| Sumario: | Purpose: In sequential Objective Structured Clinical Examination (OSCE) all students initially take a short screening OSCE. Examinees who pass are excused from further testing, but an additional OSCE will be administered for the remaining examinees. Previous investigations on sequential OSCE were based on the Classical Test Theory. We aimed at designing and evaluating screening OSCEs based on Item Response Theory. Methods: We carried out a retrospective observational study. In each of a 10-station OSCE, the students’ performance was graded on a Likert-type scale. Since data was polytomous, difficulty and discrimination parameters, and students’ ability were calculated by graded response model. To design several screening OSCEs, we identified the five most difficult stations and the five most discriminative ones. Five, four, or three stations were selected for each test. Normal and stringent cut-scores were defined for each test. We compared the results of each 12 screening OSCE to the main OSCE and calculated the positive and negative predictive values (PPV and NPV), as well as the exam cost. Results: 253 students (95.1%) passed the main OSCE. 72.6% to 94.4% of examinees passed the screening tests. The PPV values ranged from 0.98 to 1.00, and the NPV values ranged from 0.18 to 0.59. Two tests effectively predicted the results of the main exam accurately and could result in 34-40% financial saving. Conclusion: Sequential OSCE, if stations with highest values of IRT-based discrimination and stringent cut-scores are applied to the screening test, can be an efficient and convenient way to conduct an OSCE.
|
|---|