Volume and outcomes related in cancer surgery

Experience leads to lower mortality rates

Are more experienced surgeons necessarily better ones? In some complex procedures, research indicates that they are.1,2 That connection between case volume and outcomes has spurred some physicians to advocate directing patients toward specialty centers, particularly for high-risk cancer surgery in which mortality rates vary greatly.

"Imbedded in high volume are well-trained people, dedicated staff and equipment, and well-established clinical protocols," says Toby Gordon, ScD, vice president of strategic planning and marketing for Johns Hopkins Medicine in Baltimore and associate professor of surgery and health policy and management at Johns Hopkins University School of Medicine. "We’ve been able to established over and over the relationship between volume, cost, and outcome," she says. "The more you do, the less it costs per case, and the better the clinical outcome. That’s after case-mix adjustment."

Gordon analyzed data from 795 cases at 43 Maryland hospitals of patients undergoing pancreaticoduodenectomy, also known as the Whipple procedure, a complex operation for pancreatic cancer. The operative mortality rate over the 12-year study period ranged from 1.8% to 14.2%.1

The study used a cut-off of at least 20 procedures a year to separate high-volume from low-volume hospitals. Only Johns Hopkins performed more than 20, with an average of 51 per year, while the others had an average of slightly less than one per year.

The research on the Whipple procedure grew out of concerns from managed care organizations that the care at Hopkins was too costly. The payers were "very receptive to the data," says Gordon, and as the number of procedures performed at Hopkins grew, the statewide mortality rate associated with the procedure fell. A similar connection between outcomes and volume exists with other complex procedures, such as esophagectomy and liver resection, she says.

Evidence for a link between volume and outcomes in surgical oncology grew stronger with another study at Memorial Sloan Kettering Cancer Center in New York City.

Colin Begg, PhD, head of the cancer center’s department of epidemiology and biostatistics, studied short-term operative mortality for 5,013 patients using a Medicare database. The patients underwent pancreatectomy, esophagectomy, pneumonectomy, liver resection, or pelvic exenteration, or procedures for cancers of the pancreas, esophagus, lung, colon, and rectum, and genitourinary cancers. "There were trends favoring high-volume hospitals for all five of the procedures although the strength of the trends varied, and it was statistically significant only for four of them,"2 he says. "It was a fairly restricted study, but it gave us very clean-cut results."

Begg plans to expand his research into these broader questions: Do specialization and expertise impact other outcomes? When is expertise important and when does it have little effect?

"It’s not easy evidence to assemble," he says. "If there is a value-added for specialist care, it may show up in quality of life outcomes or morbidity from cancer treatments. These kinds of endpoints are not generally available in administrative databases."

What might be done in response to evidence that expertise in some complex surgeries reduces mortality? Payers can and often do steer patients to regional centers for certain difficult and costly procedures, and hospitals may set volume standards as a part of credentialing.

However, states and medical societies have been reluctant to regulate a minimum volume, although studies do show the potential for a "cut-off."

"Is there a linear relationship, so the more you do, you get better and better? Usually it levels off at some point, so that is the minimum," says Gordon.

Yet setting a lower limit may be more problematic than it seems. When the Chicago-based Society of Thoracic Surgeons considered the relationship between volume and outcome in coronary artery bypass graft surgery, the society’s committee chose not to recommend a minimum standard.

One study found a higher ratio of observed vs. expected mortality among practices with less than 100 cases per year.3 A study of 44 Veterans Affairs hospitals showed a relationship of higher observed to expected mortality at low-volume hospitals but didn’t find a volume threshold.4 The society’s Ad Hoc Committee on Cardiac Surgery Credentialing determined that volume should not be used as a criterion for credentialing but that surgeons should track their outcomes and compare themselves to peers with a national database.

"There’s some relationship [with volume and outcomes], but it’s not as tight of a relationship as you would like to develop a standard," says Don Turney, the society’s assistant executive director.

A minimum standard can have unanticipated consequences, he notes. "There are lots of reasons that [physicians] may have low volume for a particular year that has nothing to do with surgical ability. They may have broken an arm; they may have been on vacation," he says.

Instead, the society has focused on a national database that allows surgeons to compare their outcomes with peers and to use the data for quality improvement. "This databasing effort has had a significantly positive effect in outcomes and improving care," says Turney.

Meanwhile, Bruce Hillner, MD, professor of medicine at the Medical College of Virginia at Virginia Commonwealth University in Richmond, says he hopes the research on variation in mortality from cancer procedures will spur a greater focus on outcomes in oncology.

"It’s a very major challenge," he says. "The issues of practice variation and indirect quality indicators are not high on the national agenda in the cancer scene."


1. Gordon TA, Bowman HM, Tielsch JM, Bass EB, et al. Statewide regionalization of pancreaticoduodenectomy and its effect on in-hospital mortality. Ann Surg 1998; 228:71-78.

2. Begg CB, Cramer LD, Hoskins WJ, Brennan MF. Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998; 280:1,747-1,751.

3. Clark RE. Outcome as a function of annual coronary artery bypass graft volume. The Ad Hoc Committee on Cardiac Surgery Credentialing of The Society of Thoracic Surgeons. Ann Thorac Surg 1996; 61:21-26.

4. Shroyer AL, Marshall G, Warner BA, Johnson RR, et al. No continuous relationship between Veterans Affairs coronary artery bypass grafting surgical volume and operative mortality. Ann Thorac Surg 1996; 61:17-20.