Editorials

The surgeon as a risk factor

Determinants of outcome include technical skill, volume of work, and case mix

Surgery can be a risky business. Members of the public are now attuned to that fact and appreciate that their surgeon's performance is a key determinant of success. Outcome after surgery is of course relatively easy to assess; you survive the operation or you don't, the anastomosis holds or it doesn't, the hernia recurs or it doesn't, and so on. In an era of increased scrutiny it is perhaps no surprise that surgeons feel under pressure or that they account for a third of the referrals to the newly established National Clinical Assessment Authority in England.¹

What then are the determinants of surgical performance, and is a poorly performing surgeon easy to spot? Technical skill is vital, but it is by no means the only essential ingredient for success. Thorough training, compassion, sound judgment, good communication skills, honed clinical skills, and knowledge are all critically important. Surgeons do not work in isolation and success depends on effective collaboration and team working. This is not to submerge the surgeon in anonymitysurgical teams need enlightened leadership if they are to serve their patients well.

Increasing emphasis has been put on the relation between volume of surgical activity and outcome. One always imagined that the two were relatedafter all, even the greatest golfers have to play and practise regularly. There is now abundant evidence that hospitals with higher volumes of activity tend to have better outcomes and emerging evidence that surgeon's volume of work is also a determinant of outcome.^2-6 Areas where the relation between hospital volume and outcome is now clear include major surgery for certain cancers, cardiac surgery, liver transplantation, and major vascular surgery. However, words of caution are needed. The relation is not linear; some low volume units achieve good results whereas higher levels of activity do not necessarily guarantee good outcomes.⁷

Complexity of case mix must be allowed for; databases must be sufficiently large to allow robust comparisons; and trainees must be able to learn their craft. Comparison between hospitals is difficult, but it can be even more difficult to identify a poorly performing individual. Hospital outcomes can hide large differences between surgeons, and referral to a high volume hospital does not guarantee that surgery will be performed by a high volume surgeon. Even "busy" surgeons may take a long time to accrete enough performance data to allow valid comparison with their peers, particularly in low volume specialties such as neurosurgery.

The European carotid surgery trial provides a lesson in some of the pitfalls when assessing surgical performance.⁸ In this trial the overall risk of major stroke and death within 30 days was 7%. Seventy one of the 147 participating surgeons did not encounter strokes or deaths during the operation. However, for a 0% risk of stroke or death to be significantly lower than the overall risk of 7% a surgeon would have had to operate on at least 50 patients, whereas most of the 71 had operated on less than five. At the other end of the spectrum, surgeons with the highest risks (50% and 33%) had operated on only two and three patients, respectively. Only surgeon X seemed to have a significantly increased operative risk (11 strokes or deaths in 50 patients), but after correction for case mix even he ceased to be a statistically significant "outlier." This is not to say that alarm bells should not ring when a surgeon seems to be performing suboptimally, merely that care must be taken when interpreting what seem to be unusually high or low risks.

Our understanding of medical error and misadventure owes a great deal to the work of James Reason and the recognition that shortcomings of the system rather than the individual are often responsible.⁹ The system approach recognises that individuals are fallible, sees errors and mishaps as consequences not causes, and deals with failings in the organisation of the system concerned. There are encouraging signs that the United Kingdom is moving away from a knee jerk "name, blame, and shame" approach to medical misadventure and undue reliance on raw league tables.

While there may be occasional surgical "rotten apples," most errors in surgery (as in other disciplines) are committed by well trained, well motivated individuals. This is not to absolve individual surgeons from responsibility for their performance or lessen the importance of audit and peer review. In their haste to acquire new technology, surgeons have not always acquitted themselves well,¹⁰ although there are now encouraging signs that specialised techniquesfor example, total mesorectal excision for rectal cancercan be acquired widely with appropriate training programmes. ^{11 12} We need to be sure that members of the public do not have unrealistic expectations of their surgeon and their surgery and that they understand the risks beforehand. Equally we need to ensure that patients at higher risk are not denied surgery because no one is willing to operate on them. We need to work collectively to develop a truly open system that limits the incidence of error, recognises risk, allows surgeons (and all healthcare professionals) to learn from mistakes, and replaces blame and retribution with an opportunity for learning and training.

David Carter, professor emeritus, University of Edinburgh. 

PPP Foundation, London W1G 0PQ

Footnotes

Competing interests: None declared.