Still Think Health Care Costs Make Sense?

We know that health care costs are too high, and we know the rate at which they're increasing is actually accelerating.  But some people might be under the assumption that this process is uniform, or perhaps operating in some market-driven, logical manner.  A new toy from Robert Wood Johnson foundation - an interactive map of the U.S. showing variances in health care spending by hospital region, with information supplied by the Darmouth study - not only demonstrates the huge variances in spending throughout the U.S. but that the factors driving up cost are anything but logical.

All the information from the map is based on Medicare reimbursement per enrollee.  Now it's easy to imagine some factors coming into play that would influence costs even in a system like Medicare, where coverage for individuals is intended to be more or less uniform.  One is the natural geographic variances in cost of living that we all come to expect.  Major cities like New York, San Francisco, Boston, L.A. and San Diego are very expensive to live in.  Miami is about in the middle.  Corpus Christi, Houston and Lubbock, TX are among the cheapest cities to live in.  The second thought would be that states with the highest concentration of Medicare might influence the costs - either higher or lower.  So we'd expect Florida, California and New York - the three states with the highest numbers of Medicare beneficiaries - to look different from other states.

If you play along on the interactive map, you'll see that's not the case at all.  Start with the percentage annual increase in per capita spending, and you'll see Lubbock and Corpus Christi are dark brown, well over 4% annual growth on top of inflation, and Houston's not far behind at 3.81%.  Miami, too, has a very high annual growth at almost 5%.  But every expensive cost of living city has low growth - at the average 3% or below.  In San Diego's case, well, well, below.  So at the same time that everything else about living in San Francisco was on the rise compared to the rest of the nation, their health care costs weren't rising as quickly.  Switch to the green colors for the total reimbursements per enrollee, and it's a similar story.  Now Los Angeles, New York and Boston show up as high spending areas, but Lubbock and Corpus Christi have caught up to them, as has most of Louisiana.  In fact, go to the state view on that same map.  Florida and New York are dark green, as we'd expect if we're looking for concentrations of retirees, but Texas and Louisiana are nearly as much in terms of costs per capita.  California is average.  Keep in mind that the health care outcomes for Medicare recipients by state bear absolutely no relation to how much money that state is spending per person.

Let's put this another way - Monroe, LA has one of the highest per enrollee costs in Medicare in the nation at well over $10,000.  Does that make any sense?

Couple this with a similar report in the New England Journal of Medicine trying to figure out what causes these regional disparities.  Their first conclusion: how much these regions are spending has nothing to do with health, or even access to new technology (unless you want to explain to me how much of Louisiana has better access to the latest technology than Chicago).  Instead, they ran some studies involving hypothetical scenarios.  They found that doctors in high-paying areas were just as likely to suggest treatment where there was strong evidence that it was needed as the low-paying areas, but much more likely to use "discretionary services" - to respond to a grey area with treatment, even if the evidence wasn't conclusive.  The article gives the example of a patient who's 85 years old with end-stage congestive heart failure, when the muscles of the heart have stopped responding to treatment.  There's not much that can be done short of a heart transplant - which would be significantly dangerous at that age.  Doctors in high-paying areas were three times more likely to admit the patient into an intensive care unit, an expensive move despite the unlikelihood that this would make the patient any better, and 30% less likely to discuss palliative care with the family.

Why is this the case?  To be blunt, there's just not enough comparative effectiveness research to help guide decision-making when it comes to grey areas.  The 85 year old with heart failure example is extreme, but physicians must make these choices every day for ailments on a spectrum ranging from mild discomfort to mortal danger for their patients.  It's not just to treat or not to treat - it's choosing from a whole range of treatment options.  Without better clinical data not just on whether a program is effective or not, but whether it's just as effective or much less effective as a cheaper alternative is information that can help us come to terms with what the real cost drivers are.

Ignorance is not bliss in this case.  It's just a recipe for more skyrocketing costs that don't make us any healthier.