Diagnostic errors, defined as failures to reach timely and accurate explanations for a patient’s presentation, affect the majority of our patients, because most will experience at least one diagnostic error in their lifetime.1 We report the case of a patient who was inappropriately diagnosed with acute bacterial rhinosinusitis and later found to have lymphoma. We use this case as a paradigm for introducing the role that diagnostic errors and cognitive biases play in clinical care, and discuss strategies for minimizing them.
A 72-year-old woman with no significant medical history presented to the hospital with a rapidly enlarging neck mass. Four months earlier, she was seen at an urgent care clinic with otalgia and hearing loss, and the diagnosis of acute bacterial rhinosinusitis was made. Her symptoms persisted despite antibiotic therapy, and over the next 4 months, she was treated with 3 additional antibiotic courses. She then noticed left-sided facial fullness, prompting presentation to the hospital. Biopsy confirmed the diagnosis of lymphoma, and further imaging showed diffuse disease. Her disease was refractory to multiple chemotherapy regimens, although she is now receiving immunomodulatory therapy with improvement in disease burden.
A better understanding of the diagnostic process can help minimize errors such as those made in this case. The diagnostic process relies on patient-, systems-, and clinician-related factors. The clinician-related factors include aspects such as knowledge base and the reasoning process, the latter of which is susceptible to cognitive biases.
We identify several cognitive biases that likely played a role in this case. Premature closure, or the acceptance of a diagnosis before its verification, was likely implicated, because the patient was diagnosed with acute bacterial rhinosinusitis despite the lack of cardinal features such as fever or facial pain. Because of availability bias, which is the error of making diagnoses on the basis of recent experience, providers are less likely to consider diagnoses that are less prevalent in their practice.1, 2 Certainly, acute bacterial rhinosinusitis is more commonly encountered than lymphoma at urgent care centers. In addition, anchoring, or the maintenance of one’s initial diagnosis despite new findings, may explain the prescription of multiple antibiotic courses without reevaluation of the diagnosis in a timely fashion.1, 2
Identifying cognitive biases requires a better understanding of the reasoning process. This process can be categorized as system 1 or system 2 thinking, also known as “fast” and “slow,” respectively.3 System 1 thinking is quick, relying on pattern recognition, heuristics, and prior experiences. Meanwhile, system 2 is analytic, deliberate, and time-consuming, involving techniques such as hypothesis generation and diagnostic modification.1 Both methods play key roles; using system 2 thinking alone would be taxing and unsustainable in our high-volume health care system, whereas system 1 thinking may be more error prone. Therefore, it is important to determine the appropriate uses for each strategy, optimizing efficiency and accuracy.
To read this article in its entirety please visit our website.
-Ritika S. Parris, MD, Alexander R. Carbo, MD
This article originally appeared in the February 2017 issue of The American Journal of Medicine.