Cancer support-group data challenged - research conducted by David Spiegel of Stanford School of Medicine is questioned - Brief Article

Science News, May 9, 1992

In recent years, carefully designed studies documenting the benefits of support groups for cancer patients have revitalized research in this area. Perhaps the most influential of these studies, directed by psychiatrist David Spiegel of the Stanford University School of Medicine, found that breast cancer patients randomly placed in weekly support groups for one year lived markedly longer than controls -- patients randomly assigned only to regular medical care (SN: 11/4/89, p.302).

But debate has now surfaced concerning the actual effects of Spiegel's intervention and the ethics of denying psychological treatment for cancer to control groups in the name of science. Several researchers discuss these issues in the spring ADVANCES, a journal that publishes articles describing research on the inrerplay of mind and body.

Psychologist Lawrence L. LeShan, who has studied the relationship of emotions to cancer progression since 1950 and now practices in New York City, launched the debate in the same journal last year. "We're dealing an important, unresolved question," LeShan asserts. "Did Spiegel's design extend the life span of the experimental group or shorten that of the control group? Or both?" Many cancer patients harbor a strong sense of rejection by others, LeShan maintains. When researchers tell them about a potentially helpful procedure but then withhold the intervention from some patients in order to obtain a suitable control group, feelings of rejection -- along with death rates -- may soar among controls, he argues.

A preliminary analysis of federal data on U.S. cancer survival rates from 1978 to 1986 -- conducted by psychologist Bernard H. Fox of the Boston University School of Medicine -- offers some support for LeShan's theory. Fox finds that women receiving group therapy in Spiegel's study died at a slightly higher rate than women with breast cancer in the national sample, while controls died at a substantially higher rate. Spiegel's sample came from the San Francisco area and may differ in some respects from the nations sample, Fox point out. Fox plans to pull San Francisco survivals rates out of the national data and compare them with Spiegel's findings.

Until then, Fox considers LeShan's argument "provocative" but open to question. For instance, he says, Spiegel's small sample of controls may have deviated form the national survival rate purely by chance.

Spiegel recjects LeShan's hypothesis. He notes that his team initiated the support groups in the 1970s, before approaches A became popular. Cancer patients in the experimental group reported plenty of reticense about attending group meetings, but only a few controls cited disappointment at missing the chance for group therapy, Spiegel argues. Any therapeutic eggect observed in the study underestimates the survival advantage for cancer patients "in the real world who seek out and pay for supportive therapy, he adds.