By: Norman Bauman for Veins1
St. George's, London, is the first hospital in the U.K. to publish death rates for its surgical procedures. Click "Mortality rates" on its web site, and you can find "league tables" showing their death rates by department over the last five years. For example, the vascular surgery department, which repairs the carotid, aortic and iliac artery, has a jagged curve that bounces up and down like a stock chart, with peaks of 7 percent and troughs of 2 percent, with an average of 5 percent.
| Take Action |
| The simple rule for picking a surgeon of a certain type, like a vascular surgeon, is to go to a hospital that does a lot of vascular surgery, since the institution is more important than the individual surgeon. Once there, get a surgeon who specializes in your procedure, or a surgeon who has done a lot of procedures. There are many exceptions, but this is a good overall strategy. |
So if you need surgery, you might be tempted to go shopping for the hospital with the lowest death rate.
But as St. George's web site explains, that doesn't quite work. Even the best hospitals and doctors have worse outcomes for patients who are sicker (or just older). Some patients can't be saved. Conversely, any surgeon can improve his outcomes merely by refusing difficult cases. Those jagged curves are hard to interpret.
Nonetheless, doctors and hospitals keep their own statistics, first to monitor themselves, and second to figure out how to do better surgery.
More is better
The simple rule for picking a vascular surgeon is to go to a hospital that does a lot of vascular surgery, since the institution is more important than the individual surgeon. Once there, get a surgeon who specializes in your procedure, or a surgeon who has done a lot of procedures. But that rule is an oversimplification, and there are a lot of exceptions.
Visiting the United States last year, Peter J. Holt from the vascular surgery department at St. George's, explained to the vascular surgeons at the VEITH ™ Symposium in New York how he interpreted vascular surgery death statistics.
Individual studies give inconclusive or contradictory results, because they have too few patients. Because of this doctors do meta-analyses, which combine individual studies. Meta-analyses give an overview, and find patterns you can't see in individual studies. But they have their own limitations.
Dr. Holt did a meta-analysis of every good study he could find of the death rates of surgery for abdominal aortic aneurysms. In an aneurysm, the aorta balloons up like a bicycle tube. When it gets to the size of a doorknob, it's very likely to burst. It must be repaired, because if it bursts, the death rate is very high.
Dr. Holt found 30 good articles on elective surgery for abdominal aortic aneurysms, which included over 400,000 surgical cases. The overall mortality rate was 9.5 percent. He lined up the studies according to the number of surgeries at each hospital.
The break point was at 43 surgeries per year. Hospitals performing 43 or more surgeries per year had a 34 percent lower death rate than hospitals performing less than 43 surgeries per year. "There's a 34 percent volume-related reduction in mortality for elective aneurysm surgery," said Dr. Holt. This is applicable across all health care systems, in the U.K. and the U.S. as well. "The effect is certainly substantial."
Dr. Holt didn't give advice to patients, but he did give advice to doctors and health care policy-makers. "Centralization of aneurysm services is required," he said, "and we suggest a minimum threshold of 43 elective aneurysms per annum."
Why is more better?
"It's certainly plausible that high-volume hospitals have better outcomes," said Dr. Holt. "These larger units tend to have specialist vascular surgeons, each of which has higher volume." Furthermore, the whole system is set up more efficiently. There are specialists in intensive care, kidney disease, cardiology, and respiratory disease – all the doctors who treat the patient if he runs into trouble after the surgery is over. The anesthetists are more familiar with and specialized in vascular surgery. The intensive care departments are larger.
"They may be more effective working in larger multi-disciplinary teams," said Dr. Holt. "It makes sense always to tie them together, and certainly in the U.K. the small hospitals don't have all those other services."
At the VEITH™ Symposium, doctors asked him what was going to become of the vascular surgeons in small hospitals. "If you can't provide the minimum volume of surgery, it's quite clear that the mortality rate is worse," said Dr. Holt. So the surgery should only be done in large, high-volume centers, with a minimum of 43 cases per year.
The surgeon or the hospital?
Which is more important: the surgeon or the hospital? One doctor in the audience described a study he had done in New York state 10 years ago, which found that they had higher mortality not only with the low-volume hospitals but also with the low-volume surgeons. "We found that surgeons which had fewer than five aneurysm repairs per year had double the mortality."
Dr. Holt's study didn't look at that, but "the best results are generally specialist surgeons working in high-volume centers," he said.
There are individual variations among surgeons, said Peter R.F. Bell, Leicester, U.K., who moderated the panel. But the variations among hospitals are more important.
Doctors vary, said Dr. Bell. "Not everybody's good at everything equally well, are they?" In the good hospitals, said Dr. Bell, "you have better anesthetists, you have better teamworking, you have better ICUs, you have better controls."
But raw mortality numbers can be misleading because some hospitals and doctors choose easier patients. "It's no good saying that this is the cardiac surgeon with the best results," said Dr. Bell, "because if he starts doing the bad patients, he'll get bad results."
Meta-analyses find general trends, but there may be exceptions. The U.S. Department of Veterans Affairs also does extensive studies to monitor and improve their outcomes. In 1996, they published a study of 3419 elective aneurysm repairs over three years in 116 VA centers. Centers that did 32 repairs or more had lower mortality than centers that did 31 repairs or less. But "many lower-volume centers had excellent results," they said. They recommended that doctors refer patients to higher-volume centers, "or to lower-volume centers that have excellent results."
