Well-Being and Medicare Spending

Okay, just let me say up-front that I think the way this study was done and the conclusions it purports to reach are basically garbage.  In this study, published in the JAMA Network, the authors were supposedly attempting to assess whether a broad notion of  a population’s well-being was associated with lower health spending.   (JAMA Article)   They used the Medicare population to measure health spending.  There measure of well-being is completely vague, some undefined notion of emotional, social and economic health, in addition to physical health.  For some reason using data going way back to 2010, the used the county-level Gallup Sharecare Well-Being Index and annual Medicare FFS spending per beneficiary.  You can probably spot some of the problems already.  Let’s see, let’s take a well-being measure across a whole population, but only look at spending in Medicare and then only FFS Medicare.  Oh and did we mention that the well-being measure is basically a self-reported survey.

The questions within the well-being survey are in the categories of life satisfaction and optimism; emotional health; physical health, specifically chronic illness or a recent medical problem; healthy behaviors; work environment (how many Medicare recipients work?)  and basic access to needs like housing, safety and medical care itself.  Is it just me or all of these just basically a proxy for income and wealth level?  The researchers also looked at certain county characteristics as another potential reflection of well-being.  Counties with as few as 4 respondents to the survey were used in the analysis.  Hmmm.  Counties in the highest quintile of fully-adjusted well-being scores cost Medicare a whopping average of $992 less per year per beneficiary than did counties in the lowest quintile.  Income and health system capacity had the greatest association with lower spending.  Income level has two health care effects.  One is that it tends to be associated with life circumstances and personal characteristics, like education, that lead to people being healthier and spending less.  The second is that people with more money can and do tend to spend more on the same health needs.  The first effect obviously predominates.  What is actually being measured in this study is just the well-known association between better life circumstances, largely related to financial situation, and better health.  The authors try hard to portray the results as reflecting some amorphous state of well-being, but I don’t see that at all.