Do general practice selection scores predict success at MRCGP? An exploratory study

Background: Selection into general practice training is undertaken using a competency based approach. The clear advantage of this approach over traditional methods has been demonstrated through evaluation of its validity and reliability. However, the relationship between selection and performance in the Royal College of General Practitioner examinations (MRCGP) has yet to be explored. The MRCGP comprises of an applied knowledge test (AKT), a clinical skills assessment (CSA) and workplace-based assessments (WPBA). Aim: To explore the predictive validity of general practice selection scores using the AKT and CSA elements of the MRCGP as a final outcome measure. Methods: This study carried out a retrospective analysis of 101 trainees from the Wales Deanery who were successfully selected on to general practice training in 2007. Selection data consisted of an overall selection score as well as scores from each individual stage of selection. Correlation was used to explore associations between selection scores and examination scores. Results: The score for overall performance at selection achieved statistically significant correlation with examination performance (r = 0.491 for the AKT and r = 0.526 for the CSA, P