Differences in Length of Hospitalization Studied : Medicine: Contrast in patients’ conditions and their outcomes don’t justify most variations, study finds.
A heart attack patient spends 12 1/2 days in a prominent Boston hospital. In California, a comparable patient gets out in fewer than nine. A puzzling question in health care is: What accounts for that pricey, 3 1/2-day difference?
A new study of men and women treated in six California and Massachusetts hospitals challenges the usual explanations. The study found that neither differences in the patients’ conditions nor in their outcomes justified most disparities in lengths of stay.
“In terms of long-term outcomes, a day or two shorter length of stay for most of these conditions probably will do no harm,” said Dr. Sheldon Greenfield, an author of the paper published today in the Journal of the American Medical Assn.
“It’s important in the big cost crunch,” said Greenfield, a professor of medicine at Tufts University School of Medicine in Boston. “If everybody is trying to save money, any money that can be shaved here and there is worth shaving.”
Others, however, said it is important first to determine what medical practices and procedures tend to lengthen hospitalizations, and whether those procedures make a significant difference in the quality of patient care.
“This gives us a mechanism for monitoring and maximizing quality of care,” said Paul Cleary, a Harvard Medical School researcher and lead author of the paper. “Our main goal is not to save money. . . . Our main goal is to be very effective in providing care.”
It is well known that medical practices vary throughout the United States. Surgery rates, for example, have long differed from one region to another. In recent years, studies have also found regional differences in rates of hospitalization and lengths of stay.
In general, patients stay longer in hospitals in the Northeast than on the West Coast, research suggests. But experts disagree on the extent to which differences in severity of illness, overall health, and social and demographic attributes of patients explain the gaps.
In the new study, the researchers examined the cases of 2,484 patients who had been hospitalized for one of six conditions at six teaching hospitals, including UCLA, UC San Francisco and Stanford medical centers.
The researchers began with patients who were relatively similar to one another. They also used statistical calculations to take into account differences in severity of illness, overall health, and other medical factors that might explain disparities in length of stay.
“We wanted to see if we could explain the differences in length of stay on the basis of case mix,” said Cleary, an associate professor of health care policy. “And if we could not, did those differences have impact on the health of the patients?”
The group found clear differences in length of stay for five of the six medical and surgical procedures--cardiac bypass graft surgery, gall bladder removal, prostate surgery, total hip replacement and treatment after a heart attack.
For example, the average length of stay for heart attack treatment at one Boston hospital was 12.4 days, compared to 8.3 at one on the West Coast. For prostate surgery, patients in one Boston hospital stayed 6.5 days, compared to 3.7 in one California hospital.
Only in the case of total hip replacement did differences in the types of patients seem to justify variations in length of stay. In all the other cases, patient characteristics explained little of the variation. The disparities remained unexplained.
The researchers also found that the time a person spent in the hospital seemed to have no significant impact upon subsequent risk of death, their ability to function after discharge, their probability of readmission or their satisfaction with their care.
Cleary and Greenfield said it remains to be seen what might account for the differences among hospitals. One possibility is that procedures, such as certain time-consuming diagnostic tests, are done routinely in some hospitals but not in others.
“There is a pressing need for . . . research to help determine the specific medical practices that are related to longer lengths of stay,” they wrote in the paper. “Only when such data are available will we be confident that it is possible to reduce length of stay without compromising the health and safety of patients.”
