Home » Departments, Meetings

A Conversation with Anirban Basu

1 October 2019 No Comment
The International Conference on Health Policy Statistics (ICHPS) has played a vital role in the dissemination of statistical methods in health policy and health services research throughout the past 20 years. In preparation for the next conference, which will take place in January 2020, we’re running a series of interviews and articles about previous Health Policy Statistics Section award winners.

James O’Malley is a professor of biostatistics in the department of biomedical data science and Dartmouth Institute for Health Policy and Clinical Practice at Dartmouth.

Aasthaa Bansal is an associate professor at the Comparative Health Outcomes, Policy, and Economics (CHOICE) Institute at the University of Washington.

Key Publications by/About Anirban Basu
Basu, A., and P. Rathouz. 2005. Estimating marginal and incremental effects on health outcomes using flexible link and variance function models. Biostatistics 6, 93–109.

Basu, A., J. Heckman, S. Navarro-Lozano, and S. Urzua. 2007. Use of instrumental variables in the presence of heterogeneity and self-selection: An application to treatments of breast cancer patients. York Health Econometrics and Data Group (HEDG) Working Paper 07/07. (with Editorial) Health Economics 16, 1133–1157.

Terza, J., A. Basu, and P. Rathouz. 2008. Two-stage residual inclusion estimation: Addressing endogeneity in health econometric modeling. Journal of Health Economics 27, 531–543.

Basu, A. 2015. Welfare implications of learning through solicitation versus diversification in health care. NBER Working paper # 20367. Journal of Health Economics 42, 165–173.

O’Malley, A.J., and A. Bansal. 2018. A conversation including 39 questions with Anirban Basu. Health Services and Outcomes Research Methodology 18(4), 287–297.

At the 2018 International Conference on Health Policy Statistics (ICHPS), held in Charleston, South Carolina, Anirban Basu was awarded the Mid-Career Excellence Award from the ASA Section on Health Policy Statistics (HPSS). After receiving the award, Anirban was interviewed about his life, including early influences and his opinions related to hot topics in statistics. A summary of this interview by James O’Malley and Aasthaa Bansal follows.

The Mid-Career Excellence Award is for those who are within 15 years of their most significant degree. Having earned his PhD in 2004, Anirban met the eligibility criteria for the 2018 award.

Anirban transitioned from a PhD student in 2004 to full professor in 2014, subsequently receiving a named and endowed full professorship in 2015. He is currently the Stergachis Family Endowed Professor at the University of Washington and director of the Comparative Health Outcomes, Policy, and Economics (CHOICE) Institute.
Anirban’s rapid rise was accompanied by a large number of awards, including student awards from the ASA at JSM 2003 and HPSS at ICHPS 2003, a Society for Medical Decision Making (MDM) comparative effectiveness award in 2009, three awards from the International Society for Pharmacoeconomics and Outcomes Research (2007, 2009, 2016), and ASA Fellow in 2016. He has also received lectureships and other honors and published close to 100 publications in journals across multiple disciplines.

Anirban’s service work is meritorious, especially for one so young. He chaired the 2010 International Conference on Health Policy Statistics (ICHPS), was HPSS program chair for the Joint Statistical Meetings in 2008, and was a member of the planning or advisory committee for four other ICHPS meetings. In addition, Anirban has chaired or co-chaired conferences of importance to HPSS, including the MDM Annual Meeting in 2011, after co-chairing its Scientific Committee in 2009 and 2010. He also founded the Annual Health Econometrics Workshop (AHEW) in 2009 and has chaired or otherwise overseen each subsequent edition.

Anirban is also prodigious in his receipt of grants and generous in his commitment to grant review and editorial work. He is loved by his students and those he has mentored. And he has a reputation for being a champion for junior investigators, having helped mentor them through the arduous process of obtaining grants. At the time of submitting his nomination for the mid-career excellence award, he had advised or mentored 22 students or postdoctoral fellows.

Timeline and Highlights

Anirban was born in a small town in the state of West Bengal in India. He became interested in statistics during his master’s program in pharmaceutical sciences and changed fields to pursue a master’s in biostatistics, during which he discussed statistics with P.K. Sen at The University of North Carolina. “Looking back, it was one of the best decisions I made in my career,” said Anirban. “I always had a curiosity for economics, partly because my father was an avid follower of financial news.” Combining his love of statistics with his passion for population health, Anirban earned a PhD in public policy at The University of Chicago.

Anirban remained at The University of Chicago, becoming a faculty member in 2004, until moving to the University of Washington in Seattle in 2011. In Anirban’s first year at Chicago, statistician Paul Rathouz encouraged him to apply for the HPSS student paper award for his work on estimating marginal and incremental effects on health outcomes. “I was one of the student winners that year and immediately hooked to the section and its members,” said Anirban.

A seminal paper he co-authored on the two-stage residual inclusion method seemingly popularized the control function IV approach in the social science and statistical literature, sparking a lot of comment and investigation by statisticians. His work in causal inference was also influenced by a collaboration with James Heckman. Anirban said of this experience, “Working with a Nobel Prize winner is always intimidating. I learned a ton from Jim and his works, really the underpinnings of causal inference, especially in the context of instrumental variables.”

Anirban brings to statistics approaches and reasoning developed in other fields. In his 2015 Journal of Health Economics paper, Anirban took a well-known theoretical model of behavior known as Roy’s model, which captures self-selection by individual agents based on perceived costs and benefits of alternatives, and applied it to study conditions under which patients would agree to enter a randomized controlled trial (RCT) when both the comparators are freely available to them under health insurance.

Asked about data science, Anirban had an interesting perspective. “I think the evolution of the field of data science, beyond statistics, is mostly driven by the type of data we see these days, which spurs innovation in statistical methods,” he said. “In economics, causal inference is the king, as many economic problems aim to forecast effects of policies and prices on behavior. So, data sciences, more specifically machine learning methods that focus on prediction and classification, do not immediately fit the requirements within the field. Health is one area where both pure factual prediction problems and problems of counterfactual predictions exist.” Anirban continued, “There has been a lot of uptake of data science for the first set of problems. But unlike traditional economics, there is limited scope for large-scale manipulation of data production to apply machine learning for counterfactual predictions. But that has not deterred some researchers from applying these methods erroneously to causal inference problems, without a clear identification rationale. More recently, interesting methods that combine deep learning algorithms with instrumental variable approaches are beginning to show promise. But then again, there are clinical researchers who are so tied to RCTs that they refuse to believe that randomization can exist outside of controlled environments.”

When asked about his future plans, Anirban said he will “continue to develop new methods and also figure out appropriate application of instrumental variable methods for observational data research.” He added, “Another big thrust of my work is to understand heterogeneity in effects and how behavior (patient, provider, policymaker) changes when facing evidence about heterogeneity.”

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)

Comments are closed.