Last year I made the case that health care technology is used inefficiently in the U.S., and more by me is forthcoming on this point. In this post I just want to point to two papers that have helped me think more deeply about this issue.
A very nice typology of health care technology that lends itself to thinking about its efficiency is found in a paper by Amitabh Chandra and Jonathan Skinner, about which I’ve written here. Another very good paper in this area is by Russell Korobkin, who explains why health insurance covers treatments, even if they offer minimal benefit for a high cost:
Consistent with this legal straitjacket, health insurers usually pay for any treatment recommended by a treating physician that offers the potential for any positive clinical benefit unless explicitly excluded from the contractual scope of coverage (Hall 2003: 655, 658, 671). When insurers do deny a physician’s treatment proposal and subsequently defend their position to external review boards, the issue is nearly always either whether the disputed treatment is at all effective for the patient’s condition or whether a requested procedure is cosmetic or lifestyle-related rather than medical in nature (Bloche 2011: 28; Hall 2003: 658). [...] Today it appears to be rare, if not completely nonexistent, for a private insurer to refuse to cover a physician-recommended treatment with recognized positive expected clinical value, unless specifically excluded by contract, on the ground that the treatment is not cost-justified. [Links added]
Korobkin goes on to propose a new way the health insurance market might be organized to permit plans to differentiate themselves according to the value of technology they cover and in a way that’s transparent to consumers. In time, Amitabh and I will say more about this idea, but Korobkin’s paper is worth reading in full, as is Amitabh’s with Jon Skinner. They cover issues and limitations to a level of detail about which we’re aware, but which Amitabh and I won’t have the space to address.
For example, one issue, among many, that Korobkin raises is that the value of health technology is heterogeneous.
As noted above, the benefits side of the equation will probably be based on the marginal QALYs that a particular treatment offers, but there is no value-neutral method of calculating QALYs (Neumann 2005: 31–34), and a heterogeneous society is unlikely to ever reach consensus on a single methodology. [...]
This problem can only be justified, if at all, by observing that the alternatives are arguably worse. In our current private insurance system, consumers must pay for deeper coverage than a great many would wish to purchase, with carve-outs that are specified ex ante in insurance policies that are undoubtedly inefficient for many customers, but few read or understand them before they find themselves in need of treatment. [...] Attempting to get the macro decision ‘‘right’’—that is, providing the architecture that enables consumers to make informed and stable choices about the allocation of their resources to medical care—requires simplifying information on the benefits side of the equation, even though it is clear this strategy will fail to account for heterogeneous preferences for different types of benefits. [Links added]
As is true of most issues in health care, there are no perfect solutions to the health care technology management problem. There are only those who delude themselves into thinking there are.