Death Plan

Death on the Installment Plan

Now we know: Rejecting the Medicaid expansion could kill nearly 6,000 people each year.

By HAROLD POLLACK

Like many other liberal health-policy wonks, I’ve written a lot about the value of health reform in improving access to preventive care, protecting people against crippling medical debt and improving people’s physical and mental health.

I haven’t written much about how better access to health care can actually save lives. The argument for the Affordable Care Act, President Obama’s signature health-care-law, doesn’t ride on this. Moreover, the connection between health insurance and mortality is really hard to pin down, even if insurance truly has strong protective effects. The uninsured in America are mainly non-elderly adults. Deaths are really rare in this population, on the order of 0.4 percent per year. according to an Urban Institute study. Real-world randomized clinical trials—even those with thousands of patients—are just too small and too brief to reliably determine how much we might reduce mortality by extending coverage to the uninsured.

On Monday, though, a beautiful study was published in Annals of Internal Medicine that provides some of the best data we have connecting health coverage to saved lives. It’s changed my thinking, too. I’m more confident than I was last week that the ACA will save many thousands of lives every year.

Ironically, the study examined the impact of the bipartisan insurance expansion enacted in Massachusetts in 2006—a.k.a. “RomneyCare,” which provided the basic model for the ACA. Three of the best researchers in the business—Benjamin Sommers, Sharon Long and Katherine Baicker examined a decade’s worth of mortality data in Massachusetts counties, comparing trends to those found in carefully chosen comparison counties in other states. This wasn’t a randomized trial, but it was the next best thing, tracking the experiences of hundreds of thousands of people for years before and after the enactment of Massachusetts’ reforms.

Here’s their bottom-line result: Insurance coverage reduced mortality rates by about 30 percent. For every 830 people newly insured, Massachusetts prevented one death per year.

The sheer craftsmanship of this study makes it a pleasure to read (at least, if you’re a health wonk like me). It includes several smart checks to rule out potential biases. For example, Sommers, Long and Baicker show that mortality rates among elderly Massachusetts residents were basically unaffected by the 2006 reforms—which makes sense because almost everyone in this group was already insured through Medicare. The authors also demonstrate especially strong mortality reductions for conditions that are actually amenable to medical intervention, such as strokes.

Do these results generalize to the national expansion of coverage under the Affordable Care Act? Nobody really knows. Massachusetts has done a better and more enthusiastic job implementing RomneyCare than many states (and the federal government) have done thus far with ACA.

On the other hand, Massachusetts experienced the strongest survival benefits in low-income areas that contain many uninsured people. These counties look more like those in less-prosperous states most affected by health reform. Massachusetts began its reform as a prosperous liberal state with effective public health polices and a strong infrastructure of safety-net care. Other states are starting with a much less favorable baseline, and thus hold more dramatic possibilities for improvement. A state like Kentucky, which just provided coverage for the first time to hundreds of thousands of very poor people, might well see larger effects.

One thing is for sure. If anything close to these results apply, the ACA is saving many lives every year. The new law is projected to cover more than 20 million adults who would otherwise go uninsured. The Massachusetts estimates imply that the ACA will prevent something in the neighborhood of 24,096 deaths every year (simply: 20 million divided by 830). That’s more than twice the number of Americans killed in gun homicides. It’s considerably more than the number of Americans who die from HIV/AIDS.