Hospitals are then ranked in quartiles based on their self-reported answers.
Dudley and his colleagues looked at 155 hospitals that took part in the 2006 survey and found little difference among those in the highest and lowest quartiles. The in-hospital death rate from lowest to highest quartile was: 1.97 percent, 2.04 percent, 1.96 percent and 2 percent.
Last year, the survey was reduced to questions about 13 practices. The streamlined survey also did not correlate to patient death rates.
Hospital administrators have complained that the Safe Practices Survey is too cumbersome and the questions too vague to assess safety.
"We strongly believe some of the Leapfrog issues don't correlate to patient outcome," said Debby Rogers, vice president for quality and emergency services at the California Hospital Association, which represents 400 hospitals in the state.
Nancy Foster, vice president for quality and patient safety policy for the American Hospital Association, which has 5,000 member hospitals, said several survey questions, such as whether a hospital has an informed consent procedure, would have limited impact on death rates.
"Most of these safe practices are the right thing to do. They could avert an error, but they are unlikely to have led to the prevention of death," Foster said.
HealthDay was unable to reach Leapfrog for comment.
Previous research has shown it's not easy for patients to find solid information about hospitals or individual physicians. Last year, a study in the Archives of Surgery that looked at six hospital-comparison Web sites found there were inconsistent results, and the sites used inappropriate or incomplete standards to measure a center's quality.
The sites included the U.S. government's "Hospital Compare," "Quality Check" from the Joint Commission on Accreditation of Healthcare Organizations and Leapfrog's survey.'/>"/>
All rights reserved