There are over 40,000 medical students in the UK. If Janet Yates’s figures are taken as representative of medical schools, it means that 5,120 of these students (12.8%) will struggle at some part of their training, with 1,600 (4%) eventually leaving. But it’s the gap between these two figures – the 3,520 individuals who struggle in medical school but manage to complete their degree – that I worry about. Doubtless some of them will go on to be good-enough doctors. But some won’t.
*
There have been a few studies of doctors who were disciplined for professional misconduct by the GMC (in the UK) or by their State Licensing Board (in the US).8 Convincing evidence has emerged from these studies that such misconduct was associated with early academic difficulties (in the UK) and prior incidences of unprofessional behaviour in medical school (in the US).9 So there is good evidence that difficulties in medical school (be they academic failures or examples of unprofessional behaviour) increase the risk of subsequent professional misconduct. However, only a small proportion of doctors – approximately 1% – in the UK and in the US – end up being disciplined by the relevant regulator. Although these findings are important, they therefore don’t tell us anything about those doctors who struggle through medical school but whose later clinical performance doesn’t warrant a referral to the regulator.10 Based on my experience of supporting junior doctors who were failing to thrive, I strongly suspect that some of these doctors who struggle at medical school continue to experience great difficulties at work – even if the difficulties are not serious enough to warrant disciplinary action.11
From my perspective as a psychologist working within the medical education system, I sometimes witnessed a marked reluctance to curtail a doctor’s training. In part, this is admirable; medical training is both long and demanding, and medical students and junior doctors should rightly expect to be adequately supported. But I suspect that other forces are in play as well. Sometimes I encountered a form of institutional denial, characterised by a distinct inability to consider whether the initial decision to offer a student a place at medical school could, perhaps, have been a big mistake.
Jennifer Cleland, an academic at Aberdeen Medical School, is one of the few people who have researched ‘failure to fail’ within the medical training system.12 In one study she highlighted how supervising clinicians struggle to give negative assessments to medical students, even when their performance warrants it. One respondent quoted in the study gives a taste of the problem:
I think part of the difficulty is you know we can all give patients bad news but medical students are potential colleagues and I think we’re very bad at communicating that sort of thing to our colleagues.
I find this extraordinary; what this doctor is saying is that she finds it easier to tell a patient that they are dying than to tell a medical student that they are failing. It’s another example of the way doctors unconsciously separate themselves from their patients – seeing them as belonging to a different group – in order to manage the inevitable anxieties of clinical work. And the result of this unconscious manoeuvre is that delivering catastrophic news to a patient becomes easier than talking to a medical student about their serious career difficulties. Other issues that supervising clinicians mentioned in Cleland’s study included lack of time, and lack of clear guidance on standards.13 Senior clinicians can feel unsure whether the clinical performance they observe when a student is with them constitutes a failing grade or not. This is hardly reassuring for future patients who might be treated by these medical students once they qualify.
In a later review spanning studies in the UK as well as North America, Cleland goes a step further, and questions the whole approach to remediating underperforming medical students:
The ethics of supporting students to progress to the next stage of training only to continue to perform poorly are, at best, questionable. It is also debatable whether scarce faculty resources should be used to support progression without improvement, which may take weak students further towards registration as potentially weak doctors when the evidence suggests that faculty members find it harder to fail senior students.
This is a step in the right direction – but it still doesn’t go far enough. What never gets said is that it is inevitable that some medical students will fail. Acknowledging this truth appears to be a step too far for the medical training system. Yet in order for all medical students to be perfectly suited to their chosen course, two conditions would have to hold true. First, the process of selection into medical school would have to be 100% accurate; and secondly, students would never change throughout the four, five or six years of their course. Clearly both assumptions are absurd, and even if the numbers are not huge, there are invariably going to be some individuals in the system – at medical school, and beyond – who shouldn’t be there.
This ‘failure to fail’ doesn’t occur only in medical training; it has been noted throughout the health and caring professions. Undoubtedly in both the UK and the US, a growing fear of litigation complicates the picture. Failing students can seek legal redress from the medical school, and gaining a reputation for removing students from the course doesn’t sit well with the financial need medical schools have to attract future applicants. But the unwillingness to tackle the issue of medical students who are unsuited to the practice of medicine doesn’t only pose risks to future patients – it also creates misery for the students themselves. Some of the students and doctors I’ve encountered should never have been allowed to remain in the system for so long. I wouldn’t want Kesia to suffer such a fate.
*
Wherever in the world one is applying to train as a doctor, significant weight will be given to one’s academic achievements. How that achievement is measured, the final weight that is attached to it in the overall selection decision, and what other factors are also taken into account, differ considerably between different countries and even between different universities within the same country. But everywhere in the world, prior academic achievement will form a central piece of the jigsaw that medical school selectors piece together when making their decisions.
In the US all prospective medical students have to sit MCAT (the Medical College Admission Test), a standardised computerised assessment.14 In medical schools across the country, admissions panels attach great weight to MCAT scores, and they have routinely been shown to predict results on the first set of national examinations (STEP1) that medical students take at the end of their second year. But all this means is that results on one computerised standardised assessment (MCAT) are significantly correlated with results on another computerised standardised assessment (STEP1) taken two years down the line. In contrast, the ability of MCAT scores to predict later clinical performance as a doctor seems, at best, to be extremely weak.15 (The examination was revamped in 2015, and now includes sections on psychology and sociology. It’s too early to know whether these new sections will better predict future clinical performance as a practising doctor.)
In the UK, A-level grades are used to assess academic potential for medical school. A-level grades predict both written and practical tests that are taken later during medical school. Chris McManus, a professor of medical education at University College London, uses the metaphor of the ‘academic backbone’ to describe the idea that, in medical education, current learning and achievement is critically dependent upon achievement at earlier stages. For McManus, this ‘backbone’ comprises ‘the accumulation of “medical capital”, that set of knowledge, theories, experience, understanding and skills that comprise successful medical practice’.16
According to this argument, it is essential that adequate weight is given to the academic achievement of applicants to medical school; if your backbone is weakened, you are not going to be able to stand up, let alone run. McManus also counters the frequently levelled criticism that choosing medical school applicants on the basis of school exam grades leads to the selection of students who are good at passing exams, but not necessaril
y good at the practice of medicine. He points out that ‘knowledge is generally preferable to ignorance, and clinical knowledge underpins clinical practice’. As a patient, it would be difficult to disagree with this; who wants to be treated by an ignorant doctor? And when I encounter doctors who have struggled academically from the beginning of their medical training, McManus’s warning that one needs to give adequate attention to prior academic achievement rings true.
But of course, academic achievement isn’t the whole story. Whilst as patients we want knowledgeable doctors rather than ignorant ones, we also need our doctors to have a whole host of other qualities such as empathy, an ability to handle pressure, to work well with their colleagues, and personal integrity. Academic achievement alone isn’t enough, when it comes to selecting doctors.
Basing selection decisions on academic achievements also introduces a social bias into the process because applicants from higher-income families tend to achieve better A-level grades. For example, a national survey of first year foundation doctors found that 31% attended a private school whilst only 7% of children in the UK were educated in the private sector – more than a fourfold over-representation.17 A similar social bias is seen in the US: the percentage of medical students from the highest 20% of parental incomes has never been lower that 48%, whilst the percentage of medical students from the lowest 20% of parental incomes has never exceeded 5.5%.18
Part of the problem, however, is that as medical schools have traditionally demanded the highest academic standards, there are little data available on whether it is feasible for a medical school to lower its academic entrance grades and still produce competent doctors. King’s College London, however, since 2001 has admitted students from low-achieving secondary schools, with lower grades than those normally required for medical school admission. Students admitted to this Extended Medical Degree Programme (EMDP) have an extra year to gain their final medical qualification.19
Perhaps not surprisingly, given the association between A-level grades and medical school exam performance, EMDP students are significantly more likely to fail finals. McManus therefore cautions against lowering entry grades without research into how much they should be reduced, and for which groups. A depressing conclusion – and also, perhaps, an incomplete one. Because even though EMDP students are more likely to fail finals than their peers who were admitted with higher A-level grades to the standard five-year programme, there is some suggestion that they are also over-represented in the most highly performing students. If you lower the entry grades for students who went to poorly performing schools you will increase the proportion of failures, and you will also identify brilliant students who otherwise would never have been given the chance to train as doctors.20 Tough choice. Isn’t there a better way of selecting the future medical workforce?
*
The holy grail of medical selection is a test that measures the potential of an applicant independent of the educational opportunities that the applicant has previously experienced. In this way, so the argument goes, able applicants can be identified whose secondary school grades were reduced because of the poor quality of education that they received. In the UK, two such tests have been devised – the UKCAT (United Kingdom Clinical Aptitude Test) and the BMAT (BioMedical Admissions Test). But do these tests provide an objective measure of ‘potential’, as the test suppliers claim? Do they tell us any more about the candidate than their A-level grades? And does their use lead to the selection of a more socially diverse set of medical students?
A five-year study of over six thousand entrants to medical school in the UK provides some good answers. What the study found was that UKCAT scores were significantly predictive of later performance at medical school, even when prior A-level results were taken into account. In other words, UKCAT results did add value to the selection process. The authors also argued that relying on UKCAT scores rather than A-level results could help to widen participation, as the former seems to be less influenced by the quality of secondary school education. So good news on both counts.21
The evidence about BMAT is much less impressive. The only part of the BMAT test that predicts later performance at medical school is the knowledge component. This part of the test is close in structure to a standard academic test such as A levels, and students who do well on it also do well in their A levels. So as yet there’s no evidence that the BMAT provides an education-free test of ‘potential’ as the test suppliers claim, or will do anything to widen participation. Is it a coincidence that this is the entrance test favoured by elite medical schools in the UK such as Oxford, Cambridge and Imperial College?22
Over the years I have spent supporting doctors, I’ve certainly encountered some exceptionally able doctors from working-class backgrounds; doctors whose parents were van drivers, or shop assistants, or who worked in a factory. But they are very much the exception rather than the rule. And as we have seen in earlier chapters, doctors who don’t come from professional backgrounds often struggle to feel that they fit in. On ward rounds consultants might criticise their accent or grammar and they often tell me that they feel less confident than their privately educated colleagues.
But it’s not just my own impression that medicine in the UK remains elitist; in 2012 the profession was named and shamed because of its poor record on social mobility. The Independent Review on Social Mobility and Child Poverty had this to say:23
Medicine lags behind other professions … It has a long way to go when it comes to making access fairer, diversifying its workforce and raising social mobility … The profession itself recognises that the skills which modern doctors require include far greater understanding of the social and economic background of the people they serve … It now needs to be matched by action. Overall, medicine has made far too little progress and shown far too little interest in the issue of fair access. It needs a step change in approach.
In terms of widening access, the profession isn’t keeping up. And this really matters, for a number of reasons. First there is the issue of social justice, and the desirability of having a fairer, more fluid society. According to this argument, the cohesiveness of society is at stake if individuals from particular groups feel that they have no chance of improving their social circumstances, even if they have the ability to do so. Then there is evidence that diversifying the medical workforce (be it in terms of social class or ethnicity) produces doctors who are more likely to work in the communities from which they were drawn. So for example a Scottish study found that GPs from less affluent backgrounds were more likely to work in GP practices serving the most deprived communities, compared to colleagues who came from more middle-class families. And it also seems that students who study in a more diverse medical school end up having more positive attitudes towards patients from minority groups than those who study in more ethnically homogeneous medical schools.24
*
In response to the damning criticism, the Medical Schools Council launched an initiative to make selection into the profession fairer and more transparent. Previously medical schools had been criticised for the somewhat opaque way in which selection panels operated.25 Beyond academic excellence, what else were they looking for in an applicant? In order to counter the charge of lack of transparency, in 2014 the Medical Schools Council compiled a list of the key skills and attributes needed to study medicine.26
The list kicks off with ‘motivation to study medicine and genuine interest in the medical profession’, which is a reasonable starting point. Next is ‘insight into your own strengths and weaknesses’, which also seems sensible. In total, seventeen personal qualities are listed. Perhaps I’m reading more into the order than is warranted, but I can’t help feeling an element of disquiet that ‘empathy’ and ‘honesty’ are at the bottom of the list. Studies that have asked patients what they look for in their doctor indicate markedly different priorities, with empathy coming near the top of the list.27
Whilst I might quibble with the order, overall the list of
attributes contains those that people hope to see in their doctors. But the key question is: how can they best be evaluated in a medical school applicant? Traditionally in the UK and in North America, medical schools have used a combination of written personal statements submitted in advance by the applicant, and interviews. But what’s the evidence that these methods work?
A major review of best practice in the selection of medical students commissioned by the GMC had this to say:28
Research evidence suggests that autobiographical submissions (like personal statements) are more susceptible to contamination and input from third parties than many other common selection methods, which disadvantages applicants from lower socioeconomic groups who are less likely to have the appropriate networks and resources to provide this.
The GMC report concluded that evidence supporting the utility of personal statements was ‘limited’. Nowadays, the Medical Schools Council produces a guide that lists the selection methods used by every institution in the UK. The 2017 Guide shows that the majority of UK medical schools no longer score personal statements – so clearly they have taken recent research findings on board. But what about being interviewed by a panel?29 Is this method any better at predicting who will turn out to be good doctors?
Also Human Page 25