Medical researchers at the College have concluded that the amount of medical care Americans receive in their final months varies widely with geographic location, and is often a function of the amount of care available in the region, rather than of necessity.
The Dartmouth Atlas of Health Care 1998 study, released yesterday by the American Hospital Association and covered by The Washington Post and the Associated Press, examines the likelihood that an individual is in a hospital at the time of his or her death, Dartmouth Atlas Editor Megan McAndrew Cooper said.
The study looks at all Medicare beneficiaries who died in the two-year period from 1994 to 1995 and focuses on the circumstances of the last six months prior to death in each case.
The study also took into account differences in the age and the rates of illness of residents living in different communities across the country.
Cooper said the research is particularly interesting since past medical studies concerning preferences have found 80 percent of people do not want to die in a hospital bed, yet it appears that up to 50 percent are indeed hospitalized at the time of their death.
The study draws some important conclusions regarding the respect for patient preferences, as well as the amount of federal Medicare money wasted in supporting dying patients who do not want such treatment.
In addition, the Dartmouth Atlas research suggests that the likelihood of dying in a hospital bed is highly correlated with the local supply of these beds.
"If there are a lot of beds, you're more likely to get put in one," Cooper said.
"When people are sick, a doctor might have to make a decision about whether to put them in the hospital to treat them in the last week of their life," she said.
Cooper said there are a greater number of hospital beds per capita in the southeastern United States and in urban areas than there are in the Midwest or West.
As a result, due to the influence of supply on treatment, people in the eastern United States are much more likely to be hospitalized during the last six months prior to their death. In addition, the likelihood of undergoing various common medical and surgical treatments is much greater in these areas.
This "powerful influence of supply," Cooper said, "is absolutely related to geographical location and the local supply of beds and doctors."
In Ogden, Utah, for example, the average number of days spent in a hospital bed during the six months prior to death was only less than five, a nationwide low.
More than 2,000 miles away, in Newark, N.J., patients spent almost 23 days, on average, in the hospital over the same time period.
This correlation between available local supply and treatment given extends not only to the likelihood of being in a hospital bed at the time of death, but also of being in a hospital bed or an intensive care unit within six months of death.
Even in communities with similar demographics, Cooper said, the medical treatment given to individuals in those two communities is very different.
Individuals in St. Petersburg, Fla. and Sun City, Ariz. -- both relatively wealthy retirement communities -- nonetheless received vastly different medical care in the six months prior to their death, according to an article in The Washington Post.
While Sun City residents averaged only a half-day in intensive care in the six months prior to their deaths, St. Petersburg residents spent an average of almost five days, according to The Washington Post article.
The Dartmouth Atlas of Health Care is a new organization at the College which looks at differences in health care resources and utilization in the United States, Cooper said.
