Survival benefit for cancer trial participants does not persist in adjusted analyses

In studies using designs addressing sources of bias and confounding, there is no evidence of a survival benefit for cancer trial participants, according to a study published online May 20 in the Journal of the American Medical Association to coincide with the annual meeting of the Society for Clinical Trials, held May 19–22 in Boston.

Renata Iskander, from McGill University in Montreal, and colleagues conducted a systematic review and meta-analysis to examine whether patient participation in clinical trials is associated with greater survival benefit than usual care using data from 39 publications comprising 85 comparisons of trial participants and routine-care patients.

The researchers found that when all studies were pooled, regardless of design or quality, trial participants had a statistically significant overall survival benefit (hazard ratio [HR], 0.76; 95 percent confidence interval [CI], 0.69 to 0.82).

However, in study subsets that matched trial participants and routine-care patients for eligibility criteria, survival benefits were attenuated (HR, 0.85; 95 percent CI, 0.75 to 0.97). When only high-quality studies were pooled, the benefits disappeared (HR, 0.91; 95 percent CI, 0.80 to 1.05). When estimates were adjusted for potential publication bias, benefits also disappeared (HR, 0.94; 95 percent CI, 0.86 to 1.03).

“These findings may strike some trial advocates as discouraging, given how hard they work to improve patient outcomes within trials,” the authors write. “However, a more reassuring interpretation is that there is no evidence that excluding patients from trials due to geography, nonavailability of trials in their condition, or ineligibility deprives them of survival opportunities.”

Copyright © 2024 HealthDay. All rights reserved.