine, Stanford, California USA
coached. Acad Med 2010;85:1118-1124.
61. Redelmeier DA, Shafir E, Aujla PS. The beguiling pursuit of

Fellow, Critical Care Medicine, Stanford University School
more information. Med Decis Making 2001;21:376-381. of Medicine, Stanford, California USA
62. Gruppen LD, Wolf FM, Billi JE. Information gathering and
integration as sources of error in diagnostic decision making.
Med Decis Making 1991;11:233-239.
63. Bordage G. Why did I miss the diagnosis? Some cognitive
explanations and educational implications. Acad Med
1999;74:S138-S143.
64. Kostopoulou O, Wildman M. Sources of variability in uncertain
medical decisions in the ICU: a process tracing study. Qual
Saf Health Care 2004;13:272-280.
65. Sackett D, Haynes R, Guyatt G, Tugwell P. Clinical
Epidemiology, a basic science for clinical medicine. Boston:
Little, Brown & Co.; 1991. 69-152.
66. Kirch W, Schafii C. Misdiagnosis at a university hospital in 4
medical eras. Medicine (Baltimore) 1996;75:29-40.
67. Cunnington JP, Turnbull JM, Regehr G, Marriott M, Norman
GR. The effect of presentation order in clinical decision
making. Acad Med 1997;72:S40-S242.
68. Leblanc VR, Norman GR, Brooks LR. Effect of a diagnostic
suggestion on diagnostic accuracy and identification of
clinical features. Acad Med 2001;76:S18-S20.
69. Soreide K, Korner H, Soreide JA. Diagnostic accuracy and
receiver-operating characteristics curve analysis in surgical
research and decision making. Ann Surg 2011;253:27-34.
70. Trowbridge RL. Twelve tips for teaching avoidance of
diagnostic errors. Med Teach 2008;30:496-500.
71. Vazquez R, Gheorghe C, Kaufman D, Manthous CA. Accuracy
of bedside physical examination in distinguishing categories
of shock: a pilot study. J Hosp Med;5:471-474.

168 Decision making in critical care CM&R 2015 : 3-4 (December)