Going Beyond Hospital Report Cards

Death rates for hospitals, adjusted for patient sickness, are often used to measure quality of care. Hospitals with high adjusted death rates are then identified as needing improvement. This method has two problems, however: (1) the risk-adjusted death rates include both preventable and unpreventable deaths, and the percentage of preventable deaths is unknown; and (2) many preventable deaths occur at hospitals with average or below average death rates. Thus, these hospital "report cards" can only be a starting point in determining how to reduce the number of preventable deaths.

A study recently published in Circulation analyzed a sample of 347 inpatient deaths following bypass surgery in Toronto hospitals and determined that as many as one-third of the deaths could have been prevented with better care. The researchers also showed that the hospitals' report card grades were not correlated with the rates of preventable deaths. Even in the hospital with the lowest risk-adjusted mortality rate, about 20 percent of deaths were preventable. Preventable deaths occurred frequently among both low and high risk patients, and were actually more common in lower-risk patients. A large proportion of preventable deaths were related to problems in the operating room (86 percent) and intensive care unit (61 percent), and the study produced a list of specific improvements needed in the care of bypass surgery patients.

The method used in this study is one that could be used in other situations to improve quality of care, even in hospitals that are already performing well. In the study, experienced cardiac surgeons who were blinded to patient, surgeon, and hospital reviewed the medical records of the patients who died following surgery to identify preventable deaths. Two reviewers examined the records for each death, and a third reviewer was used when the two reviewers disagreed. For each preventable death, specific improvements in quality of care were identified that could have prevented the death. The use of reviewers who were expert in cardiac surgery allowed the reviewers to identify and provide solutions to the problems identified in preventable deaths.

In an accompanying editorial, Yale cardiologist Harlan Krumholz emphasized that the post-bypass death rates in Toronto hospitals were already low and the system "could have been content with its successes." Krumholz noted that high-profile medical mistakes that clearly cause harm represent only a small portion of the causes of preventable deaths. Thus, an important feature of the study was that the investigators did not require that a preventable death be one that the medical record reviewer was certain was preventable, but only one where the reviewers felt that the likelihood that it could have been prevented was greater than 50 percent. This standard was designed to identify more deaths where care was not optimal so as to maximize the potential for quality improvements.

How can the number of preventable deaths be reduced? Krumholz advocates that hospitals routinely and systematically identify problem areas that contribute to bad outcomes. Quality improvements that reduce or eliminate these problems should reduce the number of preventable deaths, even though it generally will not be possible to determine the particular deaths that were prevented. Krumholz notes that there is a need for a shift in culture so that assessments of preventable deaths are part of the expectation of every hospital, provider, and patient. "How else will we attain the goal of creating the high-reliability, high-performance institutions that we prefer for our practices and that our patients deserve?"

-- PM