Better Alignment of Incentives in Healthcare
Posted on July 2, 2020 by Alec Howard
As healthcare approaches 20% of GDP in the US, there has been increased focus on its economic structure. This structure, unique among economic sectors, is characterized by multiple inefficiencies – economic rents (in which producers can earn excess profits), uneven distribution, consistently and excessively rising costs, a large and confusing system of middlemen, and importantly, mis-aligned incentives that create adverse results in multiple and unexpected ways.
Two articles that emphasize – in different ways – the difficulty the healthcare market has in aligning incentives between payers, providers, and patients are: “Does Disneyland Provide a Guide to the Future of Personalized Medicine?” by Dr. Tomas Philipson (Council of Economic Advisors, on leave from the University of Chicago) and “To Save on Healthcare, Change What the Doctor Orders” by Dr. Peter Orszag (Head of North American M & A, Global Co-head of Healthcare, Lazard Freres & Co LLC, former Director of the Office of Management and Budget).
Each article highlights different aspects of the incentive alignment problem. Dr. Philipson discusses the disincentives created by our current payments structure to the development of personalized medicine, in which diagnostic tests are relatively cheap, but procedures, services, and drugs all require incremental expenditure and are increasingly costly. This creates a disincentive to innovators to develop better diagnostics, which in turn retards the development of personalized medicine. As a result, there is a mis-allocation of resources – either proper treatments are not pursued, or improper treatments are. In both cases there are substantial (and presumably to some extent avoidable) direct economic and human costs.
Dr. Orszag reports on a recent study in which economists determined that patients do not shop for MRIs, despite what appeared to be every incentive to do so. Factors which one might expect to foster comparison shopping: wide availability of the service, little, if any, quality difference from one provider to another, and significant variability in price, even between local providers, had little, if any, effect on patient decision-making. From this Dr. Orszag concludes that consumer choice is likely to be an ineffective driver of reduced healthcare costs, and that if healthcare costs are to be meaningfully affected, then doctors will need to change the way that they price both their own services and recommend ancillary services.
It’s interesting to explore these two premises in greater depth. In the first, Dr. Philipson uses the Disneyland model as a potential (and better) template for personalized medicine (by inference, all of healthcare). Admission to the park is relatively expensive, but once in, the rides are free or inexpensive. This creates an incentive to use the system optimally – each customer rides as much as they like. Since Disneyland is owned by one entity, it’s able to use a pricing model that optimizes user enjoyment of the entire park experience. Furthermore, there is a clear incentive to develop rides that create value or enjoyment for all users.
By contrast, personalized medicine is completely different – the economics of diagnostics (analogous to park admission) are completely divorced from those of drug development (the rides), and the pricing is the exact opposite of the amusement park– diagnostics are cheap, and drugs are expensive. Furthermore, since drug development and diagnostic development are undertaken by different economic entities, pricing for optimum value is impossible. Lastly, feelings of resentment and ill-usage are far more likely.
Dr. Philipson attributes the lag in the development of personalized medicine to this: given the poor reimbursement/payment regime for diagnostic tests, there is little incentive to develop new tests, which are the necessary complement to the new therapies. This is particularly so for cancer treatments. Many of the new therapies in development are matched to, and appropriate for, patients with a specific genetic profile – diagnostic tests are critical to determine which therapies will be, and conversely will not be, effective.
In the second article, Dr. Orszag discusses a study done by academics and doctors at leading universities. The study examined in detail the market for lower-limb MRIs and whether patients independently sought lower cost options. The study corrected for bias factors – for instance, MRIs associated with emergency situations were excluded. Overwhelmingly patients chose to follow the recommendation of their physician, rather than price shop.
This conclusion rings true – people are happy to comparison shop for cars, houses, appliances, etc. but are far less inclined to so for services in general and critical services in particular. They are much less likely to ignore the recommendations of their service provider – typically because these economic decisions usually involve correcting a problem rather than acquiring an item, and because a service relationship typically has a higher personal and trust level. If your auto mechanic tells you your car needs a brake job, or your plumber tells you your pipes need better insulation, the most typical response is to simply proceed. It is simply not reasonable to expect patients to buck this.
Dr. Orszag concludes by discussing methods by which physician behavior might be changed, including value-based medicine and better clinical support.
So, in both cases, poorly structured or misplaced incentives lead to suboptimal results. Why does this happen? In the case of personalized medicine, separate payment for diagnostics and drugs leads to an artificial distinction between the value-added pieces of a clearly linked process. In the case of MRIs and a lack of comparison shopping, doctors have no incentive to minimize the cost of a near-commoditized procedure.
How might this be improved? Both authors cite value-based medicine as a potential solution. While this is appealing, putting it into practice appears daunting – plus, it seems likely that its practice could lead to unintended (and potentially undesirable) consequences. For instance, if payment is predicated solely on whether a treatment works, might that discourage physicians from making treatment decisions that they otherwise would?
One alternative structure with promise is the direct primary care model, in which patients pre-pay (on a subscriber basis) a group of doctors or other healthcare providers for a specific range of services. Direct primary care follows the “Disneyland” model as Dr. Philipson recommends – once the subscription fee is paid, standard services are covered with no additional cost. Physicians have discretion over both clinical and economic decisions, and to Dr. Orszag’s point, their incentive is now focused on providing their services efficiently – there is no incentive to recommend a commodity service at a marked-up price. Lower administrative costs could also be expected to result from this structure, and it might also may engender genuine competition. And, if it worked as expected, perhaps analogous structures could be developed for more specialized procedures and conditions. Better economic structures – even on a small scale – could provide both a template and an incentive to larger scale reform. Continuing work on the economic structures of healthcare is critical to achieving better economic and clinical results.
Get Started With David
Get Your own complimentary* Discovery Analysis™ from B2B CFO®
*Complimentary for qualified business owners