Many studies have suggested that patients who have surgery at hospitals that perform a large number of a specific surgery do better than patients who have the same surgery at low-volume hospitals. But two studies in the November 21 Annals of Internal Medicine find that simple calculations of volume vs. outcome usually do not consider patient characteristics, the quality of overall hospital care, the actual cause of death, or "clustering," the tendency of the outcomes of one provider's patients to be more alike than those of a different provider's patients. In one study, researchers re-analyzed data from three of their own previous studies that had shown a volume-outcome relationship. This time, they showed that adjusting for clustering sometimes weakened the relationship between high volume and good outcomes (Article, p. 658). In the other study, colon cancer-specific outcomes did not vary with provider volume even though overall death rate was lower in high-volume hospitals (Article, p. 649). Editorial writers say, "relationships between volume and outcome are neither as straightforward nor as robust as much of the literature suggests" (Editorial, p. 696). They state that efforts to direct patients to providers with the most experience, such as plans to regionalize surgery, may be based on false inferences about individual outcomes from aggregate data obtained from groups.

Topical Treatment Appears Successful for Bedsores

Treating pressure ulcers or bedsores on the foot with topical nerve growth factor completely healed eight ulcers in a group of 18 patients, while, in the control group, only one ulcer healed, a new study found (Article, p. 635). Bedsores are among the most common health problems in nursing homes, particularly in the frail elderly and in people who cannot move about. Bedsores often increase the cost of medical and nur
'"/>

Contact: Penny Fuller
pfuller@acponline.org
215-351-2656
American College of Physicians
20-Oct-2003