Death: the surgeons' best-kept secret

Imagine buying a restaurant guide in order to go out for a good meal. This guide tells you how many tables there are, how many meals are served in a year, how fast the service is and how long you will have to wait for a table. But search as you may through its 200 close-typed pages, there is not a word about the food.

The NHS Performance Guide 1994-5, published yesterday tells us equally little about the real performance of doctors and hospitals. Even so, predictably, the professionals were indignant: the BMA passed a unanimous emergency motion regretting the publication of "misleading and unhelpful league tables.''

Some doctors do sometimes get away with murder. Like most professionals, they have evaded effective scrutiny of their work, they are reluctant to blow the whistle on killer colleagues, and most do not know how well they personally compare with the best (though some surgeons have been participating, voluntarily, in a Royal College of Surgeons project which tells them, privately, how they compare with others in their field).

All surgeons have to take part in "medical audit", a group session discussing their results, and some doctors take this very seriously. But a great many meet, say, once a fortnight, with a few colleagues behind closed doors, with no records kept and no manager allowed in.

If you ask a surgeon for statistics on his outcomes, only the most punctilious will have any idea. How many patients have to be readmitted? How many get infections in their wounds? How many feel better afterwards? Most doctors don't know. "I'd have to get out all my case notes to find that out, and I haven't time for that sort of thing," said one consultant, plainly not acquainted with computers.

When patients die in hospital after operations, the information is collected by the National Confidential Inquiry into Peri-Operative Deaths, inside the Royal College of Surgeons. But "confidential" is the watchword. All information is anonymised and shredded, used only for general surveys, so that no dangerous doctor has ever been identified through CPOD, even privately to the doctor himself.

The redoubtable Dr John Yates, former Department of Health adviser and relentless thorn in the side of the medical establishment, has compared outcomes of (anonymous) doctors in several (anonymous) districts. He found a national average of 2.5 per cent deaths for general surgeons, but some had rates three times higher. One small group had up to 6.5 per cent death rates, but because they all had the same results when they compared notes, they thought it was normal.

The story of the introduction of laparoscopic surgery is one example of what can happen in the secrecy of the operating theatre when no figures are published showing doctors' results. A disastrous series of avoidable surgical accidents could have been identified and stopped immediately.

Laparoscopic, or keyhole, surgery has been the first great surgical revolution since the introduction of anaesthetics. By using tiny remote-controlled scalpels and internal cameras, a surgeon can operate though a minuscule incision, watching everything on a video screen. It causes patients far less pain, allowing them to leave hospital within days, to the delight of managers. Manufacturers of the equipment pressed their wares on eager doctors, as managers and patients clamoured for it.

The surgical skills required are utterly new. While some pioneers trained and studied the new method with scrupulous care, in the mad dash hundreds of others thought they could learn by a couple of days of sitting next to Nelly. "It was see one, do one, teach one," says one surgeon, appalled at how some of his colleagues took up keyhole techniques with virtually no training.

The Royal College of Surgeons was warned that patients were being used as human experiments and many were being seriously injured. It is easy to misinterpret the images on the screen and nick an organ accidentally, without realising until the patient later develops agonising pain or acute blood poisoning. It was partly the fault of our bizarre animal protection laws that surgeons were free to practise on real live patients but not on live anaesthetised animals.

Throughout the late Eighties the Royal College, instead of taking action, continued to assert that its members were trained professionals who could decide for themselves what techniques to use. Only far too late did it introduce guidelines on laparoscopy training, by which time most surgeons had already learnt by trial and error.

The point about this story is that it couldn't have happened if medical outcomes had been published. Last December Scotland published crude clinical outcome tables, but as the press fell on these "death league tables" with glee, there was an aggrieved outcry from Scottish doctors - and they had a point. The very best surgeons may have the highest death rates, because they take on the most dangerous cases, including the botched jobs of other surgeons. The Scottish Office is not sure when, or if, it will repeat the exercise.

Now the Health Secretary has said that a pilot study is to collect clinical outcome data in England and Wales, with a view to possible publication in two years' time. Not surprisingly, Jim Johnson, chairman of the BMA's Consultants' Committee, was much taken aback by this news. He sits on the Department of Health's Clinical Outcomes Group, which has been attempting to devise sophisticated ways to measure doctors' success rates, taking account of the age and frailty of each patient and the severity of their condition. He is, of course, in theory, in favour of outcome measurement, so long as it's fair. "The Scottish figures were counter-productive and very, very unhelpful to patients," he says. "Our group is still a very long way down the line from finding a way to measure outcomes fairly. We have certainly not yet drawn up a model that works."

If the Health Secretary has jumped the gun on her own committee, perhaps it is because it would take for ever to perfect a system that would be acceptable to all doctors.

Yet if politicians had published outcome results - however crude - years ago, there would now be a far better educated public. Take one example:people are sentimental about small local hospitals. But would they feel the same if they knew that general surgeons in small units who dabble in a bit of everything get far worse results than big specialist hospitals? They would learn that it is often better to travel a long way to a major trauma centre than to arrive sooner at a small local casualty department with no specialists on duty. They would ask penetrating questions before being operated on by any surgeon who does fewer than 50-100 identical operations a year.

Professionals will always try to prevent the publication of league tables, but the only way to squeeze better information out of them is to publish relatively crude figures now. It will frighten doctors into hastening to provide better quality information in future and it might encourage them to weed out their own rotten apples who drag down the statistics of a hospital.