How many bankruptcies are medical?
Thursday, June 4 2009
Heard about that new study of medical bankruptcies? I don’t think it proves what it says it does. A lot of the blogging on the study has apparently been in response to the press release, but here’s the link to the actual .pdf of the study. Reading that document raises more questions that it answers.
The article is maddeningly vague about the model and variables used, which makes parsing the results more difficult than it should be. For example, the only measure of medical debt in the model is an ordinal variable of <$1000, $1000-$5000, or >$5000 (p. 2). Anybody who falls into the last category counts as a “medical bankruptcy” in their analysis (p. 3). But in the analysis, the mean net worth (“assets – debts”) of a medical bankruptcy is about -$45,000. That means a person could have $5000 in medical debts, $40,000 in other debts in excess of their assets, an unknowable amount of other debts to match their assets, and still qualify as a medical bankruptcy. In theory, you could have $400,000 in assets, $440,000 in unpaid library fines, and the study would still call you a “medical bankruptcy” for having $5,001 in medical expenses.
Moreover, the difference between the net worth of people counted as medical vs. non-medical bankruptcies is about $7,000, and is not statistically significant. This means that a bankruptcy with as much as $7,000 in medical debt (and possibly much more) is indistinguishable from a bankruptcy with $5,000 or less. You could make an argument about straws, camels, and broken backs, but it would be a stretch for a significant number of the cases.
So far, this is only a sideshow for the real issue: the biggest problem with the study is that it never tests its main claim. They say that medical problems “contribute” to bankruptcy, but that hypothesis never shows up in their models. It’s like paying for the circus, but only getting the clowns.
And it ought to be easily testable: instead of running the model on two separate groups, they could have included a dichotomous (i.e. dummy) variable for cases meeting their definition of medical bankruptcy across the total sample. Or they could have included a variable indicating the level of debt. Or they could have run similar models on the medical, non-medical, and total samples, then figured out from there what difference medical bankruptcy makes. But they didn’t do any of that, and this is a real problem; without a direct test, the study doesn’t offer any real evidence that medical debts make a difference in bankruptcy, and I have a hard time accepting its claims.
I realize that puts me in opposition to a great many progressive bloggers who have reported on this study – including Tim Foley at Change.org, Paul Testa at New Health Dialogue, Jason Rosenbaum at NOW! Blog, etc. Even Merrill Goozner posted the study’s numbers. (By contrast, here’s Gary Langer of ABC news questioning those numbers.) In fact, you should know that the lead author of the study is the National Spokesman of Physicians for a National Health Plan, and a co-author is the current Secretary of that organization. PHNP, of course, is one of the leading voices for single-payer health care in this country, and I have a statisically significant amount of respect for Drs. Himmelstein and Woolhandler.
Like the authors – and other progressives – I am concerned about the number of people being driven into bankruptcy because of medical debts. It’s embarrassing – beneath us as a nation – the extent to which health care pushes people into debt and poverty in this country. Moreover, this is a very real, very immediate problem for me: I am broke, and right on the cusp of not being able to afford my COBRA payments. I fully agree that keeping sick people from going bankrupt is among the many important goals of health care reform.
But I don’t think it serves our cause to inflate, exaggerate, or distort the extent to which medical bankruptcy is a problem. We should be as careful and precise in our assessment of the problem as possible. I have written previously about the AHA’s “cost-shifting” study, that it’s a blatant ploy to frame the debate in terms favorable to the hospital industry. The sad fact is that this PNHP study is not much better, if that; it comes dangerously close to distorting the evidence in favor of ideology.
It’s not that I have a problem with ideologically motivated research. Plenty of useful science gets done by people motivated by ideological agendas. I don’t think progressives should refrain from exploring the many problems of our current health care system. I do believe, however, that we need to hold that research to a very stringent standard.
We – progressives – ought to trust that the evidence supports our views. And we need to be diligent and careful in how we assemble and present that evidence. And if the evidence doesn’t support our views, we need to be change our views. The problems with the PNHP study don’t mean we need to change our views, but nor is it credible evidence in their favor.
[Note: this post edited for clarity, 4 Jun 2009]
1. You have a fantastic scheme buried in here that could help out a lot people in major financial straights. You get two people whose medical debt is so high they can hardly pay their bills. In fact, they cannot even afford to maintain their cars anymore. Pad them up well, make sure the airbags work and have a demolition derby moment. They blame it on a pothole, a dog ran into the road and then out. You both walk away with insurance money from your totaled cars and use it to pay off our medical debt. What do you think of that? Am I going to end up in jail just for typing this.
2. On a more serious note, this post is amazing. We need more progressives with your kind of spine. I agree with what you wrote. Medical debt is so rampant that we hardly need to inflate numbers to make it a significant issues.
Best,
Kairol
http://everythingchangesbook.com/
Kairol – I should mention that you’re responding to an earlier version of this post, where I admitted that I am using my insurance settlement from a car accident to pay my bills until I can find another source of income. Sorry for the stealth edit – I wouldn’t have changed it if I had seen your comment beforehand.
The only problem with your plan – apart from the illegality – is that you only get the resale value of the car if its totaled. So our hypothetical people could, in theory, just sell their cars legally and use that money. I will admit that it never occurred to me to sell my car to pay my bills, but now that I have the money and no car, I’m not really in a hurry to get a new one. Of course, I have the advantage of living in a walkable neighborhood with a person who will let me drive her car, so I don’t really need my own car.
As for 2., thanks for the kind words. It’s frustrating to know medical debt is rampant, but that we still don’t have a clear picture of the extent of the problem.
I am a physician, a small business owner, and have a chronically ill child.
I pay several hundred dollars per month (over $800) for health insurance as the small business owner. This is for a high deductible plan. Despite the health insurance my medical bills (over the cost of the insurance) exceed $5000 per year. This has been the case for the last few years. I have not had to file for bankruptcy.
My family has had to change its lifestyle though. My husband had to quit his job entirely as my daughter’s illness required someone to be home with her at least 10 days per month for over a year and these days were not known ahead. Some of the times were emergency. She has 6 doctors that she regularly sees for care at 2 major medical centers in 2 different states. Both centers are over an hour away from our home.
No I am not bankrupt, but the added problems with health insurance have made her illness even harder to manage.