The differences may not seem great at first, but considering how large the sample size, they are likely significant. While 9.8 percent of patients of five-star hospitals died within 30 days of being discharged, 11.2 percent of patients in one-star hospitals perished.

Similarly, five-star hospitals readmitted less than a fifth of their patients within a month of discharge, while hospitals with lower ratings all saw at least a fifth of their patients again in that timeframe.

“If you use the star rating you’re more likely to end up at a high quality hospital,” Dr. Ashish K. Jha, a study co-author, told Reuters. “But I wouldn’t use only the star rating to choose a hospital.”

Dr. Joshua J. Fenton of the University of California, Davis, pointed out that the hospitals with the highest ratings also shared a number of characteristics that weren’t necessarily linked to the quality of their service. No large hospitals received five stars; nor did any hospitals that included an intensive care unit.

“I don’t think these data are enough to by themselves to suggest that (patients) should use the star rating as a single guide to choose an institution,” he said.

Granted, the reliability of readmission rates as performance measures has been the subject of skepticism as well. Indeed, a separate study from Harvard last year suggested that hospitals with higher than average readmission rates often serve lower-income populations that are more likely to suffer from chronic conditions and less likely to manage their illness as prescribed because of poor access to necessary services, including a primary care physician or a pharmacy.