Confirmation bias has been known in healthcare for years, but clinicians still are learning how to avoid this pitfall. Emergency physicians are particularly susceptible, according to the study that most often is cited in explaining the problem.
The severity and volume of emergency care means many emergency physicians often must rely on heuristics, such as rule-out protocols, as a guide to diagnosing and treating patients, the author notes. The use of heuristics or protocols can be potentially misleading if the initial diagnostic impression is incorrect, notes author Jesse M. Pines, MD, MBA, an emergency physician at the Department of Emergency Medicine (ED) at the University of Pennsylvania in Philadelphia.1
"Confirmation bias occurs when people selectively focus upon evidence that supports their beliefs or what they want to believe to be true, while ignoring evidence that serves to disconfirm these ideas," Pines explains.
The author notes that ED physicians are particularly susceptible to cognitive errors of all kinds because they are required to integrate their knowledge base with new situations to create a diagnostic and management plan. Combining that situation with the fact that diagnosis errors are the foremost cause of malpractice claims makes the effects of confirmation bias even worse.
"Because of the rapidity with which EPs [emergency physicians] must work and the importance of an accurate diagnosis, it is important that EPs be cognizant of the possibility that diagnoses may be compromised by confirmation bias," Pines writes. "Put simply, this means that one may have an initial or a preconceived idea about something and interpret subsequent information or data so as to confirm that idea," which can mean incorrectly confirming a diagnosis.
Confirmation bias is closely related to "anchoring bias," in which an incorrect initial impression or diagnosis determines the course of action even contradictory information presents. Pines cites an example in which a patient who has been treated frequently for headaches actually has a subarachnoid hemorrhage on the next ED visit but the headache diagnosis prevails.
Noting that is no single solution to avoiding confirmation bias, Pines suggests that healthcare institutions should teach clinicians about the phenomenon and to be aware of its insidious nature.
"When the initial clinical impression is not corroborated by objective data, EPs must be open to revisiting the possibility of an inaccurate diagnosis and may have to start again at diagnostic time zero or, alternatively, defer to an appropriate inpatient or outpatient workup," the author concludes.
1. Pines JM. Profiles in patient safety: confirmation bias in emergency medicine. Academic Emergency Medicine 2006; 13:90-94.