The Role of Certainty, Confidence, and Critical Thinking in the Diagnostic Process: Good Luck or Good Thinking?

The diagnostic process in emergency medicine is often less than elegant. Unlike the clean, straightforward descriptions in text books, real patients come packaged with all sorts of challenges. In the resident portfolio entry by Dr. Caraballo, he colorfully describes the mental process of sorting, filtering, and interpreting an assorted collection of facts and observations into a meaningful construct, eventually a diagnosis, and finally a plan of action.1 This process is often imperfect, and the ability to recognize these imperfections is an important step in figuring out how to avoid mistakes and optimize care. The “facts” of a case are often not straightforward. History-taking produces a narrative account subject to interpretation, not a checklist of facts that easily fit into preformed templates (despite our efforts to make them do so). The art of medicine requires us to filter out “noise” to appreciate the signal; the ability to discern relevant from distracting signals is a key skill in clinical practice. Kathryn Montgomery describes this in her observations of medical students, explaining how medical student charts are filled with “facts” but often miss the “big picture.”2 The process of filtering requires interpretation and judgment and can be subject to cognitive bias.3 The potential for affective bias is particularly strong in emergency medicine.4 In this resident portfolio, the author demonstrates a keen awareness of the potential influence of cognitive bias in clinical judgment. He is quick to recognize the pitfall of thinking that a melodramatic patient is over-reacting and probably not sick, trusting his instinct that the patient “looked sick” despite the relative indifference of a paramedic, a nurse, and a tired consultant. He carefully avoids the “psych out” bias (attributing a complaint to a behavioral or psychiatric condition). When he encounters an unusual condition, he avoids “zebra retreat” (the tendency to avoid an uncommon diagnosis) as well as the “availability bias” (the tendency to consider something recent in one’s experience). He acknowledges how his confidence wavered when others failed to see cause for his concern. His introspection reveals insight into how cognitive bias can influence perception and judgment, and helps him avoid underestimating a serious illness. The knowledge and skill of recognizing potential cognitive traps is useful content for resident training and deserves reinforcement. This resident portfolio is an excellent example of reflective writing and mindful practice, a process in which one unpacks a series of events, carefully reconstructing each decision and thought, analyzing what types of factors swayed the author’s impressions and actions.5-8 There is much to applaud in his story. Although inexperienced, this clinician had a drive to “get it right” and to persist in his clinical questions, even if feeling lonely in the quest. The unwillingness to give up and the tenacity to continue thinking, questioning, and asking for help—even the gall to awaken a reluctant consultant—are traits we can admire. Even more, he followed this patient through his lengthy hospital course to see the follow-up 45 days later. Additional time spent in reflection also provides opportunity for the resident to fix his knowledge gaps. Many questions in this case deserve more thought. What are the best tests to establish the diagnosis of Boerhaave’s syndrome? Chest x-ray? Computed tomography? Esophagram? If contrast is necessary, should one use gastrografin or barium? Is direct esophagoscopy safe? When is intervention (surgery) indicated? When is conservative treatment adequate? The answers to these questions can help the author plan an approach for the next case or share it with others who can benefit from his experience. The author’s story is a fascinating account of an interesting case, but his analysis reveals how precarious the diagnostic process can be and how uncomfortable the process can become. Even having successfully managed an unusual life-threatening illness, his writing reveals a lingering dissatisfaction and unease that is rather telling. He writes openly of how unsettling the process was, how he grappled with indecision, uncertainty, and a temporary loss of a sense of control. Is there more that can be gained from his introspection? There are other valuable lessons not fully crystallized in the author’s writing. The instinct that served him well at the start of the case, the carefully honed and highly valued intuition that emergency physicians rely on to recognize and manage emergencies, was insufficient to carry him when tests were unrevealing. When the diagnosis was not so easily forthcoming, what then? The pressures mounting from the persistent influx of new patients, and the inability to solve this diagnostic conundrum, left him overwhelmed. He was stuck! He needed additional cognitive strategies. The dual process model of reasoning can help address his quandary.9 The model describes two modes of reasoning: System I and System II. System I describes the intuitive model in which experts recognize common patterns and react almost effortlessly in a clear and decisive manner. Emergency physicians dwell in System I whenever possible. Simple, decisive, quick! However, when cases are neither simple nor clear, or when a diagnosis lies outside one’s experience, it becomes necessary to resort to a more analytical and thorough approach, consider alternative diagnoses, and think through strategies to rule-in/rule-out multiple possibilities (System II). System II is painful for emergency physicians; it requires mental effort, perhaps a few more tests, and a greater investment of time. Yet the few quiet moments of thought that System II requires may well be gained back by clearing one’s mind to plot a new course and make a plan that will likely be more efficient than waiting and hoping for an answer to appear by chance or fortune. Once a plan is set, it is easier to reengage other tasks and other patients, helping turn chaos once again to order. Trainees need to spend time in System II until they manage a wide range of conditions. With time and experience, conditions that at first require System II thinking can be relegated to System I. How might the cognitive process have been improved in this case? While the author was aware of some of the traps of cognitive bias, he suffers from others, particularly anchoring (the tendency to lock on to features early on in the course of a presentation … although in this case he was correct), premature closure (accepting a diagnosis before it is verified), confirmation bias (the tendency to look for evidence to support rather than refute a given diagnosis), and the unpacking principle (failure to seek all relevant information or limit history taking).10 He admits actively ignoring portions of the history that he decided were irrelevant, facts that may have provided critical clues. While he celebrates his good fortune that his first thought was correct, it might not be next time. Ultimately, these common biases can have catastrophic consequences when serious illnesses are not considered or critical information is ignored. A metacognitive check, a brief foray into System II thinking, might give rise to important questions: “Just how good is a screening x-ray for this diagnosis? What other tests might prove/disprove the diagnosis? If not a perforated viscus, what else might explain this presentation?” This case required a workup better directed at the specific diagnosis under concern or a broader search for other potentially devastating diagnoses (myocardial infarction, aortic dissection, mesenteric ischemia, or volvulus). And in fact, in this case, it was really only when the author allowed himself to consider an alternative diagnosis that he stumbled into the test that confirmed the diagnosis. System I thinking helped keep this author convinced that the patient was sick. In that regard, intuition served him well. But System II was required to solve the diagnosis and might have helped him plot a methodical course with a greater sense of control and perhaps provide him a language to frame better questions for his consultant. Rather than wrestling with self-doubt and a sense of losing control of the ED, it could have given him improved focus and a more defined direction. The reflective process can be helpful, but should be guided to avoid incomplete or wrong conclusions. The author’s conviction that his intuition served him well is only partly correct. Fixated in System I left him struggling for several hours until a broader differential and a look beyond his first impression helped him actually find the answer. The resident’s reflection offers several lessons. Trust your instincts but recognize their limitations. Your instincts embody the sum total of your experience and training and should not be disregarded; however, they can (and should) be tested and verified. Awareness of cognitive bias will help avoid common mistakes. In the face of inexperience and indecision, recognize the need to resort to System II thinking: seek other methods to prove/disprove the diagnosis or consider alternate diagnoses. Thinking about clinical decision-making and formulating strategies for handling uncertainty are important. Reflective writing is useful and provides valuable insight, but should be guided by experienced mentees who can offer additional perspective and make sure that the lessons gained are complete and balanced.