The academic group that put out a much-covered and deservedly much-debated study on so-called "medical bankruptcies" a few years ago has released an update today. In some important ways it improves on the earlier version, which I’ve blogged about before. But in other, fundamental ways, reservations remain in order.
The study again was led by a Harvard professor who co-founded an advocacy group for a single-payer national health care system. From my perspective, it makes leaps, inflates numbers and questionably characterizes results. Its authors clearly have a different view; since these are the sorts of numbers that can inform presidential pronouncements and policy directions alike, a good look is warranted.
The study, published in The American Journal of Medicine, concludes that "62 percent of all bankruptcies in 2007 were medical." In fact, though, the survey on which it’s based asked respondents if medical expenses or problems were one of “a number of different reasons” for their bankruptcy. (“People file for bankruptcy for a number of different reasons. Why did you, or a spouse or partner, file for bankruptcy? Check all that apply.”)
It strikes me as a mischaracterization to portray any one item or group of items in the list that followed as the sole or even chief cause of bankruptcy. The study seems to me to do just that, using phrases such as “medical debtors,” “medical bankruptcies” and, in the accompanying news release, “bankruptcies attributable to medical problems.” Its key data table, titled “Medical Causes of Bankruptcy,” lists various items as “reason for bankruptcy,” dropping the qualifying word, “a” reason.
In reality there could be any number of equally important or more important contributors. The survey could have asked respondents to rank them in importance; it didn't. Given what was asked, I can see saying that medical bills or problems contributed to bankruptcies – not suggesting they caused them.
A next question is how the researchers got to that 62 percent. The number of respondents who actually said medical or health care bills were among the reasons for their bankruptcy was 29 percent. Thirty-two percent also said their or a spouse’s medical problems (not specifically bills) were among the reasons, and 11 percent cited another family member’s medical problems (not specifically bills). Aggregating these, the study says 44 percent cited medical bills or medical problems as a reason for bankruptcy.
That’s a plenty big number, and cause for real concern. At the same time, sticking with what the survey actually asked, it’s not about medical costs per se, but about either medical bills or undefined “medical problems.” And again these are not the reason, but one of a number of reasons, for bankruptcy.
It’s also not 62 percent. The researchers arrive at that sum by adding in people who do not report a medical reason for their bankruptcy, but fit other conditions – e.g., lost two weeks or more of income because of their or a family member’s illness, or had more than $5,000 in uncovered medical expenses in the previous two years. These people are counted as having a medical problem among the reasons for their bankruptcy even though, when given an opportunity to say so themselves, they did not.
In another headline result, the authors report a nearly 50 percent increase since 2001 in bankruptcies for which medical problems were among the reasons. I’m having trouble seeing how they got there, because the two surveys asked different questions (as well as being done among different populations). We’ve asked for details, and I’ll update this note when we learn more. (Separately, the study describes people who used a home equity line of credit to pay medical bills as having “mortgaged [their] home” to do so.)
As I noted at the start, this study is in some ways better than its 2001 predecessor. It uses a national sample as opposed to a sample in just five federal court districts. The “greater-than-$5,000 in medical expenses” condition is less bad than the greater-than-$1,000 threshold used in '01. And some of the most debatable ’01 conditions have been dropped, such as a gambling problem or the death of a family member, both previously included as a reason for “medical bankruptcy.”
Again, the lead author, Dr. David Himmelstein, is a co-founder of the “Physicians for a National Health Program,” an advocacy group calling for a “public or quasi-public” single-payer health system. One wonders whether other researchers, with other perspectives, might have presented the data differently.