Methods of assessment: selection of students for medical school

Research published in BMC Medicine shown that the UKCAT scores can be used to predict performance throughout medical school. Here, Sam Leinster discusses it in the context of challenges in medical school student selection.

The intended outcome of medical education is doctors who deliver effective and compassionate healthcare to the public. In order to achieve this, medical schools need to select students who can be educated to deliver such care. This selection remains one of the major challenges for medical schools.

Medicine remains a popular career choice and the number of applicants for medical school places greatly exceeds the number of places available.

Medicine remains a popular career choice and the number of applicants for medical school places greatly exceeds the number of places available. In theory, this should enable medical schools to select those students most suited to a career in medicine.

Conceptually, one should be able to define the attributes and competencies necessary for the successful practice of medicine and test the applicants for possession of the attributes and aptitude for the competencies. This approach has, after all, been used for many years by a wide variety of employers.

In practice, medicine is so diverse that it is difficult to define a core set of competencies in a form that is amenable to testing. Instead, prior educational attainment has commonly been used as the main criterion for selection.

In practice, medicine is so diverse that it is difficult to define a core set of competencies in a form that is amenable to testing.

Prior educational attainment has been shown to be highly correlated with subsequent performance in medical school and postgraduate examinations. However, a meta-analysis by the same authors showed that prior educational attainment accounted for only 65% of the variance in performance in first year examinations at medical school leaving 35% unexplained. This reduces the effectiveness of educational ability as a selection tool.

One effort to overcome this deficiency has been the introduction of the UK Clinical Aptitude Test which is currently used by 26 of the 34 UK medical schools as an aid to selection. This test consists of four subscales verbal reasoning, quantitative reasoning, abstract reasoning and decision making evaluated using a selected response format.

A recent well conducted study by Tiffin et al has shown a small but statistically significant additional predictive validity for examination performance in all years of the undergraduate medical course in comparison to prior educational attainment alone.

A recent well conducted study by Tiffin et al has shown a small but statistically significant additional predictive validity for examination performance in all years of the undergraduate medical course in comparison to prior educational attainment alone.

The authors suggest that this test could be used as a screening tool to exclude students who were predicted to do poorly in medical school examinations albeit at the expense of excluding a similar number of students who would have passed all of their examinations at first sitting.

Pragmatically, screening would reduce the number of applicants who were called for interview and so lessen the workload of medical school admission teams. However, it would do so at the expense of excluding students who were capable of graduating without problems.

In common with previous studies, this study has used performance in examinations as a surrogate for success in a medical career. This is a fundamental problem in any discussion of medical student selection. Clinical studies that use surrogate markers as outcomes measures are rightly criticized.

The same standards should apply to educational studies but there is no consensus on measuring effective and compassionate care. Knowledge and skills are obvious prerequisites so some correlation between examination results and effective practice are expected but while knowledge and skills are necessary they may not be sufficient.

New methods of selection are constantly being tried. Perhaps we need to put more effort into new methods of assessing outcomes so that the utility of the selection processes can be tested properly.

View the latest posts on the On Medicine homepage

Comments