Selection to medical school in Great Britain

British Medical Journal, April 3, 1999 by C Michael Steel, David Jackson, David W Sinclair, Stephen R Magee, D A Levison, D Parratt, J M Bland, Sheila M Gore, Chris McManus

Admissions procedure at St Andrews is driven by purely academic criteria

Editor--Readers of McManus's paper might be misled into believing that, in 1997, it would have been unwise for a prospective medical student from an ethnic minority to apply to St Andrews.[1] The reality is that for those who met, or were predicted to meet, our clearly published academic entry requirements, the chance of receiving an offer was 98.5% (compared with 97% for comparably qualified white applicants). It is doubtful whether these offer rates were equalled by any other medical school in the United Kingdom.

The difference between perception and reality can be explained as follows. First, McManus based his calculations on achieved A level or Highers grades whereas, for the former at least, selection to medical school has to be based largely on predicted grades. At St Andrews, and probably elsewhere, the relationship between predicted grades and likelihood of an offer is far from linear but shows a marked threshold effect, offer rate below the threshold being effectively zero. Second, for those applicants attempting to achieve the required entry grades at a second or third sitting, we take account of achieved grades at previous sittings and at GCSE, which were not included in McManus's dataset.

There are other deficiencies in McManus's approach. It may, however, be sufficient to note that fewer than 50% of candidates from ethnic minorities met our academic threshold (compared with over 70% of whites), that ethnic minority applicants who were resitting had correspondingly weaker academic credentials when GCSE and first sitting A level results were assessed, and that, overall, the simple odds ratio for receiving an offer from St Andrews (ethnic minority versus white applicants) corresponds exactly to the odds ratio for meeting our academic entry requirements.

Although the number of applicants from different subgroups taking up places in particular medical schools measures student choice as well as the schools' selection procedures, it may be worth recording that the proportion of St Andrews entrants in 1997 who were from ethnic minorities exactly matched the proportion who applied.

The difficulties inherent in digesting and analysing limited and incomplete data are properly emphasised by McManus but, given that this is by no means the first publication to address the issue of equity in selection for places at medical school, it is unfortunate that more time was not taken to analyse other evidence and draw more reliable conclusions. The important point to remember is that the evidence clearly confirms that the University of St Andrews' approach to applications is driven by purely academic criteria.

C Michael Steel Professor in medical science
David Jackson Admissions officer medical science
David W Sinclair Prodean medical science
Stephen R Magee Director of admissions
University of St Andrews, School of Biomedical
Sciences, St Andrews, Fife KY16 9TS

[1] McManus IC. Factors affecting likelihood of applicants being offered a place in medical schools in the United Kingdom in 1996 and 1997: retrospective study [with commentaries by A Esmail and M Demetriou]. BMJ 1998;317:1111-7. (24 October.)

Dundee University Medical School has some questions

Editor--McManus's study on factors affecting admission to medical schools' seems to contain discrepancies in the data that are misleading--namely, in table 2 and its comparison with figure 3, which deals with the possible deleterious effect of ethnic origin on admission to medical schools. In the table, Dundee University Medical School is "black spotted" and the results are described as "combined" from the 1996 and 1997 data from the Universities and Colleges Admissions Service.

Although it is not clear how such a combination of data has been achieved it would be reasonable to assume that the mean odds ratios from figure 3 have been used. If this assumption is correct we reason that an error may have occurred. The mean odds ratio for Dundee is 1.80, and this is awarded the "black spot" in table 2. However, Aberdeen (1.85), Belfast (4.0), Oxford (2.15), and Glasgow (2.0) all have greater mean odds ratios but are shown in table 2 as free from ethnic minority bias. McManus needs to address this paradox or explain the derivation of the data in table 2 in more detail. If such an error exists, the validity of the whole content of table 2 has to be questioned. Given that the table has been recommended to applicants and their advisers in schools and colleges as a guide to where it is best to apply, the consequences are serious. We are confident that our own admissions policy is fair. Our internal audit of 1997 applications using predicted examination data and personal achievement scores for all applicants shows an odds ratio of 1.12 for "whites" versus "ethnic minorities." It seems that, contrary to what is shown in McManus's table 2, applicants from ethnic minorities would be well advised to consider applying to Dundee.