Reliability of Mini-CEX Assessment of Medical Students in General Practice Clinical Attachments

Mini Clinical Evaluation eXercise (mini-CEX) involves observation of routine clinical encounters, initially developed to assess clinical competencies of postgraduate doctors. This study aimed to measure its inter-rater reliability in assessment of medical students in general practice settings.

General practitioners (GPs) supervising medical students were invited to complete online teaching on how to conduct a mini-CEX. This included three randomly presented scripted films of clinical scenarios representing different levels of student performance. Consenting participants completed an online mini-CEX. Mean marks were calculated for each case, Intraclass Correlation Coefficients (ICC) for overall clinical and four individual competencies, one-way analysis of variance to compare scores, and internal consistency measured by Cronbach's alpha.

RESULTS were collated for the first 100 completing GPs, majority aged 40-59 years (71%), male (59%), New Zealand European (58%). Forty-four percent were in rural practice, with 21 mean years in practice. Mean mini-CEX grades increased as standardized performance increased, indicating that GPs reliably agreed about ranking of student performance from poor to very good. The intraclass correlation coefficient (ICC) for overall clinical competency was 0.78 (95% confidence interval 0.48-0.99), indicating good reliability regarding their agreement with each other. A Cronbach's alpha calculated with the overall scores was 0.85, indicating good internal consistency.

Mini-CEXs in undergraduate general practice attachments provide a reliable measure of assessing performance. However, they may be less useful in identifying exceptional performance or weaknesses in key competencies. In addition, caution must be applied in relying upon mini-CEXs to supply a summative assessment.