Suitability of selection criteria as a measure of outcome of medical graduates: University of Colombo

Abstract

A prime obstacle faced by a medical educator is selecting the right student to be trained as a doctor, and the general consensus is that this is also the most difficult task. This study was designed to evaluate the effects of selected outcome measures on outcome performance of medical undergraduates of the University of Colombo. A retrospective cohort study was conducted using the performance (marks) of students of 4 batches GCE (A/L) 1993 through to 1996). GCE (A/L) aggregate marks, attempt of entry, district of entry, English language proficiency and sex were tested as predictors of success. Results of main assessments were considered as measures of success. Relationship between outcome measures and outcome predictors were assessed using the multiple logistic regression model. Data of 699 students were analyzed and 82% of students entered from the Colombo district. A higher percentage of first attempters (at GCE A/L) performed well and obtained classes. Entering medical school from first two GCE A/L examination attempts was a significant positive predictor of passing any examination (odds ratio 3.2 to 7.5) or obtaining honors (odds ratio 2.8 to 16.0). Attempt of entry predicted 5.4% of the outcome (pass or fail) in university performance. Correlation between the GCE A / L aggregate mark and the student's position in order of merit for the internship appointments was -0.37 (p<0.001). A combination of factors should be used in the selection process of students to embark on the undergraduate process as any single factor is a poor predictor of outcome of performance. We believe that the number of attempts allowed to sit for GCE A/L in order to gain entry to a medical school as well as other degree courses should be confined to two attempts.