Acceptance into medical school and racial discrimination

Subjects, methods, and results

We obtained permission from the deans of all medical schools in the United Kingdom to analyse the data from the Universities and Colleges Admission Service on candidates who applied and were accepted for places in 1992. We classified candidates as belonging to an ethnic minority group when they identified themselves as being Chinese, Bangladeshi, Indian, Pakistani, Afro-Caribbean, or black. Candidates who did not provide information on their ethnic origin and those from overseas were excluded from the analysis. Applicants were stratified according to score at A level (30-26 or 25 or less, grade A scoring 10 points, grade B eight, grade C six, grade D four, and grade E two, with the maximum possible score being 30). Applicants with passes in the Scottish Certificate of Education were excluded because of problems in comparing the two examinations. To adjust for A level score a stratified analysis using a Mantel-Haenszel test1 was carried out and expressed as an odds ratio—that is, the odds of white candidates being accepted into medical school compared with the odds of candidates from ethnic minority groups being accepted.

The table shows the likelihood of being accepted into medical school according to ethnic group after controlling for A level score. The overall weighted Mantel-Haenszel odds ratio for acceptance to medical school was 1.15 (95% confidence interval 1.02 to 1.29) for those with A level scores of 30-26 and 2.02 (1.89 to 2.72) for those with scores of 25 or less.

Comment

Many factors could explain the differences between white applicants and those from ethnic minority groups in their acceptance rates to medical school. We were able to stratify A level scores into only two bands. Candidates within these bands might have had a wide range of results, and ethnic differences could account for the number of points scored.

In some medical schools the confidence intervals of the odds ratios were wide because of the small number of applicants from ethnic minority groups and the fact that we used data from only one year. Although these results should be interpreted with caution, we found significant differences in the odds ratio in several schools.

The data we analysed relate only to academic achievement, but other factors are considered equally important in selecting students for medical school. We cannot comment on how factors such as type of schooling, regional differences, and previous experience influence whether an applicant is offered a place. Factors other than ethnic group—for example, sex—could account for some of the differences that we observed.

A higher proportion of applicants to medical schools than in the population overall are from ethnic minority groups.2 Medical schools may therefore inadvertently be trying to restrict the overall numbers of students from ethnic minority groups to reflect population distributions in the United Kingdom. This may explain why white students with lower A level scores (25 or less) seem to have a greater chance of being accepted than their colleagues from ethnic minority groups with similar scores. There did not seem to be any selection bias in favour of white candidates among applicants with scores of 30-26.

The process of selecting students for undergraduate medical courses has long been a problem for those having to select. Admission policies are not standard or defined so the policies of individual medical schools vary greatly.3 Our data suggest that some medical schools could be accused of practising discriminatory admissions policies.

AE and DP were commissioned by the Medical Practitioners Union, London SE1 1UN, to carry out this study.