Conference Report: Society for Medical Decision Making (SMDM)

Karen Smith and Prashant Rajan

Hey everyone! Welcome to OrACORe and PIVOT’s first official blog post! We wanted to take this opportunity to provide an update on a recent trip that we took to the Society for Medical Decision Making (SMDM) conference in Vancouver. The SMDM meeting brings together an interdisciplinary group of researchers who study how medical decisions can improve individual and population health. The theme of this year’s meeting was “From Uncertainty to Action.” SMDM has three main areas of activity:

• Cost-effectiveness analyses
• Preferences, shared decision making, and decision aids
• Diagnostics and clinical decision making

The conference organizes research presentations, instructive courses, and even professional networking sessions around these areas of activity. It starts off with a day of short courses designed to offer instructive information on methodology. One of this year’s courses, Advanced Topics in Decision Analytic Modeling, provided cutting-edge information in modeling techniques. The course emphasized the implementation of continuous-state Markov models (as an alternative to the conventional discrete-state models) to more accurately depict disease states. In addition, the course instructors provided detailed instructions on how to vary cycle lengths accurately by adjusting transition probabilities and discount rates.

One of the talks we were most excited about was “Modeling value-based payment adjustments for timely acute stroke care: cost-effectiveness analysis modeling meets pay-for-performance”. The project, presented by Ankur Pandya from the Harvard T.H. Chan School of Public Health, aimed to use cost-effectiveness analysis to determine the optimal payment values for pay-for-performance incentives. Value based payments are becoming popular, but haven’t been linked to healthcare value as estimated by cost-effectiveness analysis; the talk framed a new way of thinking about translating cost-effectiveness analyses to actual policy changes.

Another conference highlight was the keynote speech by John Ioannidis. Probably most famous for his 2005 paper, “Why Most Published Research Findings Are False”, Dr. Ioannidis spoke to SMDM about how to make medical research credible and useful. Among other things, he pointed out that the majority of medical studies are either false positives or overly exaggerated effects (why try to replicate if it might kill your major discovery?). Fortunately, there are ways to improve research, for example, changing incentive structures to reward replication experiments and reproducible discoveries.

Overall, the SMDM meeting was incredibly useful, adding depth and breadth to our training in decision analysis and cost-effectiveness at OrACORe/PIVOT. Through poster presentations and plenary sessions, SMDM exposed us to the many different avenues for work in the field of medical decision making. But by also providing structures for more detailed instruction in modeling, SMDM provided us with a greater methodological understanding that we can carry forward in our own training and future careers. See you next year in Pittsburgh!