Using comprehensive feature lists to bias medical diagnosis

Area of Science:

• Medical Diagnosis
• Clinical Decision-Making
• Cognitive Psychology in Medicine

Background:

• Clinicians often underreport observed features during initial case assessments.
• Discrepancies exist between initial reports and later consensus on case features.
• Standard diagnostic processes may not fully capture all relevant clinical information.

Purpose of the Study:

• To evaluate the impact of comprehensive verbal descriptions on subsequent diagnostic accuracy.
• To investigate how pre-exposure to detailed features affects clinical judgment.
• To understand the cognitive mechanisms underlying diagnostic performance.

Main Methods:

• Generating complete and accurate verbal descriptions from medical photographs (dermatology, internal medicine).
• Presenting these descriptions to clinicians of varying expertise for initial diagnosis.
• Comparing the diagnostic accuracy of clinicians who received descriptions versus those who did not.

Main Results:

• Initial diagnosis based on comprehensive verbal descriptions impaired subsequent diagnostic accuracy when clinicians viewed the photographs.
• Experienced clinicians showed a greater decrease in diagnostic acumen compared to less experienced ones.
• The effect suggests a potential interference or bias introduced by the detailed verbal information.

Conclusions:

• Pre-exposure to a comprehensive list of features, even if accurate and present, can negatively impact diagnostic performance.
• Cognitive factors may lead to reduced diagnostic accuracy when detailed information precedes direct observation.
• Further research is needed to explore the reasons for this counterintuitive effect in clinical settings.