Home Clinical A surprising look at the relationship between hospital volume and quality

A surprising look at the relationship between hospital volume and quality

December 2003

Published in the December 2003 issue of Today’s Hospitalist

In recent years, a growing body of literature has suggested that patients who undergo surgery at high-volume centers fare better than those treated in low-volume hospitals. This volume-outcome relationship, which is based on the age-old notion that practice makes perfect, has been vigorously embraced by large purchasers of health care. Through initiatives like those promoted by The Leapfrog Group, employer groups have tried to steer their patients to high-volume centers on this basis.

Two studies in the Oct. 21 Annals of Internal Medicine, however, suggest that the volume-quality relationship is anything but a slam dunk. Instead, those articles show that many other factors “from physicians’ practice patterns to surgeons’ patient bases, to name a few “are also driving hospitals’ outcomes.

In one study, investigators looked at the possible effects of what they call the “clustering phenomenon” “the tendency of certain types of patients to seek or obtain care from certain types of physicians “and how that might skew results in studies looking at the link between volume and outcomes.

In another article, Jeffrey A. Meyerhardt, MD, of Boston’s Dana-Farber Cancer Institute explored the relationship between volume and quality in colon cancer surgery. He and his co-authors discovered that patients who underwent the surgery at high-volume hospitals did indeed have slightly higher long-term survival rates than patients who underwent procedures at low-volume centers.

When researchers narrowed the criteria to look only for deaths related to colon cancer, however, they found that patients who went to high-volume centers experienced almost identical death rates and tumor-recurrence rates. In other words, they found that the higher death rate of patients who had gone to low-volume hospitals was due to conditions unrelated to colon cancer or the surgery.

Dr. Meyerhardt and his colleagues decided to examine the link between volume and quality in colon cancer surgery precisely because the procedure is basic, and it seemed unlikely that volume alone could be to blame for variations in outcomes. “We weren’t convinced that for something like colon cancer surgery, which is a fairly common operation, the hospital effect was truly the cause,” he explains.

“The survival difference was somewhat interesting and surprising,” he adds. “The lack of recurrences was not so surprising. Putting those two results together, however, shows that the relationship between volume and outcomes is something we need to look at further.”

In a conversation with Today’s Hospitalist, Dr. Meyerhardt talks about why it is important in the context of care quality and patient outcomes to look at multiple factors above and beyond per-surgeon and facility patient numbers.

What makes you think that volume measures may not be the best indicator of quality?

There is so much more to the issue than simply how many operations you do in a particular center. Our study is further evidence of that. It shows that how many colon cancer operations a center performs does not affect colon cancer outcomes.

If volume isn’t a surrogate for quality, what role do you think it plays in patient outcomes?

That’s the question we’re trying to answer. Volume is obviously a surrogate for something that seems to be affecting how long patients live. Is that related to the fact that smaller hospitals treat potentially sicker patients, people with more co-morbidities or factors we weren’t able to adjust for in our study? Or do these patients have less access to health care? We know from multiple studies that access to health care plays a very important role in patients’ overall outcomes.

There may be another component of low-volume hospitals, a socioeconomic factor that helps determine who receives care at these facilities. Perhaps those patients can’t get to the doctor as frequently because they’re working. Or maybe doctors at those centers do other procedures besides colon cancer surgeries. They may have different practice patterns than doctors at the larger centers.

Those are potential components that should give us pause when we look at some of the volume studies, at least in terms of long-term outcomes.

It’s also important to emphasize that in our study, we didn’t look at 30-day mortality from surgery or complications from surgery in the immediate future. The patients we studied were at least six weeks out from surgery, so we couldn’t look at those factors.

Do you equate low-volume centers with community hospitals?

In general, yes, because high-volume hospitals are generally located in the cities. That said, there are certainly community hospitals that have an expertise in certain diseases or procedures. We know that there are some community hospitals that do more colon cancer surgeries than the academic center down the road, for example, so I wouldn’t say that a community hospital is necessarily a low-volume center.

When people discuss the relationship between volume and outcomes, it’s often in the context of complex, high-risk procedures such as open-heart surgery, not more basic procedures such as colon-cancer resection. Do you think the volume-affects-quality argument is more valid with more complex procedures?

There is clearly some level of complexity to certain surgeries that means the more you do, the better the results. But we have to remember that there are two types of volume: the surgeon’s volume and the hospital’s volume.

The hospital’s volume reflects not just the skill and expertise of the physicians, but factors like the nursing staff and how often they see this type of procedure or patient; the pathologist and how well he or she is able to prognosticate the patient’s disease from looking at the sample; and the anesthesiologist’s experience in providing anesthesia for that particular operation.

I think that there can be a relationship between volume and quality with pancreatic cancer, because there is some level of complexity to the procedure. So volume may become important in the long-term outcomes of certain procedures.

Heart surgery is the one procedure that corporate America and many report cards view as the sine qua non example of what to look at in terms of experiences, and not just in terms of volume, but also complication rates. Those rates, however, have to be taken with a grain of salt. The sicker patient almost always goes to a particular surgeon, and that surgeon is the one who will end up treating the sicker patient. So some of those volume-based measures are problematic.

What role do physicians such as hospitalists play in complication rates and outcomes in post-surgical patients?

Our study didn’t examine this, but I think we’re learning more and more about the importance of hospitalists in the care of these patients. The literature showing that hospitalist use is resulting in shorter hospital stays and potentially fewer re-hospitalizations is growing.

And as physicians and surgeons have become increasingly busy, their ability to spend time in the hospital with their patients has become more limited. As those pressures on the outpatient setting have increased, there is clearly an important role for the hospitalist. But I can’t venture whether there’s going to be a long-term outcome difference.

Where do we go next in evaluating how varied processes of care affect outcomes?

We clearly need to look at the in-hospital processes of care and their effect on long-term outcomes. There is an effort underway right now that should help with that.

The National Cancer Institute has a new grant program, the Cancer Care Outcomes Research and Surveillance Consortium (CanCORS). This program will prospectively look at 5,000 patients with colon cancer and 5,000 patients with lung cancer at various of stages of the disease and follow their care.

That should give us a prospective way of looking at how the processes of care affect quality. I think that research will provide us with a lot of very interesting information about how care processes “in inpatient and outpatient settings “affect outcomes.

Bonnie Darves is a freelance writer specializing in health care. She is based in Lake Oswego, Ore.