Unexplained differences in hospital death rates

Most of the two-fold variation between U.S. hospitals in heart attack death rates remains a mystery, according to a new study.

Researchers found that factors often assumed to influence the quality of hospital care - such as location, size, and the economic and social standing of patients - explained less than 20 percent of the differences between institutions in patient survival after acute myocardial infarction, a severe form of heart attack.

The findings, the investigators say, point to a need to identify more direct measures of quality to use in targeting hospitals that need improvement.

Just getting to a larger teaching hospital, for example, was not enough to ensure a heart attack victim a better chance of survival, lead researcher Elizabeth Bradley of Yale University in New Haven, Connecticut, told Reuters Health in an e-mail.

Bradley and her colleagues investigated 513,202 post-heart attack discharges from 2,908 hospitals between July 2005 and June 2008, on a hunt for signals about where care might be better.

On average, just over 16 percent of a hospital's heart attack patients died within 30 days of admission, report the researchers in The American Journal of Cardiology. But they found that these mortality rates ranged widely among institutions, from 11 to 26 percent.

After adjusting for differences in the patients' medical history, other current illnesses, age and gender, the team uncovered several characteristics of hospitals and their patient populations that were tied to a portion of this variation.

Hospital features associated with better post-heart attack prognosis were large numbers of beds and of heart attack discharges, connection with an academic medical center, an urban location and high socioeconomic status in the patient population - a measure that encompasses education, income, occupation and housing value.

However, the influence of those factors was modest. The difference in survival rates between hospitals where more than two thirds of patients came from the lowest socioeconomic-status zip codes, for example, versus hospitals where just 2 percent of patients came from those areas, was less than 1 percent.

Combining all the significant factors only explained about 17 percent of the total disparities in heart attack survival across the nation's hospitals.

The researchers point out that the observational nature of their study restricts them from proving any direct cause-and-effect relationships between specific factors and death rates.

Nevertheless, the findings add to evidence hinting at what does - or more often what does not - seem to affect the quality of medical care in the U.S. The Dartmouth Atlas Project, for example, has assessed variations in health care spending across the U.S. and found that more money does not always buy better care.