Global overall rating for assessing clinical competence: what does it really show?

Context  Single‐item global overall ratings are often used as a method of assessing learners’ clinical competence at the end of a clerkship. The purpose of this study was to identify which aspects of clinical competence are assessed through these ratings. Methods  At the end of a clinical clerkship in primary health units, 106 Year 4 students are routinely assessed by faculty staff of three disciplines (obstetrics and gynaecology, internal medicine, paediatrics), using a single global numeric rating (on a scale of 0–10). Faculty scores across disciplines for each learner are averaged to produce a global overall rating (GOR). In this study, the same students were assessed by the same faculty staff 2 weeks later using a newly developed, more detailed form composed of 13 domains, of which six related to technical skills and seven to humanistic skills, each scored on a scale of 0–10. Scores for each domain across disciplines were averaged as global itemised ratings (GIRs). Statistical analysis included Cronbach’s α coefficient and Pearson’s correlation coefficients. Statistical significance was set at P  ≤ 0.05. Results  The internal consistency of GIR items was high (α coefficient = 0.935). Global overall rating scores were higher than most technical domains of GIRs and lower than the humanistic domains of GIRs. The highest significant correlations were found between the GOR and the technical domains of the GIR. Conclusions  When faculty staff attribute a global single‐item overall rating to a student’s clinical competence, they tend to focus more on technical skills.