Assessing Surgery Risks

October 1, 2006

Duke Medical Center surgeons have developed a simple scoring system based on five patient characteristics that can predict which candidates for gastric bypass surgery would have the highest risk of dying. Gastric bypass surgery, a weight-loss-inducing surgery for people who are morbidly obese, involves stapling off a large portion of the patient's stomach. Roughly 170,000 Americans underwent gastric bypass surgery in 2005, according to the American Society for Bariatric Surgery.

The surgery is generally safe, according to the Duke researchers, but carries the risks of any type of surgery, including adverse side effects or death. The key, they say, is determining which patients are at the lowest risk.

The researchers analyzed the outcomes of all 2,075 patients who underwent the surgery between 1995 and 2004 at Virginia Commonwealth University, where Eric DeMaria, Duke's director of bariatric surgery, worked before coming to Duke in 2005.

They found that thirty-one patients (1.5 percent) died within ninety days of surgery. In analyzing those who died, DeMaria and his colleagues identified four factors that were independently predictive of increased risk: a body-mass index of greater than 50 (a range of 18.5 to 25 is considered normal); male gender; hypertension; and risk of pulmonary embolus. In addition to the four primary factors, the researchers added a fifth—age—to their scoring system. Based on the results of past studies, patients over the age of forty-five are known to be at the highest risk for death after bariatric surgery.

"In using our system, each one of the five characteristics is worth one point," DeMaria explained. "Those patients with a score of zero are at the least risk, while those with five points are at the highest risk."

If validated by additional prospective studies, the new scoring system not only would give surgeons concrete information on which to base treatment options, but also would help patients make informed decisions about potential risks.