• Supply sensitive care

    “Supply sensitive care” is health care delivered at a volume that responds to availability of provider supply and in ways that cannot be explained by other factors like the health of the patient population. Of course all care is supply sensitive to some extent, but some things are more so than others, particularly when there is a lack of evidence or guidelines to inform best practice.

    This Dartmouth Atlas project topic brief (pdf) on the subject is a quick and easy read.  Here’s just one illustration from it:

    The blue dots represent how hospitalization (discharge) rates for non-surgical conditions vary with number of acute care beds across 306 regions of the US (hospital referral regions). The green dots are for hip fractures, for which hospitalization is indisputably warranted. For hip fractures,

    the hospitalization rate is similar across the country, with no correlation to the supply of hospital beds. By contrast, more than half of the variation in hospitalization rates for medical (non-surgical) conditions is associated with bed capacity.

    There’s a huge literature on regional variations in health spending, which I won’t review here. (See Bernstein, Reschovsky, and White and/or the CBO, both ungated PDFs.) It’s natural to ask to what extent any of it could be due to patient preferences, for instance cultural differences that drive variation in the demand for care and care intensity. That’s a difficult question to address.

    Amber Barnato and colleagues take it on in Are Regional Variations in End-of-Life Care Intensity Explained by Patient Preferences? (ungated PDF). They conducted a survey of Medicare beneficiary and elicited responses to questions about preferences pertaining to end-of-life care.

    Outcomes included concern about receiving too little medical treatment in the last year of life or receiving too much medical treatment, preference for dying in an acute care hospital, for life-prolonging drugs with side-effects, for palliative drugs with potential for life-shortening, and for mechanical ventilation.

    They did not find a relationship between patient preferences and end-of-life spending.

    We did not find a pattern of greater concern about receiving too little medical treatment, less concern about receiving too much medical treatment, preference for spending one’s last days in a hospital, for life-prolonging drugs despite side-effects, and for mechanical ventilation to achieve 1 week’s and 1 month’s life extension across respondents living in regions with progressively greater EOL spending. The observed relationship between respondents’ preferences for avoiding potentially life shortening palliative drugs and greater spending regions was explained by the confounding effect of race/ethnicity. Taken together, the lack of cross-sectional association between preferences and spending in our study is unsupportive of the hypothesis that differences in preferences explain regional variations in EOL spending. […]

    In summary, the results of this survey do not support the hypothesis that observed regional variations in EOL spending are attributable to differences in goals and preferences for care among residents of those regions.

    If the study findings can be extrapolated to care other than at the end of life, patient preferences have very little to do with geographic variation in health spending. If it’s not a demand phenomenon, then it must be a supply one. Care and care intensity seem to be driven in large part by the level of supply, though there are other factors. If you build it, offer it, and certainly if you promote or prescribe it, they will come and Medicare, as well as private insurers, will pay, whether it is good for health or not.

    Expect many more posts on these subjects in the near term (this week and next), and likely for months and years to come. In particular, next week I’ll explain the difference between supply sensitive care and supplier-induced demand.


    • Very interesting stuff. One counter-intuitive implication (of supplier-induced demand and potentially care as well) I could see is that supply restrictions (e.g., limited medical residency slots, long training periods, certificates of need for providers) do have some positive externalities on the system to counterbalance the arguments libertarian-leaning have against them.

      Now, of course payment reform conceivably could remedy this even in the context of removing supply restrictions but all else being equal, increasing suppliers (whether PCPs, specialists, NPs or PAs) is not an unalloyed good…

    • I’m just curious–why would the authors say that the correlation they observed between patient preferences and intensity of treatment was ‘explained away’ by race or ethnicity? In my clinical experience, it did seem like patients and family members from certain racial or ethnic groups were more likely to ask for life sustaining care at the end of life, maybe because those racial/ethnic groups tended to have a long history of neglect by the medical establishment. For those patients and their families, the recommendation to set aside life-sustaining treatment or forgo life-extending interventions seemed to raise suspicions (understandable, given the history involved) that I never saw hospitals do a great job addressing.

      To me, these patient preferences were very real and related to race/ethnicity in important ways, and not just statistical artifacts. I’m surprised the authors didn’t think to explore this more.

    • Thanks for this interesting post and graph. Again we see that the “demand” for use of medical services (in this case hospitalization) is driven by the medical industry (hospitals and doctors) and not by patients.
      I really don’t think we can blame any significant amount of “overuse” or “overcharging” on patients. The medical industry sets supply and creates demand and also sets prices (high). Patients just do what they are told by their doctors. Patients don’t get sick just to use the hospital but when they do get sick, the medical industry determines how much and at what price of their services will be used. Patients also don’t have any control over price. First, it is nearly impossible to find out the price of medical care before treatment. Second, there is no method to compare price and quality.
      Again we have seen…. Don’t blame medical costs on patients. Its the medical industry that drives costs.

    • -Is this analysis assuming that both the distribution of non-surgical conditions, their intensity, and the overall health status of patients is equal/random in all HRA’s?

      -It’s also worth asking the extent to which the demographic and socioeconomic characteristics of a given HRA have on the typical number of inpatient beds. When you look at the geographic distribution of “High quality, low cost” Medicare HRA’s they tend to concentrated in areas where population density, crime rates, and poverty are low and social networks outside of hospitals are quite robust – and hospitals tend to be smaller. Things like the home environment that a senior citizen with dizziness and mental confusion will be returning to may not matter in theory, but they matter in practice when it comes time to determine who will be admitted for observation and who will be sent home.

      -Given the weak and generally negative association between Medicare spending and private-payer spending trends in Medicare HRA’s that has already been demonstrated in the literature, I’d argue that folks who are partial to this narrative need strong empirical evidence that this association holds for private payers, and that the size of a hospital has no effect on the kinds of patients that seek care within its walls before concluding that more beds always and everywhere equals more unnecessary care.

      At the very least, this is an idea that should be vetted with extensive empirical testing before it’s rolled out nationwide..

      • I would have to dig into the studies, but I get the sense that much of the early work on supply sensitive care was not based on Medicare populations per se. I highly recommend Jack Wennberg’s Tracking Medicine if you have not read it. A deep review would include reading all his papers and all of those by his Dartmouth colleagues.

        • 1. That’s certainly a useful suggestion.

          2. I’d be very interested in learning which of their findings has withstood critical scrutiny.

          3. Rant….

          If anyone out there in the readership can tell me which of the conclusions based on the regional statistical associations that Darthmouth has documented haven’t been confounded a closer examination of real differences in the people that inhabit different geographic regions, or expanding the data-set beyond Medicaid, or measuring things like the percentage of patients with equally severe illness that survive in a given setting instead of comparing how much was spent on the patients that died, etc, etc, etc. That would be most helpful.

          Ditto for citing evidence from real clinical settings that have used regional practice variations to reduce costs and improve quality while treating the same patient cohort.

          How much of the statistical Monet still looks like something real under magnification?

          Are the conclusions of the study below *really* a surprise to anyone? Anyone else think this is the kind of data we should have before making dramatic policy changes based on high-level statistical associations?

          “Low-Quality, High-Cost Hospitals, Mainly In South, Care For Sharply Higher Shares Of Elderly Black, Hispanic, And Medicaid Patients”

          Does the following text from the study below not give anyone else pause?

          Despite widespread interest in understanding high-quality, low-cost providers, we are unaware of any previous studies that examined hospitals that were simultaneously on the extremes of both quality and cost. Therefore, in this study we sought to answer several questions: What are the structural characteristics of high-quality, low-cost US hospitals—the “best” hospitals? How do these best hospitals differ from other types of hospitals, especially high-quality, high-cost or low-quality, high-cost hospitals? Are underserved patients, such as minority and poor patients, more likely to receive care at one or more of these types of hospitals? And finally, how do clinical and patient-reported outcomes vary among these groups of hospitals? “

          • Can you be specific about what policy conclusions you’re concerned about? I have concerns too, and I’ve written about them on this blog. I just wonder if we’re on the same page or if you’re aware of those posts. Here’s just one: http://theincidentaleconomist.com/wordpress/health-reform-by-hatchet-or-scalpel/

            • -I believe (and I think that the data support this conclusion) that it costs more to take care of poor people, and their outcomes are worse even when they receive the same care – for complex reasons that doctors and hospitals can’t always (or often) do much of anything about.

              -My concern is that if we construct policies that reward providers and hospitals primaril on the basis of cost and outcomes without an adequate understanding of how who they are treating affects these metrics – then we could wind up shifting resources away from (punishing) those who treat the poorest, sickest populations despite having done nothing wrong, and rewarding those who treat the best-off patients despite having done nothing to merit such rewards.

              The fact that the Health Affairs paper is one of the first to examine the correlation between the socio-economic cohort treated in a given hospital and it’s cost/quality metrics *after* policy decisions based on cost/quality metrics which have clearly been confounded by these factors have been incorporated into law is…very unfortunate.

              I have other concerns about policies constructed on the intellectual scaffolding provided by data-sets confounded in ways that the people who have constructed them don’t seem to be aware of, but I’ll leave it at that for the purposes of this discussion.

    • Someone’s been reading “Tracking Medicine” recently, I take it.

    • Why is it that people who claim to have an understanding of economics look at a positive correlation between supply and utilization and DISMISS demand as a causative factor??????