Examinations (OSCEs) for Medical School Graduation: A Comparison of the borderline and Hofstee methods.

Introduction. Setting the cut-off point in objective structured clinical examinations (OSCEs) is a controversial aspect of assessment. In resource-limited settings, the Hofstee method requires additional tasks from other teachers outside the examination time. In contrast, the borderline group method is applied during the assessment, allowing for a more efficient use of time and resources. Objective. To compare the reliability of the borderline group and Hofstee methods applied in a graduation OSCE at an Argentine public university, providing local evidence to an internationally relevant debate. Population and methods. Cross-sectional study of 56 students in a 12-station OSCE. Two standardsetting methods were applied: borderline group (using observers during the exam) and Hofstee (electronic consultation with expert judges). Cut-off points, failure rates, and reliability (phi coefficient λ) were compared using generalizability theory. Results. The average score was 66.1 (SD 4.7). The cut-off point using the borderline group method was 54 (overall) with a reliability of 0.89 and no failures. The Hofstee method defined cut-off points of 60.7 (overall), with 3 and 1 students failing, respectively, and reliability of 0.68 and 0.82. Conclusions. Both methods show adequate reliability; however, they differ in their practical consequences, as the borderline group method was more lenient, generating a higher number of passing students.