by Allan Joseph
Welcome to Week 4 of Project Millennial’s 2013 summer journal club. Previous posts can be found here. This week and next week, we’ll discuss whether insurance works, using some of the most famous research in the field. This week’s post focuses on “Does Medicare Save Lives?” by David Card, Carlos Dobkin, and Nicole Maestas.
We know people buy health insurance to turn unpredictable expenses into predictable ones — that’s was the first post of this summer’s Journal Club. But what we haven’t talked about is this: Does health insurance even work?
It’s important to figure out exactly what health insurance “working” means. For our purposes, we’ll look at two separate questions: 1) does health insurance provide financial protection for those who have illnesses, and 2), does health insurance have any effect on health itself? The first is a really important question to answer, and many believe that’s the core function of health insurance. While there’s plenty of research on the first question, it’s the second question we’ll focus on today — does having health insurance make you healthier, and can more generous health insurance help, too? Card et al. put together a clever analysis of this question using Medicare as an example, but understanding the paper takes a bit of patience. It’ll be worth it, I promise.
Medicare, the national health-insurance program for the elderly in America, is one of the most generous forms of health insurance available. The elderly can sign up for Medicare on the first day of the month in which they turn 65, and Medicare Part A, which covers hospital bills, is free. As you can see from the figure, turning 65 results in a big jump in how many people have health insurance:
This gives us a great opportunity to study what happens when people sign up for Medicare, whether they were previously uninsured or whether they simply had less-generous insurance. The key assumption here — and this is really crucial — is that the only important difference between the people who just turned 65 (and received Medicare) and the ones who haven’t yet is the presence of Medicare. If that’s true, then we can treat this as a quasi-randomized trial. The authors go to great lengths to document that this is true in this study, and they’ve designed it so that the biggest possible problem doesn’t apply.
As Card et al. note, patients often wait to address elective issues (think hip replacement, or back pain) until after they enroll in Medicare, which creates a difference that could ruin the analysis. But there’s a solution to this problem: analyzing only conditions that people can’t wait to address (heart attack, stroke, trouble breathing). People are going to the emergency room regardless of how old they are when they have these issues, and thanks to a 1986 law, hospitals have to treat them until they are stabilized, even if they don’t have insurance. If you look only at the true emergencies, you can get a good idea of what happens only because of the difference between Medicare and other insurance (or not being insured at all), and not any other confounding factors.
Using emergency-department records and death records from California hospitals over an 11-year period, the authors were able to examine two questions: Does having Medicare affect how much treatment a patient receives, and does having Medicare affect how likely a patient is to survive? Fortunately for us, they come up with clear answers.
As the figure shows, patients did receive more treatment once they moved past age 65 (and thus were eligible for Medicare). It’s a small result of only about 3-5% more care, but it is statistically significant and suggests having Medicare allows patients to receive more treatment beyond that what is required by the law.
If they’re getting more treatment, do patients who are Medicare-eligible come out healthier? In emergency conditions, mortality (that is, what percentage of patients die at a given interval from their admission?) is a good measure of how well the treatment works. Take a look at this graph, which is the most important result from the paper.
Patients who are admitted when they are eligible for Medicare die from these emergency conditions less often. Given that so many of those patients actually do sign up for Medicare, that’s an incredible sign that generous insurance like Medicare can, in fact, improve health at its most basic level — preventing death. Some number-crunching by the authors reveals the following results: Medicare eligibility is associated with a 14-20% reduction in 7-day mortality, a 7-9% reduction in 28-day mortality, and a 2-4% reduction in 1-year mortality relative to the 64-year-olds who provide a control sample. As they write, “The emergence of the effect within 7 days of admission suggests that the extra service or changes in the quality of services provided to Medicare-eligible patients have an immediate life-saving impact.”
There’s another important question: was this result due mostly to previously uninsured patients receiving insurance, or did Medicare provide advantages over private insurance patients previously had? The authors investigate that issue as well and decide that the result was too big to only come from those who were previously uninsured; that is, moving to more generous insurance can actually have an effect on patients’ health as well.
Of course, this can’t be immediately extrapolated to all insurance. For one, this analysis looks only at true emergencies which, while important, are only a part of the overall healthcare picture. Second, the effects may be more pronounced in the elderly, who are generally more difficult medical cases than the young, and third, this study used data from California only.
Yet Card, Doebkin, and Maestas have provided evidence that Medicare does indeed save lives in emergency situations, which is an important finding on its own. By one measure at least, health insurance does work to improve health — but as we’ll see next week, that debate is far from settled.
Next week: the most famous experiment in health policy.
Allan Joseph is a first year medical student at the Warren Alpert Medical School of Brown University, where he is pursuing an MD/MPP. You can follow him on Twitter @allanmjoseph.