Blink or think: can further reflection improve initial diagnostic impressions?

Area of Science:

• Medical Education
• Clinical Decision Making
• Cognitive Psychology

Background:

• Clinicians often rely on intuition for diagnoses based on prior experience.
• While intuitive diagnoses are often correct, they are not always accurate.
• High-stakes board examinations serve as a model for studying clinical decision-making processes.

Purpose of the Study:

• To evaluate if additional reflection on initial diagnoses enhances diagnostic accuracy.
• To investigate the impact of deliberation on diagnostic performance in a simulated high-stakes medical examination setting.

Main Methods:

• Analysis of keystroke response data from 500 residents during the 2010 American Board of Internal Medicine Certification Examination.
• Focus on 80 diagnosis-focused questions with realistic clinical vignettes.
• Utilized a hierarchical generalized linear model to correlate response time, answer changes, and accuracy, controlling for cognitive skill and case complexity.

Main Results:

• Residents changed their answers on approximately 12% of diagnosis questions.
• Changing an incorrect answer to a correct one occurred nearly twice as often as changing a correct answer to an incorrect one.
• The relationship between time spent and diagnostic accuracy was complex, indicating nuanced cognitive processes.

Conclusions:

• Further reflection on initial diagnoses appears to benefit overall diagnostic accuracy.
• The positive impact of reflection is more pronounced in cases of higher complexity.
• These findings support the value of deliberate review in clinical diagnostic processes.