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.

Share on twitter
Share on facebook
Share on linkedin
Share on reddit
Share on email

Hidden information below

Subscribe

* indicates required
Email Format