It was my honor to preside over the 2018 Society for Benefit-Cost Analysis (SBCA) Conference held March 14 – 16 at the George Washington University in Washington, DC. I am happy to report that by all indications it was another successful conference. We had 319 total registrants from 19 countries across 5 continents. U.S. participants represented 31 states and D.C. Half of all participants were affiliated with the U.S. federal government. The next largest group were academics (almost one-third); other attendees were from the private sector or from state and international government agencies. The three pre-conference professional development workshops were well attended and also drew a mix of participants from federal, state, and international agencies, as well as academics.

The conference evaluations we’ve received so far rank the overall quality of the 2018 SBCA conference at 4.4 on a scale of 1 to 5, with 5 being the highest. We are combing through the details and responses to the open-ended questions to see how we can do even better next year. We welcome more feedback, even when open-ended comments point out problems we can’t do much about, like the unseasonably cold weather in D.C. this past March!

I would like to thank again the organizing committee and, especially, the speakers for making the 2018 SBCA Conference a success. The 2018 program introduced a few new features. We had a set of Presidential Sessions that provided an overview of some of the best work in BCA. The Presidential Sessions included new work on perennial favorites such as “Mortality Risk Reductions: New Developments” and work in emerging areas such as “New Developments in Regulatory BCA.” Several Presidential Sessions also showcased the international range of BCA research: “Methods for BCA in Low- and Middle-Income Countries” and “An International Comparison of Social Discounting for Infrastructure and Regulation.”

On Thursday, Tomas Philipson, currently on leave from the University of Chicago to serve on the Council of Economic Advisors, gave our lunchtime keynote address (see my synopsis below). A plenary on Friday morning featured four former OIRA administrators reflecting on their experiences in this position between 1993 and 2017 (Daniel Hudson’s companion feature discusses this event in more detail).

Tomas Philipson is a leading health economist who has worked extensively on the economics of medical technology, as well as on population health economics. At SBCA, he spoke on “Valuing New Personalized Medicine Technologies.” Personalized medicine – for example, a genetic test to determine who will respond to chemotherapy for cancer – promises to improve health while reducing health care costs. While this might seem to make for an easy benefit-cost analysis, the design of public policies to maximize the net benefits from personalized medicine is not easy. Patients and providers must be given the right incentives to prescribe and use personalized medicine, and these incentives must be balanced against incentivizing the pharmaceutical industry’s research and development of new medicines. Philipson led the audience through a thought-provoking analysis that used economic theory to guide public policy to maximize the value of personalized medicine.

Almost paradoxically, Philipson began his analysis of the value of personalized medicine with a discussion of the problem of patient non-adherence. Patients often do not follow their physicians’ advice and many fail to start recommended treatments or stop their treatments prematurely. Philipson cited an estimate that waste due to non-adherence amounts to about 13% of total health care spending in the United States, or 2.3% of the GDP. Philipson put patient non-adherence in a new light. He argued that while physicians may be more informed about population-wide effectiveness and side effects of treatments, patients become more informed about their own individual value of a treatment. By experiencing the treatment, patients learn about its effectiveness and side effects.

Based on his analysis, Philipson concluded that the optimal rate of patient adherence is not 100%. In fact, over-adherence can sometimes be the more important problem. For example, suppose some patients’ cancers respond to chemotherapy, and other patients’ cancers don’t. The optimal choice for non-responders is to stop treatment early and become non-adherers.

At this point in his talk, Philipson brought his analysis back to the question of personalized medicine. A “companion diagnostic” provides an evaluation of treatment quality before treatment is started. For example, a genetic test might indicate, albeit imperfectly, whether the patient’s cancer is likely to respond to chemotherapy or not.

In the last and perhaps most provocative part of his talk, Philipson analyzed whether the current U.S. health care system “gets the incentives right” to maximize the value of personalized medicine. He used the analogy of the famous two-part pricing problem, where amusement parks charge a high entrance price but charge low or zero prices for the rides inside. Personalized medicine also involves two-part pricing, but the price of the companion diagnostic (the “entrance price”) is low while the price of the treatment (the “ride price”) is high. The incentives aren’t right: much of the value generated by personalized medicine accrues to the patients who do not take the treatment because of the results of the companion diagnostic test. The disconnect between the low price and the high value of the companion diagnostic test sends the wrong signals to innovators to bring that value to the market. One policy solution is to promote the vertical integration of the manufacturers of companion diagnostics and the relevant treatments. Philipson entertained questions from several audience members about this controversial implication.