Competence, professional self regulation, and the public interestBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7146.1740 (Published 06 June 1998) Cite this as: BMJ 1998;316:1740
- Rudolf Klein, professor of social policy ()
Editorial by Treasure News by Dyer
The case of the three doctors in Bristol represents a landmark in the history of the self regulation of the medical profession in the United Kingdom in terms of its length, its salience in the eyes of the public, and the issues it has raised. It has stretched over eight months and involved more than 60 days of hearings before the General Medical Council—probably the most extended and expensive case in the history of the GMC. It is the stuff of which headlines are made; it is highly charged emotionally since it concerns the deaths of children after heart operations. And it has provided a test case for the GMC's policy of seeking to ensure that all members of the profession accept their collective responsibility for maintaining standards and practising within the limits of their competence.1
Until the GMC has determined its verdict, in the light of their findings of the facts, it would be improper to comment on the actions of the individuals being investigated. However, the case raises some wider questions, both for the medical profession and for the NHS. This paper explores some of these questions from the perspective of a lay observer, drawing on an analysis of the transcripts of the proceedings.
There seems to be some confusion about how doctors should interpret their responsibility for protecting patients from harm from other doctors
Doctors seem to need training in communicating with each other
There may be a need for more explicit and stringent training requirements before surgeons are permitted to operate independently
There may be a need for more explicit requirements for retraining when results are poor
A frustrating procedure
The proceedings have been long and drawn out because they have involved the detailed examination of the circumstances of every death in the series of operations in question and every relevant conversation and meeting occurring during the five years between 1990 and 1995 at the Bristol Royal Infirmary. Expert witnesses were called to review the operations: much of the argument revolved around whether Bristol's excess number of deaths—compared to the national average for the specific procedures in question—reflected a particularly difficult case mix. Prolonged cross examinations were used to establish who had voiced concerns about the outcomes to whom, what data had been produced when, and what action had been taken as a result.
It would be difficult to imagine a more painstaking procedure. Yet it is almost certain to leave a sense of frustration among the public; indeed some of the parents involved in the case spoke of their frustration during the hearings.2 The reason for the frustration is the constraint imposed on the GMC by the legislation under which it operates and by the rules of evidence used in criminal justice. The GMC was not conducting an inquiry into what happened at the hospital in Bristol but considering specific charges against specific doctors.
In the case of the first surgeon, the charges concern a series of arterial “switch” operations; in the case of the second, the charges revolve around a series of operations for the correction of atrioventricular septal defects. The case of the third doctor, the chief executive of the healthcare trust, revolves around his responsibilities for taking action as a result of concerns voiced by some members of the medical staff. Common to all three doctors was the charge that, in view of the outcomes, the operations should have been stopped sooner than the surgeons themselves decided to stop them.
The case was not a review of the overall performance of paediatric cardiac surgery performed in the trust; the hearings were held to determine whether the outcomes in the specific operations covered by the charges represented a wider failure of systems. Although the proceedings did illuminate some of the background, they did so only fitfully and almost incidentally. The defence pointed out that some of the key players at the trust were not called to give evidence; one reason for this seems to have been that three consultant members of clinical staff had been sent warning letters by the GMC, which made them reluctant to give evidence (presumably for fear of incriminating themselves).
The purpose of the hearing was not to establish a complete balance sheet of the quality of care provided at the hospital in Bristol but, more narrowly, to establish whether specific charges had or had not been proved. The hearings were also not an inquiry into the causes of the deaths, though this had been the focus of much public concern. Again, evidence about what had gone wrong emerged only as a byproduct of the hearing; and the hearing leaves uncertainty about just how much has been revealed. Moreover, the medical members of the panel conducting the hearing—who are perhaps better equipped to follow up some of the wider issues than the counsel conducting the cross examinations—may have been inhibited from doing so for fear of appearing to be biased, and thus giving cause for appeal. (Early in the proceedings, defence counsel sought to disqualify the president of the GMC, Sir Donald Irvine, on precisely such grounds.)
However meticulous and however stringent GMC disciplinary hearings are—and the Bristol case scores highly on both counts—they cannot wholly allay public anxieties. This is not their function and they are not designed to do so, even though they make visible the profession's collective determination to maintain standards. This is a highly important symbolic function.
In high profile cases of alleged medical incompetence, particularly when there is anxiety about what are perceived to be unnecessary deaths, it would seem more sensible to appoint independent review panels to conduct a comprehensive inquiry. The review of cervical screening services at the Kent and Canterbury Hospitals NHS Trust provides one model3; it may also be that the NHS Commission for Health Improvement will develop another. In the Bristol case, the government decided against convening such an inquiry. Had it done so, the GMC might have had a simpler task, and any consequent disciplinary proceedings might have been less protracted, less stressful, and less expensive for all concerned; it would be difficult to exaggerate the strains imposed by the case both on the defendants and those hearing the charges.
If the case is likely to leave behind it a sense of public frustration, it may bequeath a legacy of concern to the medical profession. For even though the case centres on the charges against the three doctors, one theme running through the evidence is the difficulty of knowing where to draw the line between individual and collective responsibility. The two surgeons, clearly, were responsible for deciding whether to operate and for their competence in carrying out what one witness described as particularly unforgiving procedures. But the decision whether or not to refer patients to them rested on others. Similarly, the responsibility for providing accurate diagnostic information preoperatively and for providing postoperative care rested on others. To the extent that outcomes are the product of a collective effort, which was a point stressed by several expert witnesses, it may seem arbitrary to single out individuals for censure. If there are institutional shortcomings, as there seem to have been at the hospital in Bristol, who should take the blame?
One answer, of course, may be that everyone should share the blame, apart from those whistleblowers who, to their credit, raised concerns beginning in the early 1990s. They were, for the most part, outsiders: recently appointed consultants. And in what seems to have been a rather inbred culture at the Bristol Royal Infirmary—where all the main participants had been together for a long time—their actions seem to have been resented and discounted. Moreover, it probably did not help that one of the surgeons whose performance was being questioned was also chairman of the hospital medical committee and medical director for much of the relevant period, a position more likely to command prudent deference rather than encourage open criticism.
The extent to which the warnings were discounted, and to whom they were communicated, was much disputed during the proceedings. So, too, was the extent to which different participants interpreted their own duties in following up concerns. Some conceded that with the benefit of hindsight they should have been more persistent and forceful. On all these points, of who knew what and when did they know it, there was much conflicting evidence. Two general conclusions would, however, seem to follow. Firstly, there seems to be some confusion about how doctors should interpret their responsibility, as set out by the GMC, for protecting patients “when you believe that a colleague's conduct, performance or health is a threat to them.”4 How active should doctors be in following up concerns? To whom should they address their concerns? More explicit guidelines may be needed.5
Secondly, the evidence suggests that there was a pattern of misunderstanding and miscommunication, with a reluctance by staff to engage in confrontations. Even the chief whistleblower was described by the GMC's counsel as maladroit in the way he voiced his concerns. Conversations in corridors and at the end of meetings about other matters delayed far too long the day when the data were systematically examined by all relevant clinicians. This suggests that doctors need training in communicating not only with patients but also with each other.
The larger picture
One important issue raised by the case is whether the Bristol Royal Infirmary should have been carrying out the operations in the first place. It was generally conceded that conditions at the hospital were not conducive to successful outcomes; there is not a dedicated operating team. The two consultant cardiac surgeons primarily operated on adults; operations on children were only a small part of their workload. The number of operations included in the charges was small, for example 15 procedures to correct atrioventricular defects were performed between 1990 and 1994 (compared to the 30 carried out annually by one of the expert witnesses). Indeed, much of the defence case rested on the argument that everyone was conscious that Bristol had not achieved the gold standard of outcomes achieved by highly specialised, high volume units, such as at Birmingham Children's Hospital, but was striving to improve performance by concentrating all facilities on a single site and appointing a specialised paediatric surgeon. These aims were ultimately achieved and led to a dramatic improvement in outcomes.
This, however, only prompts the question of whether the surgeons at Bristol should have started performing these unforgiving operations in the first place. Given the general presumption that quality is related to quantity—that developing the necessary knack, as a surgeon from Birmingham put it, requires experience—was it wise to go down this road? Institutional imperialism (which affects hospitals as much as university departments) no doubt prompted the Bristol Healthcare Trust to stake its claim in this field. But if the self interest of the staff at individual hospitals drives them to embark on what may be initially risky endeavours, then there may be a public interest in restraining them. In this respect, the Bristol case appears to strengthen the argument for concentrating expertise in selected hospitals.
But even assuming that the surgeons in Bristol were right to start performing these operations, a further issue arises, again with more general implications. In Bristol, the high mortality experienced when the “switch” operations were started was attributed to the learning curve, a somewhat elusively elastic notion. Such “learning curve deaths” may be inevitable when new procedures are being tried out. Are they inevitable, however, when a procedure is already being carried out successfully in other places? Or could they be prevented by making more explicit and stringent training requirements before late starter surgeons (for example, those who embark on operations already carried out successfully elsewhere) are permitted to operate independently? In the case of minimally invasive surgery such requirements have been introduced; the Bristol case indicates a need for expanding this type of requirement (and perhaps also for having more explicit requirements for retraining if results are poor).
The question of how to assess performance once operations have started remains. In part this assessment depends on rigorous audit: the GMC proceedings do not provide a clear picture of whether audits were carried out. Although the surgeons in Bristol clearly engaged in much self analysis, it is not apparent just how methodologically rigorous the review of their results was over time. But audit depends on having some kind of benchmarks. And, disquietingly, the evidence given in this case underlined the absence of such benchmarks. The UK Cardiac Surgery Register does not stratify for risk, has no formal validation of data, and does not indicate the range of results at different units or of individual surgeons. It is therefore difficult to know when relatively poor performance becomes unacceptable performance; this was a problem for all the witnesses in the proceedings. Clearly, there is a need to develop adequate benchmarks; this will be an urgent task for the proposed UK National Institute for Clinical Excellence.
The case raises other issues, too, which range well beyond the particular circumstances of the Bristol Royal Infirmary or paediatric cardiac surgery. The role of non-executive members of healthcare trusts has become an issue; their absence from the Bristol story is remarkable, especially given that stories in the satirical magazine Private Eye put the issue on the public agenda (a fact which should surely have alerted everyone that there was enough dissent among staff to persuade someone to leak information to the press). It also raises questions about the relation between non-medical chief executives and the audit machinery especially once the proposal to make chief executives statutorily accountable for quality is implemented (will this mean providing information about the performance of individual consultants?). The case also prompts a look into the role of the Royal Colleges in accrediting training posts: should this not provide an opportunity to spot more general problems? (In the case of Bristol, approval for a senior registrar in paediatric cardiology was withheld.)
If the Bristol case prompts many questions, it has also provided one clear, emphatic, and welcome answer. If there were any doubts about the GMC's commitment to its contract with the public, about its determination to demonstrate the profession's collective acceptance of responsibility for maintaining competence in practice,6 they have been dispelled. And that should send a powerful message both to the profession itself and to the public.