In 1998 the General Medical Council (GMC)—the body responsible for registering doctors and for setting standards of professional behaviour in the UK—found three senior doctors from Bristol guilty of professional misconduct. The story, a complex one, centred on the work of two cardiac surgeons whose operative mortality for difficult operations for children with congenital heart disease (transposition of the great arteries and complete atrioventricular septal defects) was higher than expected.¹ There was public outcry. The surgeons were jostled outside the GMC's building by angry crowds shouting “murdering bastards” and carrying cardboard coffins.²

These events—now known simply as “Bristol”—are the subject of a public inquiry looking in depth at the circumstances of the unnecessary deaths (the inquiry can be followed on the Bristol web site http//:www.bristol-inquiry.org.uk). The focus of the GMC was the surgeons' professional conduct. The inquiry, probing into aspects of patient care beyond individual competence, has found that “institutional and organisational problems also played an important part in determining outcome, particularly in higher risk cases”.³ These issues are likely to have an impact on British medical practice far beyond the detail of complex paediatric cardiac surgery.

Cardiac surgery is only possible with organisational support. In addition to technical competence, successful heart surgery requires appropriate case selection, accurate preoperative diagnosis, detailed preparation for surgery, skilled anaesthesia, and good postoperative care. Coronary artery surgery is the most frequently performed cardiac operation. In the UK each year 25 000 operations are done; a median of 800 cases in each of the 35 NHS cardiac units.⁴ Mortality rates provide a means of comparing care among units and among surgeons; figures are available.⁴⁵ Nearly every patient survives. Over 75% of the 170 cardiac surgeons who do these operations report hospital survival rates of better than 97%, and about 90% of patients have an uncomplicated recovery and enjoy a good long term outcome. How should surgeons and units react when such high expectations are not fully met? Is the surgeon responsible or is there a problem elsewhere in the team? Or could there be organisational factors in the institution or in the complex chain of case selection and preparation for surgery? If so, what are they likely to be, how do we find out, and how do we improve?

Case selection is the first step. The purpose of cardiac surgery is to reverse or ameliorate problems that are life threatening or cause symptoms which diminish the quality of life.⁶ Benefit must be balanced against risk. All cardiac surgery is risky and some patients face more risk than others. But often it is those with the most critical cardiac disease, for whom perioperative risk is greatest, who have most to gain. The price of longer term benefit may have to be balanced by higher short term risk. A unit setting out to deliver cardiac surgery with the lowest possible perioperative mortality will disadvantage some of those with most to gain. Surgeons taking on higher risk cases will have more deaths. Hospital mortality figures may be wrongly interpreted without a sophisticated understanding of case mix.⁷

All active coronary surgeons operate on a group of patients with more complex conditions than in previous years when the expectations were set by randomised trial.⁸ In the UK, about a quarter of coronary artery surgery is on people admitted with myocardial infarction or for acute angina. About a third are over 70 years of age.⁵ Much case selection is in the hands of referring physicians and investigating cardiologists before the patient sees the surgeon. By the time the patient is in a cardiac centre with an angiogram performed the expectation is that a curative operation will follow. If the operation is not successful, to what extent are referring physicians who force a decision prepared to take responsibility? Sometimes the best the surgeon can do is to ensure that risk stratified data are kept to defend against subsequent criticism of high mortality.⁷

The operation itself is essentially mechanistic. If the coronaries are grafted faultlessly, the patient leaves the operating room with a better blood supply to heart muscle. On the perfect day heart surgery goes like clockwork. These are the days of miracle and wonder, when the heart beats in sinus rhythm, the monitoring lines are placed first time, the electronics work at the flick of a switch, and the physiological variables (circulating volume, left ventricular wall tension, pulmonary and systemic vascular resistance) are controlled to a nicety. The surgeon, working against the clock, sews small grafts to even smaller (1.5 mm) coronary arteries. There is little margin for error. Unless the patient leaves the operating theatre with a well vascularised myocardium, what are the prospects of withstanding protracted anaesthesia, major surgery, and the effects of cardiopulmonary bypass?

The surgeon's role is crucial. Technical competence is an absolute requirement of the operator; no one else in the cardiac unit can cut and sew the heart. To what extent is the surgeon answerable for the rest? Leadership and decisiveness are also necessary skills for the surgeon. The team expects the surgeon to recognise and diagnose the cause of problems as they present and to solve them,⁹ with the minimum of upset.

Surgeons cannot be answerable for all mistakes and poor case selection, errors in diagnosis, and poor theatre and postoperative management can be as disastrous as a surgeon's technical lapse Yet poor outcomes, whatever the cause, reflect on the surgeon. As team leaders, surgeons have a special responsibility. And yet unlike leaders in other fields they are unlikely to have had any training in leadership skills. The findings of the Bristol inquiry may be a watershed in our understanding of the complex set of interactions that are needed if the dangers inherent in high risk interventions are to be minimised. Many more questions will be raised than currently we have answers for. And some of the answers will not be in the technical language that surgeons are used to.¹⁰ By recognising both the technical and organisational aspects of surgery and looking at the performance of the whole, and of teams as well as individuals, we should be able to develop a methodology to enable units and surgeons whose mortality and morbidity figures are worse than expected to diagnose and correct problems in the series of events that make up a surgical procedure.¹⁰ But perhaps more importantly by understanding the organisational context of surgery we may be better able to manage the system as a whole and reduce the risk of surgery.