State of US Health & The High Cost of FFS

A couple of weeks ago, I ran across a data set, titled County Health Rankings courtesy of the University of Wisconsin. The data helps to paint a rich mosaic of the state of health in the United States. Perhaps more, importantly the accompanying data on health factors allows us to understand what behaviors and social conditions lead to good (and bad) health outcomes, which can in turn guide decisions on both policy and investment.

Part I of exploring this data looks at how the state of health and healthcare costs vary across the United States. The variations are considerable, with residents in the healthiest states being 100% healthier than the lowest ranked states. And in a perverse irony, it is the residents of the unhealthy states that are spending the most on healthcare, serving to illustrate the vicious cycles created by ‘Pay for Service’ practices in healthcare vs. Accountable Care’.

[The data can be explored interactively using this R/Shiny ‘app’: US Health Rankings. ]

The ‘State’ of US Health

Health Outcomes by State

The map above shows the US states colored by a composite measure of health outcomes (length & quality of life). Simply put, the green areas indicate healthy states while the red indicates state’s whose residents badly need a doctor’s services.

A cursory examination of the map might suggest that cold weather is good for health! The healthiest states (green) seem to be in the Northeast and the northern part of the Midwest. And conversely the South seems to have the unhealthiest population. Indeed, the (warmer) ‘Deep South’ states like Alabama & Mississippi have among the worst health outcomes in the country.

Now what does it mean to live in a healthy state? A person in Minnesota is about 20%-25% healthier than the average American. An adult in Minnesota is ~25% less likely to be in bad health than the average American (Adults in Bad Health: 12% (Minnesota) vs. 16% (National) and report fewer days of bad health a month (Bad Health Days: ~3 (Minnesota) vs. ~4 (National). Minnesotans also seem to live longer, given than they are ~20% less likely to die before 75 (Deaths before 75 & Years Lost).

Minnesota Health Outcomes Scorecard

The advantage of living in a ‘healthy state’ is obviously even starker when compared to the ‘unhealthiest’ states. Take Minnesota relative to West Virginia, the unhealthiest state in the nation.

Health Outcomes: West Virginia v. Minnesota

A resident of West Virginia, is doubly (>100%) likely to be in bad health than a resident of Minnesota. And they are only half as likely to reach the age of 75. Even children are not immune, with child morality being 50% higher in West Virginia relative to Minnesota.

The phenomenon of bad health is true across all West Virginia. 54 (out of 55) counties are unhealthier than the national median. In contrast, most (>80%) of Minnesota’s counties are healthier than 80% of the USA.

Expensive Bad Health

The residents of ‘unhealthy’ states suffer from bad health, despite spending more on their health than their peers from ‘healthy’ states. The chart below charts how much people spend on healthcare (% of Income spent on Healthcare) vs. how healthy the state is (Health Outcomes Percentile). The downward sloping line shows a NEGATIVE correlation, i.e. that the states that spend the MOST on healthcare also have the WORST health outcomes.

Health Outcomes vs. % of Income on Healthcare

There is an ~80% negative correlation between spending and health outcomes. But of course correlation is not causation and it is not that spending more on healthcare causes worse health. However this serves to illustrate the inefficiencies created by ‘Pay for Service’ payment practices.

In ‘Pay for Service’, we pay for healthcare services (doctor visits, medications etc.) any time we use them, regardless of whether the services deliver good health. Since residents of West Virginia & Alabama are sick a lot, they visit the doctor a lot, which in turn leads to higher medical bills. But these residents are clearly getting a bad deal, because the higher spending is not delivering the good health it is paying for.

And this is very much a vicious cycle. The areas with bad health also tend to be areas with low income so higher healthcare costs that a significant percentage of income is not available for consumption or investment. Residents in the states with the worst health incur healthcare costs amounting to over 20% of their income. Whether this is paid for by the government or the individuals, this limits resources available for improved decisions on diet, education or living conditions. And limitations on these resources in turn retard improvements in health, creating that vicious cycle of high costs and bad health.

Accountable care payment models could help ameliorate this cycle. While a change in payment practices would not by themselves improve health, they would presumably reduce the costs (at least relative to results). This in turn would free up resources for pro-active investments in health factors like behaviors and living conditions, which would in turn improve health outcomes.

State Level Interventions Would Work

It seems indisputable that the state of health in the United States varies tremendously by geography. And it would stand to reason then that the needs of each state would also be different. Over a fifth (20%) of the population of the bottom ten states are reportedly in bad health, suggesting a need for urgent investment in health care initiatives. And in this it might be useful to study what the ‘healthy’ states do better.

It is interesting to note that health within a state is usually quite uniform, i.e. if a state is healthy, then all/most of the individual counties in the state tend to be healthy too, rather than just the large population centers.  The opposite is also true with unhealthy states having mostly unhealthy counties. Another way to put it is that within a state, there are few differences in health outcomes across the counties. The map below shows how health outcomes (shown by the color) tends to be fairly uniform within a given state. In fact on average, health in an individual county tends to be within 10% of the state’s median health.

Health Outcomes by County

This is in turn would indicate that state level conditions are a powerful determinant of its residents’ health. Therefore state level interventions could be an effective tool to improve health outcomes in the unhealthier states. However the range of interventions a state and its agencies might pursue are incredibly diverse. They could range from large-scale interventions like universal insurance coverage to smaller-scale behavior medication efforts like encouraging exercise by making gym memberships tax deductible.  Choosing the best (and cost-effective) interventions requires an understanding of what health factors have the largest impact on health outcomes, a topic that will be further explored in Part-II