• Medicare Cuts and Mortality

    Reihan Salam has a post up commenting on the implications of an NBER working paper by Wu and Shen that finds that Medicare cuts (due to the BBA of 1997) resulted in increased hospital mortality. Austin has also previously written about this paper which finds that for each 1% reduction in hospital spending, mortality could be expected to increase by 0.4%. As Austin said:

    When cost cutting has consequences like this, maybe we shouldn’t rejoice at low* levels of cost shifting. Or maybe it’s just the payment rate changes we should think more about. By this paper, it’s not obvious those cuts led to reductions in “waste” only.

    This finding causes Reihan to reasonably worry about the potential effects of administrative cost cutting mechanisms in the Affordable Care Act. His highlighted quote of an Economics21 post on the Wu and Shen paper:

    The lessons of this study counter Administration rhetoric that spiraling Medicare costs can be blamed on overtreatment and easily be cut through centralized government control. Given the experience of the 1997 cuts, it is very unlikely that the IPAB, which relies exclusively on one-size-fits-all rate reduction, will produce solutions that do not sacrifice health care outcomes.[emphasis Reihan’s]

    Reihan goes on to offer his preferred alternative:

    This is why many of us are invested in the idea of a defined contribution model designed to facilitate bottom-up health system reform driven by decentralized business-model innovation.

    If $100 were scheduled to be spent on care, and then $95 were spent instead, would it matter how the size of the cut were identified?  All else equal, I would expect the potential effect on quality to be the same, regardless of how the cut was determined.  But, perhaps all else is not equal. That is certainly the claim of proponents of two approaches to cutting costs that are typically viewed as opposites: administrative rate setting and/or benefit exclusion, and market based insurance approaches.  To get more specific, consider two specific policy approaches that are designed to slow cost inflation: an expert-driven IPAB to help set Medicare payment policy; and defined contribution vouchering of Medicare with competitive bidding (as opposed to defined contribution with administrative pricing (like Ryan/Rivlin)).

    • The theory of the IPAB is that there is wasteful, inefficient and inappropriate care being provided, and that expert driven decisions can guide payment policy in ways that save money by getting rid of low quality care while encouraging appropriate, high quality care.  Note that the IPAB as passed in the ACA has some important limits to its administrative authority, and cannot ““ration health care,” raise costs to beneficiaries, restrict benefits, or modify eligibility criteria.”  The weak link of this approach is simultaneously putting in place the teeth necessary to obtain savings while maintaining the flexibility to match patient needs/preferences with appropriate, high quality care.
    • The theory of competitive bidding is that since patients pay the marginal premium beyond the defined contribution, they will be incentivized to be wise purchasers of health insurance (and therefore networks of providers and benefit design) and that plans will have to deliver high quality to get customers.  The weak link of this approach is providing the necessary information to consumers to inform their choices, and uncertainty regarding whether people will use such information in a way that leads to cost reductions and better outcomes.

    I am a supporter of the IPAB, and think it should be made stronger than what was passed in the Affordable Care Act. For example, a new treatment could be covered at a higher price than standard therapy, but if that therapy did not produce appreciably better outcomes over several years, then the newer treatment would still be covered, but reimbursed at the same rate as the standard therapy.  The IPAB should be expanded to make these sorts of consequential decisions, based on the best evidence.

    I have long been skeptical of competitive bidding for Medicare, but have become more open to this idea, in large part by reading Austin’s convincing work. Even though the IPAB and competitive pricing are viewed by many as ‘opposite approaches’ I think there are more similarities than differences in terms of the relative chance that either produces cost savings while improving quality.

    If we ever manage to develop a system in which we spend less per capita on health care than we would have by default, it will mean that patients receive less care than they otherwise would have and/or providers are paid less for that care than they otherwise would have been. Both approaches (IPAB and competitive bidding) provide a ‘safety valve’ of sorts, at least for persons with resources since individuals could use their own money to purchase services they desired but that were not covered, whether red-lined by administrative decision via the IPAB or through market forces via the health insurance plan a person chose. I can follow the logic for how both the IPAB and competitive bidding would produce a result of less money spent per capita and higher quality, but humility seems in order when predicting with certainty that either (or any) cost control approach will work out exactly as we imagine. That doesn’t mean we shouldn’t try.

    *Work by Wu shows that only 21% of hospital payment differences between Medicare and private insurance are ‘shifted’ to private payers. This result is lower (less cost shift) than is the intuition of many. I haven’t delved into that part of the story in this post.

    • Don
      “That is certainly the claim of proponents of two approaches to cutting costs that are typically viewed as opposites: administrative rate setting and/or benefit exclusion, and market based insurance approaches.”

      This is to imply that the IPAB’s role is primarily related to administrative rate setting. That is not clear, and given its shackles (and perhaps its future unlocking), what it can do past and present is a black box–to me at least.

      I think I am reading you correctly, or maybe not, but the quote conveys that these two interventions are diametrically opposed in approach. In fact, I see them as symbiotic, mainly, why can’t what we bid on be vetted through a value (read: IPAB 2.0) oriented process. They work together, not apart. This would leverage the strengths of both approaches.


    • @Brad
      I don’t think they are diametrically opposed. They are often put in such terms politically. In fact, IPAB or something like it could be key in providing ‘guidance’ ‘cover’ or whatever term you want to use for private insurance to deny coverage. No matter what the mechanism, the last step if we slow spending is less care and/or less pay for same care (on per capita basis). This will be hard technically, and even harder culturally.

    • Let’s say we cut medial spending in half and get to where Canada is. That means we go from 18% of GDP down to 9% of GDP and lets say that study is correct it might mean more hospital mortality but since we are 9% richer maybe life outcomes are better and we may actually be healthier. Maybe we fly more and drive less which we know saves lives. This needs to be considered and BTW the poorer a person is the more he should be willing to trade more health care for more other stuff.

      If we ever manage to develop a system in which we spend less per capita on health care than we would have by default, it will mean that patients receive less care than they otherwise would have and/or providers are paid less for that care than they otherwise would have been.

      I really think that we should try easing up on medical licensing and some other types of regulation that appear to make care less accessible and more expensive. I am not sure that it will work but it seems worth a try. BTW from what I understand Doctors in Cuba only go to school for 5 years.

    • @Floccina
      I think a more reasonable goal would be ‘lets say Medicare only goes up 2.5 times faster than inflation the next 20 years instead of 4 times inflation’. It is the rate of rise and not the level that is the biggest problem…plus, I cannot think of any nation other than former Soviet Union that decreased (literally spent less) per capita one year to the next. We need to slow the rate of growth….that is what spend less per capita than we would have be default.

    • I think the wise purchasers idea is fine in theory, but putting it into practice is where I have some concerns. This reminds me a bit of the arguments we have had in education policy when it comes to school choice- the parents who are the most educated and the most aggressive will make the better decisions for their children (or they should, assuming small sized information asymmetries and that greater parental education levels correlate with good decision-making).

      What would lead any of us to believe that the same thing would not play out in the health care market, where there are enormous information asymmetries (provider-patient, provider-insurer, insurer-insured)? Some sophisticated consumers of health insurance might fare well under such a program (competitive bidding), but there might also be a large swath of individuals who do a horrible job picking an insurance carrier. Then, they are locked into that plan until the next open enrollment. This produces problems not only for these individuals, but for the entire system-
      Do we really believe people simply will not consume health care even if their plan does not reimburse them? If they still consume, then we’ve put providers in the position of chasing down debts, as they do now. Where’s the improvement?
      Or, if higher out of pocket costs (due to lack of coverage/out of network/etc.) do discourage consumption, as they should, how certain can we be that all of the consumption foregone is unnecessary? Assuming some of the care foregone is necessary, how do we quantify the costs- to the individual, to the system, etc.?

      I would like a more efficient, and less costly, Medicare system as the next person, but I fear we are focused too much on a narrow definition of cost while ignoring the costs of unintended (but reasonably likely) consequences.

      (As I have noted to Austin before, I am also concerned that future Congresses will chip away at the definition/guidelines of minimum benefits as a way to control costs. That could lead us to a health care system of even greater inequities.)