Reliability of clinical oral examinations re-examined.

Many medical schools still use oral examinations for the evaluation of clinical competence of students in their clerkship, although it has been proven that orals have poor reliability. This study investigates the feasibility and reliability of multiple oral examinations. Students in the last week of their Internal Medicine clerkship in an outpatient clinic were given several patient-based oral examinations. The student's performance was rated on a list of items reflecting clinical competence. A global judgement of the student's performance was also given. The results indicate that it is possible to increase the number of orals and the number of examiners in the day-to-day practice of an outpatient clinic moderately. The reliability when using a number of orals is better than the reliability of the common single oral examination. The reliability using global judgements appeared to be better than the reliability of averaged item scores.