Selection of medical students: an Australian perspective

Recent articles appearing in this journal indicate that the selection of students into medical courses remains a major issue for medical schools and a controversial one1–4. Admissions policies attempt to achieve a number of objectives, not all of which can easily be reconciled. These include: The selection of medical students with the cognitive skills and other qualities to enable them to cope with the demands of a lengthy and intellectually challenging period of training, both before and after completion of the medical course The selection of applicants with the potential to become ‘good doctors’ who possess the personal and professional qualities which enable them to sustain a career in medicine without experiencing undue stress leading to burnout The development of selection procedures which are equitable, consistent and transparent. This last requirement is frequently complicated by the need to exercise a degree of affirmative action in favour of disadvantaged minorities, or to meet some other perceived community need, for example to redress the shortage of doctors prepared to work in rural or remote areas. Within Australasia and in other parts of the world, major changes in the methods of selection into medical courses have occurred over the last 10–15 years. There has been widespread acceptance that entry to medical school should not be based solely on academic criteria. One reason for this change has been the introduction of the managed care model of health-care delivery5. This developed in the 1990s and the emphasis on seeing patients quickly created the dilemma of upholding the values of the profession while responding to the needs of institutions and businesses. This places new challenges on medical schools and teaching hospitals since this shift, brought about by market forces, will require doctors of the future to possess a range of new skills, and it has been proposed that the selection process should become more thoughtful and rich, not more simplistic5. Another development of the 1990s in the USA involved a requirement by the National Board of Medical Examiners that examinees seeking to be licenced to practise must pass assessments in professionalism, communication, and interpersonal skills. Other boards of registration have followed suit, leading to pressure on the medical schools to ensure that entering students bring some of these qualities with them6. That selection policies should consider humanistic, non-cognitive qualities as well as academic success has become widely accepted6–9. The need for diversity in the profession is also acknowledged8. Agreement on the most important qualities has been problematic6,10,11 and it is also difficult to devise robust, equitable and transparent means to assess such qualities7. However, inclusion of personal qualities has been found to add to the prediction of success in medical studies12. It is reasonable to question the extent to which students’ entry characteristics predict their performance in the course. It is generally accepted that pre-entry academic scores predict performance in the preclinical years8,11, but not in the clinical years, postgraduate training or as a doctor13. Marks in science subjects, which tend to be favoured by students planning to study medicine or dentistry13,14, show an association with results of written exams in medical school. However, the association decreases later in the course15. Whereas many studies have found that students with superior academic scores on entry are no more likely to perform well as medical students and physicians than those with lesser records13, a recent UK study found that A-levels had predictive validity for undergraduate and postgraduate medical careers16. However, these authors did not advocate the use of results in A-levels as the sole basis for selection16. Many students whose performance in pre-entry examinations places them in the top 1 or 2% of school-leavers fail to gain entry into medical courses. This arouses a great degree of resentment among the applicants themselves, their parents and members of the community at large, who share the view that such achievement should be rewarded. However, many factors contribute to a student's success in final school examinations. In the Australian context, top-ranked non-government schools consistently produce a high proportion of students in this category: excellent examination results are ‘guaranteed’ to a high proportion of able students due to the excellent resources offered by such schools. Many students achieve extremely high scores by a strategic choice of subjects, with a concentration on specialist mathematics and science subjects. However, evidence suggests that students whose pre-entry studies included humanities subjects performed better during their medicine course and as interns8,13. As stated above, there has been increasing recognition of the importance of qualities other than cognitive ability for success in the medicine course. A non-cognitive factor that has been found to associate with preclinical performance is conscientiousness11,17. It has been suggested that this personality trait and its facets (competence, order, dutifulness, achievement, self-discipline and deliberation) can be reliably assessed at the beginning of the course and that it affects academic performance during the course10. High scores on conscientiousness were found to correlate with A-levels grades and preclinical performance, and it was suggested that a measure of this characteristic could be used as part of the selection process17. However, selecting for an extreme in a personal quality may be undesirable. For example, a high level of conscientiousness can be associated with practitioners who exhibit a narrowness or perfectionism to an extent which is detrimental18. Good communication skills are needed by medical students as well as by practitioners, especially as modern curricula require students to take more responsibility for their own learning than the more conventional curricula of the past. There are various ways in which communication skills of applicants can be assessed as part of the selection process, including an interview, submission of a personal statement, and provision of written references. The most widely used of these instruments is the structured or semistructured interview. This forms part of the selection process in many of the undergraduate medical programmes (UMP) and graduate medical programmes (GMP) in Australasia1 and in the UK8. Both Newcastle12 and Monash Universities19 initially conducted interviews whose outcome did not ‘count’ for selection, so that some cohorts of students entering the medicine course had widely varying interview scores. This made it possible to establish the predictive validity and reliability of the structured interview, as well as its concurrent validity (correlated with a standard psychometric test)19. One of the complexities of the Australian scene arose with the introduction of GMPs, which had the effect of dramatically reducing the number of places available for school leavers wishing to undertake a medical course. This led to many applicants applying for and ultimately accepting places in interstate medical schools. The intense level of competition for UMP places (at least 10 applicants for each place at many medical schools) has necessitated a mechanism to preselect a manageable number of applicants for interview, given that logistic and financial constraints preclude interviewing the entire applicant pool (2500 annually at Monash University). Selection interviews for UMP medical schools are generally conducted before academic results (Equivalent National Tertiary Entrance Rank (ENTER) scores) are available. The solution adopted by the majority of UMP medical schools in Australia is to use performance in the Undergraduate Medicine and health sciences Admission Test (UMAT) to preselect an appropriate number of interviewees. Some universities use the UMAT score in conjunction with a predicted ENTER. Many applicants attend as many as four selection interviews in different locations around Australia, which is costly both for the applicants and the medical schools. Despite this, proposals to introduce a single selection interview for each applicant or to share interview data (as happens with the GMP medical schools) have met with strong resistance from medical schools. The selection procedures for many of the Australian UMP and GMP medical schools have recently been summarized1. Although there are exceptions, most schools employ an aptitude test (UMAT or Graduate Australian Medical School Admissions Test (GAMSAT)) to preselect for an interview; a measure of academic performance (ENTER or Grade Point Average) may be used to impose a threshold requirement or may be incorporated into the final ranking. The fact that UMAT (or GAMSAT) results have been used to preselect for interview presents a difficulty for medical schools seeking to establish the predictive validity of these instruments, as, inevitably, all applicants gaining entry to the course will have scores at the upper end of the distribution. The consequent restriction of range of scores makes it more difficult to investigate the predictive validity of these instruments. UMAT and GAMSAT both assess cognitive and problem-solving ability. UMAT is not curriculum- or content-based, whereas GAMSAT tests content knowledge (especially basic science knowledge) as well as writing ability. Several UMAT training courses have been marketed around Australia. The cost and quality of these courses vary considerably, and their existence raises equity issues as urban applicants from high socioeconomic backgrounds are more readily able to access them. It should be noted that many such applicants are also more likely to attend top-ranked non-government schools, and to have access to private tutors, personal computers and broadband internet access at home, all of which may contribute to a better academic score. Some UMAT training programmes have expanded their scope to offer a complete ‘selection training programme’, with targeted training for all aspects of selection including the selection interview. A survey of the published literature regarding selection of medical students reveals that there are various other avenues being explored in the search for other instruments to assess readiness for medical education. These include tests of emotional intelligence20, a teamwork exercise21 and a test of moral orientation22. The published literature indicates that no single selection instrument can adequately predict performance during preclinical, clinical, intern and postgraduate training, let alone identify those applicants who will ultimately pursue successful careers in medicine. The most consistent theme emerging is that there is merit in adopting a multifaceted approach to the task of selection. Furthermore, as has been argued by many commentators, there is a responsibility on admissions committees to seek evidence that the selection instruments they use continue to deliver appropriate outcomes.