• Come back to Texas

    In 2009, Atul Gawande’s New Yorker article on health care spending differences in two towns in Texas set the world on fire.  It was a perfect example of how huge spending differences could occur without any differences in outcomes.  To recap, he showed how there were cast differences in Medicare spending in McAllen and El Paso, Texas.

    Today, a new study was published by Health Affairs that tried to see if these same differences were seen in the under-65 year old population.  You have to remember that everyone over the age of 65 is covered by Medicare.  Most people under 65, however, are covered by private insurance.  Specifically, the authors wanted to see if the same differences were seen for patients covered by Blue Cross and Blue Shield of Texas.  BCBS covers about 10% of the under-65 population in both counties.  They then looked at the spending differences between Medicare and BCBS to see if the differences seen by Gawande held.

    Let’s start with the Medicare population (I’m turning their data into charts for easier consumption):

    Well, that’s pretty much what Gawande reported.  Medicare spending per person was way higher in McAllen than El Paso.  So far, no surprises.

    The data on BCBS, however, is not what you’d expect:

    Inpatient spending for McAllen was nowhere near as high, compared to El Paso, as it was for Medicare.  Outpatient care spending was less.  And overall spending was also slightly lower per person in McAllen than in El Paso.

    Here’s the authors’ conclusions:

    We have demonstrated that the sharp differences in the use of Medicare services between El Paso and McAllen, Texas, were not generally found in the population of Blue Cross and Blue Shield enrollees under age sixty-five in Hidalgo County (McAllen) and El Paso County. We considered several explanations for these patterns, including differences in prices, health, incomes, and other factors. Ultimately, we hypothesize that some part of the puzzle may be explained by private insurance companies and Medicare exhibiting very different interactions with local health care providers.

    I think there may be something to this.  I’ve always argued that the problem Gawande uncovered in Texas was a provider problem.  Doctors and hospitals are caring for patients in a more expensive way.  In fact, I’ve argued many times that the largest segment of health system over-spending in the US is for actual care.  So this should come as no surprise.  The authors believe that private insurers are more capable of making it more difficult for doctors to practice this more expensive type of medicine.  That may be the case.

    I don’t think they are as good as some will think, however.  Here’s one last chart I made from their manuscript data:

    This is BCBS spending per person broken out by age.  Yes, private insurance is better at holding down spending for young people, and middle age people.  But, once we hit the crowd 50 and above (really 50-65, when Medicare takes over), spending in McAllen takes off.

    What we may be seeing here is that pediatricians in the two areas aren’t that different.  Maybe GPs and Ob/Gyns caring for people in their next 25 years aren’t that different.  But – and this is critical – maybe those caring for the elderly, especially proceduralists and inpatient specialists, practice very differently.  That would account for the differences, and would affect Medicare more than private insurance.

    The authors take their results and say that maybe the same providers respond differently to private insurers than to Medicare.  I think it may be that the more expensive providers spend more time dealing with Medicare than private insurance.

    Regardless, providers are the problem, and we need to figure out how we can make them practice more cost-effectively.

    McAllen and El Paso Revisited: Medicare Variations Not Always reflected in the Under-Sixty-Five Population

    UPDATE: I got a very nice email from one of the study authors.  You can read his blogged thoughts on the study here.

    Share
    Comments closed
     
    • Nice analysis. I think this is a pretty important piece of data. It bears repeating elsewhere. Reforming Medicare will be very difficult, but I think there is low hanging fruit. This is one area, utilization, where we should be able to make progress. I also think that this suggests that we need to look for and mine data on private insurance. It is a huge part of our national budget but we know surprisingly little about it.

      Steve

    • The government has now put up a website where Medicare patients can report what they believe to be fraud, double-billing, coding errors and other complaints or information they may have about their Medicare bill, and federal employees are now empowered to follow up on those reports. This should go a long way toward addressing billing problems that were not addressed by Medicare in the past, especially when word gets around that this reporting mechanism is available and when enough computer-savvy individuals age into Medicare that there is no fear of the technology. The new health care reform bill has really done some good after all, despite its not being “enough” for some people.

    • Aaron
      Plausible theories, and it has been interesting to see how quickly this paper was picked up and covered on the blogs and MSM.

      I wonder though, if you began to parse individual <65 yo DRGs, a priori apply same reasoning, and view whether framework holds up, would the one major unifying explanation begin to unravel (like it often does).

      After examining these 10 or 20 DRG's and learning that half or so dont :fit the hypothesis, and we begin to come up with why MI's in McAllen vs El Paso are this, or GIB's are that, we would be back at square one. Essentially, have we tortured data so it is talking to us, but in the end, it's still hard to make heads or tails of.

      Like all variations in care, my initial zeal for the Dartmouth short list of explanatory bullet points has waned. This is stuff is so difficult. This quick commentary in JAMA should be a must read for anyone interested in this topic.
      http://jama.ama-assn.org/content/303/23/2405.extract

      Brad