In Massachusetts, a state with a reputation for top medical care, hospital patients die at a higher rate than the national average, according to an analysis that weighted the data to account for the state's relatively low poverty rate and abundance of specialized physicians.
The analysis by a prominent Cambridge research organization also found wide variation in death rates among
Massachusetts hospitals, with patients about half as likely to die at some hospitals than at others.
In 2005, the most recent year for which the Institute for Healthcare Improvement provided mortality rates, the overall death rate for the state's hospitals was 7 percent higher than the national average.
Given the low level of poverty in Massachusetts, the high number of specialists, and other factors, "you would expect it to be even better than the national average," said Sir Brian Jarmon, a British physician who developed the methodology the institute uses to calculate mortality rates.
The institute collects data about patient care from Medicare and hospitals, and uses demographic and health characteristics to adjust the data to allow comparisons among hospitals. It pro vides the data to hospitals it works with on quality-improvement projects.
A consumer advocacy group, Health Care For All, obtained the data from the institute and provided a copy to the Globe. The report did not identify hospitals by name.
Health Care for All, headed by former state representative John McDonough, plans to send a letter to state health officials today demanding that they take immediate action to improve care at hospitals with significantly high mortality rates, and make public mortality data like the institute's figures for individual hospitals.
"A lot of people believe Massachusetts is the best place to get your healthcare in the country, if not the world," said Deborah Wachenheim, consumer health quality coordinator for Health Care For All. But "there are potentially some problems as well. We need government officials to look into what's going on and start working with hospitals that have the higher rates to bring them down."
The institute's methodology for calculating the data, called the Hospital Standardized Mortality Ratio, is not universally accepted in the United States.
Karen Nelson, a nurse and senior vice president for clinical affairs for the Massachusetts Hospital Association, said the institute's approach has not been validated as an accurate measure of hospital care by leading US health quality groups such as the National Quality Forum.
Even though the association is in favor of making some hospital performance measures public, she said the mortality data are not useful for hospitals trying to improve or for consumers making choices about where to get care. The number is general, she said, and does not show mortality for specific conditions or procedures.
Medicare, the federal health insurance program for the elderly and disabled, recently started posting on its website, www.hospitalcompare.hhs.gov, how individual US hospital mortality rates compare with the national average for heart attack and heart failure patients. For example, it says, of 65 hospitals in Massachusetts, 62 have death rates for heart failure equal to the US average, while three are better than average.
But the site does not show a hospital's actual mortality rate, and Medicare includes deaths that occur up to 30 days after discharge. IHI's data include only deaths while a patient is hospitalized.
The state's higher-than-average hospital mortality rate calculated by Jarmon does "raise questions about what's different about healthcare in Massachusetts," Nelson said. One explanation could be the large number of patients in intensive care units, which care for the sickest patients who are most likely to die.
But others are convinced that the mortality rates can be a powerful indicator of hospital quality.
Among 56 Massachusetts hospitals for which IHI computed mortality ratios - the number of actual deaths divided by the number of expected deaths, then multiplied by 100 - the range was 65 to 119. The Massachusetts average is 89.3 with the US average at 83.5 when adjusted for a wide range of patient and demographic characteristics. The lower the number the better.
Jarmon said Massachusetts hospitals with higher-than-average rates could have submitted inaccurate data to Medicare, but "it also could be that the care is not as good."
"We've used this data for several years to help hospitals see deaths and make improvements," said Dr. Donald Berwick, president of IHI, who said public reporting catalyses hospitals to improve. "I am pretty confident this is good data. I would put it all out there."
Berwick said his organization has decided not to release the data with hospital names because it is concerned it would harm the institute's relationship with hospitals, though he encourages hospitals to publicize their mortality rates.
A year ago, Paul Levy, president of Beth Israel Deaconess Medical Center in Boston, posted his hospital's mortality ratio calculated by IHI - 71 - on his blog and called on all Massachusetts hospitals to make their rates public.
Health Care For All's letter is addressed to John Auerbach, commissioner of public health, and Nancy Achin Audesse, executive director of the Board of Registration in Medicine, which licenses doctors. The group wants health officials to determine in the next three months the most reliable mortality measure.
The health department, which regulates hospitals, said it will explore Health Care For All's request.
Liz Kowalczyk can be reached at firstname.lastname@example.org.