The health care wonk-o-sphere is abuzz with the results of a new study showing a significant mortality reduction associated with the increase in health insurance facilitated by changes in Massachusetts law between 2007-2010.

Over at The Incidental Economist, Adrianna McIntyre has some discussion of the study.  She concludes:

Benjamin Sommers, Sharon Long, and Katherine Baicker estimate that overall mortality in Massachusetts declined 2.9 percent relative to control counties between 2007 and 2010; mortality amenable to health care declined 4.5 percent. This translates to one death prevented for every 830 people who gain insurance, and the effects were larger in counties with low income and low pre-reform insurance rates—the counties we would expect to be most favorably impacted by reform.

. . .

The findings aren’t bulletproof, but they’re close.

Now, I have a vested interest in the debate over the extent to which health insurance improves population health because 100% of my work in public health ethics/law/policy begins with an emphasis on the need to distinguish between health and health care, and generally proceeds to assess the ethical implications of the evidence suggesting that access to health care services is only a minor determinant of health.  If in fact there is evidence suggesting that such access can make major contributions to population health and the compression of health inequities, then that affects the framework of my running argument on this score.

In fact, McIntytre, Austin Frakt, Ash Paul, and myself engaged in an interesting and spirited debate a few weeks ago on Twitter — an admittedly poor medium for argument — on exactly these points.

So, I’ve taken a close look at the study, and these are some preliminary observations:

(1) This is a really well-done study.  I’m no methodologist, but the design of the study and the attention to ways of controlling for confounders even in a non-RCT study design is really impressive.  The results are impressive, their magnitude is surprising to me, and it should be taken very seriously.

(2) Notwithstanding #1, the study seems to have a methodological quirk common to virtually every study attempting to suss out the health impact of access to health care services as compared to other factors: it aggregates a number of interventions under the banner of “health care services” that ought IMO to be stratified.  There is no question that different kinds of health interventions seem to have different effect sizes in contribution to distributions of health and disease in human populations.  Part of the argument I and others want to make is some of these kinds of interventions (we might term them “population health-based” or some kinds of “public health-based”) seem, based on good epidemiologic evidence to have a greater impact on population health than others.  Glen Mays, a leading scholar of public health systems, makes the point well in a current blog post:

This study examines the population-level health effects of a health reform strategy that included not only insurance coverage expansions but also improvements in preventive services delivery and public health protections. Other elements of Massachusetts’ health reform strategy, such as reducing out-of-pocket costs for clinical preventive services and enhancing public health programs and infrastructure, may have contributed to the drop in mortality. For example, other studies have found large reductions in smoking prevalence and tobacco-associated health events among Massachusetts Medicaid recipients after the state added a comprehensive tobacco cessation benefit to its Medicaid program in 2006 as part of its health reform strategy. These preventive strategies and the resulting drop in tobacco exposure could plausibly explain some of the observed drop in mortality.

(3) We have an immense amount of epidemiologic evidence suggesting that access to health care services is only a minor determinant of population health.  This includes robust historical demography like the McKeown Thesis, as well as high-quality and powerful contemporary epidemiologic data flowing from, e.g., The Whitehall Studies.  The evidence base is so good on this point, that to the best of my knowledge it is frequently taken as more-or-less of a given.  The 2008 Final Report of the WHO Commission on Social Determinants of Health both canvasses the evidence for this point and states repeatedly that Closing the Gap in a Generation will simply not be possible through the provision of health care services, and that broader action on the non-medical determinants of health is not only required but is in fact the priority.  We also have another high-quality study, the Oregon Medicaid experiments, that did not show any significant attenuation in mortality associated with the expansion of health insurance coverage in the state.

My point here nothing more than the pedestrian, irritating, but still important notion that we cannot let positive results of a current and impressive study crowd out a significant amount of evidence that the very least sits in tension with the notion that access to health insurance has a large effect relative to other possible social policies in determining population health.

(4) The most important take-away for me really follows from #1-3 above.  No informed observers really dispute that access to health care services saves lives.  Of course it does.  The critical question from a public health ethics perspective is always one of priority-setting.  Given a palette of policy options A-E, which of these merits the most attention and the most resources? (And of course the criteria for “merit” is a deep and difficult normative inquiry).  As I noted in a series of Tweets on the study:

So, the issue for me is not really whether access to health insurance reduces mortality.  We have quite a bit of evidence suggesting that this is in fact the case (and again, I do not think most informed observers really dispute this).  Even as I am surprised by the magnitude of the reduction found in the Massachusetts study, I still maintain that the urgent ethical question is how the extent of the health improvements attributable to the provision of health care services compares to that attributable to action on the non-medical social determinants of health.

As terrific as this new paper is, I see nothing in it that undermines the succinct conclusion delivered by Lantz, Lichtenstein, and Pollack in a paper warning against the medicalization of health policy:

 . . . lack of access to health care is not the fundamental cause of health vulnerability or social disparities in health.


P.S.  If it is not perfectly obvious, nothing in the above should remotely be taken as an argument that expanding access to health care services is undesirable either from ethical or empirical perspectives.  The question that interests me here is rather where expanding such access ought to be prioritized relative to action on the structural determinants of health.