Related Experiment Videos
Performance of a dangerous answer subtest within a subspecialty certifying examination.
G D Webster1, S Goldfarb, J J Norcini
1American Board of Internal Medicine, University City Science Center, Philadelphia, Pennsylvania 19104.
Medical Education
|September 1, 1987
Summary
A study on nephrology board certification found that a subtest of "dangerous" answers did not identify uniquely unsafe doctors. This suggests current methods may not distinguish between certified physicians with potentially harmful practices and others.
Area of Science:
- Medical Education
- Nephrology
- Professional Certification
Background:
- Previous research suggested that some candidates passing specialty board certification exams may exhibit a high number of "dangerous" responses.
- These responses indicate a potential acceptance of actively harmful actions, raising concerns about physician competence.
Purpose of the Study:
- To confirm findings on identifying "dangerous" doctors through a retrospective analysis of a Nephrology subspecialty certification examination.
- To determine if a specific subtest of dangerous answers can isolate a unique population of certified but potentially unsafe physicians.
Main Methods:
- A retrospective analysis of a Subspecialty Board of Nephrology certification examination was conducted.
- Experts identified a subtest comprising 75 dangerous answers.
- Candidate performance on the dangerous answer subtest was compared to their overall examination performance.
Main Results:
- The dangerous answer subtest demonstrated moderate reliability and appropriately ranked criterion groups.
- The subtest correlated with the total examination score (0.71) and identified the certification status of 84% of candidates.
- When corrected for unreliability, the correlation between the subtest and the total test reached 1.0.
Conclusions:
- The study concluded that a dangerous answer subtest, in this specific examination and population, does not identify a unique group of certified but "dangerous" doctors.
- The findings suggest that dangerous answers may not represent a distinct category of physician behavior separate from overall examination performance.