• The Institute of Medicine’s geographic variation in health spending report

    It was mandated by the Affordable Care Act (ACA), was years in the making, and is now out. I have not read it (yet). But I did read Elliott Fisher and Jonathan Skinner’s summary. The key passage that explains the IOM’s findings is this:

    The report confirmed three core findings of Dartmouth’s research.  First, geographic variations in spending are substantial, pervasive and persistent over time — the variations are not just random noise.  Second, adjusting for individuals’ age, sex, income, race, and health status attenuates these variations, but there’s still plenty that remain.  Third, there is little or no correlation between spending and health care quality.  The report also effectively identifies the puzzling empirical patterns that don’t fit conveniently into the Dartmouth framework, such as a lack of association between spending in commercial insurance and Medicare populations.

    The committee also confirmed earlier work by Harvard investigators showing that, for the commercially insured population, variations in the prices paid by private health plans explain most of the variations in private insurance spending.  The committee deserves considerable credit for deepening our understanding of this irrational world of pricing commercial health care services.  Yet as the report finds, even in the commercially insured population, there are substantial differences in utilization rates across regions.  We would therefore argue that for commercial populations both price and utilization deserve attention, especially because in many regions, avoidable utilization may be easier to address than price.

    It is Medicare spending growth, however, that represents arguably the greatest risk to the financial health of the U.S Treasury, and in Medicare, variations are almost entirely the consequence of utilization of services, not prices.  The report finds that the single largest component of the variation in Medicare spending across regions that remains after risk and price adjustment is due to post-acute care (including skilled nursing facility services, home health care, hospice, inpatient rehabilitation and long term acute care). These services have also been a major source of growth.

    Click through for the rest from Fisher and Skinner, including a summary and commentary of the IOM committee’s recommendations, as well as a defense of a geographic lens.




    • Austin:
      I would be interested to learn in the Medicare and non-Medicare population what the various percentages of high utilizers to low utilizers is?
      I have seen figures that typically show that 10% of the insured population use about 80% of the benefits.
      The “trick” is that we do not know who will be in that 10% each year – it could be any one of us.
      I do think we can provide price breaks, over time, for those who are low utilizers and even those that are high utilizers.
      I suggest everyone pay fully community-rated premiums in the beginning, a ratio of 1 to 1, not 3 to 1.
      Each month, the low or no utilizers receive a discount in the following month’s premium.
      Over 36 months, the discounts could be as high as 60% off of the community-rated premium.
      Even the high utilizer is getting a break on his premium, too, in that it could, under ACA, be 3 times the lowest premium.
      This split, over time, between high and low utilizers could be done for a group of 4, or even a family of 4.
      As long as they continue to meet the10%-80% spread, the group or family pays much less in premiums, over time.
      Don Levit

    • Elsewhere in their essay, Fisher and Skinner recognize that increased spending on post-acute care is not (for the most part) driven by providers of post-acute care; it’s driven by the practices and actions of hospitals and doctors.

      For one example: hospitals and heart surgeons worked hard to shorten length-of-stay for coronary bypass patients, but to achieve that objective they needed cooperation from home health providers to hire nurses with higher skills and apply technology with higher costs. Hospitals still make bundles on surgery, but now home health agencies are providing more services to more patients that drive up overall Medicare spending.

      For another example: in the world of DRGs and public mortality statistics, hospitals wanted to reduce (a) length-of-stay for patients who are terminally ill and (b) frequency of those patients dying in acute-care settings. Hospices — some of them owned by hospitals and hospital-based systems — step up and help hospitals achieve those business objectives. Now non-profit hospices were mission-focused, working to change the way people die and to change the way people grieve, rather than to create an industry that’s hand-maiden to hospitals. But both things happened, and spending on hospice increased (as, apparently, did family satisfaction with end-of-life care provided to their loved ones). Hospice still accounts for less than 5% of Medicare spending, so if you’re going to follow the money you can’t lose sight of hospitals and doctors.

    • Austin
      Need to read more, but the findings seem to be at odds with Fisher and Skinner’s geographic focus in crafting solution.

      They cite Miami’s fraud as an example of why geography counts. However, if we take study findings to heart–assuming most cities have chicanery on one side of the street, and forthrightness on the other, we need a slightly difft approach.

      I know where they are coming from. Population approach matters. Of course.

      But they raise the issue mighty high. In house bias. My two cents.


      • It is hard for me to understand how, ultimately, geography (at some scale) isn’t a proxy. The only question in my mind is whether we know precisely what it’s a proxy for. For instance, some relatively fine (way sub-state, maybe sub-HRR) geographic divisions are a proxy for provider relationships. At some level you have (formal or informal) systems that vary in productivity, right?

    • I also found it interesting that they conclude that variation in Medicare and commercial spending is due to price markups and not utilization. Some research at Academy Health on geographic variation concluded that variation in spending was not correlated among Medicare beneficiaries and the commercially insured. However, these analyses were comparing prices and not utilization and thus, in my opinion, were flawed. The important question here is not whether or not some insurers/providers have greater market power in certain regions (although that is an important question just not for this topic). The important question is are some providers utilizing more services and is this warranted.

      • We know that some providers have far more market power than others, in part due to market concentration. Hospital mergers and acquisitions, both horizontal and vertical, have had the effect of increasing market share and thus their power to charge higher prices on the private insurance side. But just because a hospital can charge high prices to private insurers doesn’t mean it isn’t also doing its best to boost volume of services! Which is why I can’t quite understand why comparisons of privately insured versus Medicare beneficiaries doesn’t focus on utilization rather than spending.

    • Third, there is little or no correlation between spending and health care quality.

      Makes me think that very few of the medical interventions for the elderly are effective. I think that because if we cannot see a health benefit from the care that some are getting and others are not, we should not assume that the care that both groups get is effective. The care is not like vaccinations, trauma care and infant care that we know work they are mostly hard to measure.

      • It would be nice if medicare would play the bad guy and refuse to pay for most of those treatments but with old people voting in the numbers that they do and politics and politicians being whet they are that seems unlikely. What a mess.

    • I do not have hard evidence for what follows, aside from my experience consulting in the sector and listening in on MedPAC discussions. But here goes. Skinner and Fisher say that you should focus on the whole continuum of hospital and post-hospital care. The IOM report, they feel, focuses too much on post-acute care. Skinner and Fisher may be right, but post-acute care deserves a lot of scrutiny on its own.

      The most common modalities for PAC are home health care, where a therapist or a nurse comes to your home, or skilled nursing facility care, where you stay in a nursing facility. In SNFs, you probably get more intensive rehabilitation services (so do not confuse them with long-stay, “custodial” care for older adults w disabilities who can’t support themselves at home). Qualitatively, the services beneficiaries receive are similar but not identical (and in SNFs, you get quantitatively more services than in SNFs).

      There are also inpatient rehabilitation facilities, which are actually accredited as hospitals. They provide really, really intensive rehab. They’re increasingly specializing in Medicare patients with neurological diagnoses. They also do commercial work. They’re a small sector. They’re more expensive than SNFs. They offer qualitatively similar services to SNFs, as far as we can tell. Not all geographic markets have IRFs. I’ve heard MedPAC discussions where staff and commissioners think that similar patients in markets with no IRFs go to SNFs.

      Then there are long-term care hospitals (LTCHs). These are also accredited as hospitals. They provide hospital-level care for significantly longer than traditional hospitals (which I’ll call STCHs here). They serve a lot of end-of-life cases, where they perhaps unnecessarily prolong life. They provide qualitatively different care from SNFs and IRFs. They provide qualitatively similar care to STCHs, but not identical care. And they are paid a lot more than STCHs. Not all geographic markets have LTCHs. I’ve heard from MedPAC staff and commissioners in discussion that similar patients in markets with no LTCHs just stay in STCHs.

      It’s very much unclear what, precisely, the government is purchasing in post acute care. Potentially, a lot of patients could be treated in lower-intensity settings, but not all of them, and it’s hard to identify which ones. All the assessment instruments for each site are different, although CMS is making progress on a universal assessment – that way we could compare demographics, cognitive and functional status, other important variables across sites. Right now we can’t. We could perhaps craft firmer guidelines for admission criteria to each type of site, e.g. you can only be admitted to an IRF if you have a devastating neurological condition (e.g. massive car accident and brain damage, happened to one of my friends age 30-ish, and I suspect IRFs treat this type of patient as well as Medicare patients) that requires extended intensive rehab.

      I don’t want to sound anti-commerce here. However, there are a number of utter profit-maximizers in the space: a number of SNF chains are private equity-owned, there is one large publicly-traded IRF chain, I believe there are a number of for-profit LTCHs. These entities can behave in ways that distort the payment system (esp. more aggressive coding).

      Bottom line: the opacity of the post-acute care space and differences in the supply of certain types of providers enables large differences in utilization across different geographies.