Goin' Home

As I was walking through the hospital yesterday, a woman's calm, soothing voice came over the PA: "5-6 CCU, adult code blue." PA announcements are now very rare, fortunately, and they've learned to keep them cheery sounding and enigmatic when they are necessary, both of which contribute greatly to a less horrific experience for the inmates. However, as you have no doubt guessed, this was a call for the crash cart.

So it got me to thinking. While I'm sitting here pondering and writing and what not (mostly what not), not too many yards away from me, people are dying. So I got curious: what percentage of people die in the hospital? Where do Americans die? It turned out that this question was suprisingly difficult to answer. There is quite a lot of research done in Canada, where apparently their data systems make it easier. But in the U.S., it's mostly state-specific or based on samples intended to address specific comparisons such as race or disease, so it is hard to put together the big picture.

I did find a study by Weitzen et al (Medical Care;41(2):323-335) which was published in 2003, but uses data from a national survey done in 1993. Back then, they found that 58% of deaths occurred in hospitals, 22% at home, and 20% in nursing homes. The older people were, the more likely they were to die in nursing homes. Black and Hispanic people were considerably more likely than gringos to die in hospitals. These authors note that in the early 20th Century, it was far more common to die at home; that by the 1980s three quarters of deaths occurred in insitutions; and that the introduction of the Medicare Prospective Payment System in 1983 was associated with a decrease in hospital deaths and an increase in nursing home deaths. The latter is not surprising: hospitals no longer got paid for every day they kept people, instead they got a lump sum based on diagnosis, so it became in their interest to move people out.

The most recent relatively comprehensive data I could find was by Mitchell, et al (Journal of the American Geriatrics Society 53:299-305), published in 2005 and using data from 2001. They were specifically concerned with the location of death for elderly people with dementia, so they don't aggregate their numbers completely, but they present data that enabled me to do it (after a correction published a few months later). Their sample is limited to people age 65 and older, but that's most of the people who die.

About 1 3/4 million Americans age 65 and older died in that year. About 47% of them died in hospitals, 28% in nursing homes, and 21% at home. A small number died in other places, I presume while away from home, at the scene of a traffic accident or cardiac arrest. People with dementia, however, were far more likely to die in nursing homes -- 66.9% met their end there. People with cancer, interestingly, were considerably more likely to die at home -- 37.85 -- than people who died of other causes.

These authors seem to think it unfortunate that so many people with dementia die in nursing homes, since most people say they would prefer to die at home, but I don't personally have a problem with that. My father died in a nursing home, and he had no idea where he was or what was happening. If he had been at home my mother would have been burdened with the death watch and with his care, which was best left to professionals. However, I don't see much sense in people with cancer dying in hospitals, as 35.4% did. The remainder -- the majority -- Mitchell et al just lump together as "other conditions," so it's hard to make much sense of the numbers. The slight majority of these people with "other conditions" die in hospitals.

So there you have it - the predicted trend toward more people dying in nursing homes and fewer in hospitals since 1993 did occur, but it has been slow. In 2001, fewer people died at home than in 1993, with nursing homes drawing from both other settings. I don't think that's necessarily bad. It all depends.