The Fallacy of a Single Diagnosis

Area of Science:

• Cognitive Psychology
• Medical Decision Making
• Biostatistics

Background:

• Diagnostic reasoning in medicine involves navigating complex uncertainties.
• A potential bias towards a single, readily available diagnosis in uncertain clinical situations was investigated.

Purpose of the Study:

• To test for a bias favoring a single diagnosis in uncertain medical cases.
• To determine if the presence of an alternative diagnosis influences the perceived likelihood of a primary diagnosis, such as COVID-19.

Main Methods:

• Five distinct surveys were created, each presenting a hypothetical patient scenario in two randomized versions: one with an alternative diagnosis and one without.
• Participants, including community members (n=1104) and healthcare professionals (n=200), evaluated the probability of COVID-19 infection.
• Scenarios varied in complexity and risk, including cautious, risky, sophisticated, validation, and comparative cases.

Main Results:

• The presence of an alternative diagnosis (e.g., flu, mononucleosis) significantly reduced the estimated odds of COVID-19.
• Reductions in estimated odds ranged from 47% to 73% across different scenarios.
• This bias was observed in both community members and healthcare professionals, even in high-risk cases.

Conclusions:

• An easily available diagnosis can lead to premature diagnostic closure, hindering consideration of other serious conditions.
• This cognitive bias can cause individuals to underestimate the likelihood of diseases like COVID-19 when alternative diagnoses are present.
• The findings highlight a flaw in probabilistic reasoning affecting diagnostic judgments.