Clinical skills assessment

Background. From September 1996, all GP registrars completing vocational tr aining in the United Kingdom must demonstrate competence by means of a four -part assessment procedure. Aim. To look at the accuracy of one of the components of vocational trainin g: the trainer's report. Method. Seventy-five registrars completing their general practice training at the end of July 1997 were invited to fake part in a practical skills wor kshop. Eight stations were designed to test practical skills and diagnostic interpretations that were included in the trainer's report, and a clinical vignette accompanied each task. The marking schedule used was developed fr om the minimum standards required in the trainer's report. Twenty-nine regi strars (38%) took part in the workshop. Results. Only one registrar passed all eight stations. The maximum number o f stations failed by any one individual was five and this doctor was the on ly one of the sample to ultimately fail summative assessment. The majority of registrars failed by being unable to interpret clinical findings. Twenty -five registrars (86%) responded to the follow-up questionnaire. Of these, only six felt that the stations were unrealistic. All but two registrars ha d spent at least six months in their hospital training doing obstetrics and gynaecology but, in spite of this, only 31% of registrars were above minim um competence for vaginal and speculum examination. Conclusion. With one exception, registrars passed all aspects of the traine r's report. Discrepancy was found between the trainer's report and the doct or's ability to carry out clinical procedures. There is an assumption that many of these clinical skills are being taught and assessed at undergraduat e level and during the hospital component, but this cannot be taken for gra nted. Doubt must also be cast on whether the trainers are using the trainer 's report appropriately, and whether this is a Valid and reliable tool to i dentify skills deficient in registrars for summative assessment.