I am a medical doctor and statistician, and an Associate Professor of Statistics at EPFL. I hold the Chair of Biostatistics since 2020 and I serve as Director of the Doctoral School in Mathematics since 2025. I work on causal inference and statistical methodology. My aim is to develop methods for longitudinal data, including time to event outcomes, that target scientifically meaningful questions under assumptions that are transparent and testable.
Much of my research clarifies the causal interpretation of classical survival analyses and advances the theory of separable effects, a class of causal parameters that disentangle direct and indirect effects of treatments. While working on these problems, I also became interested in questions about mechanisms, often branded more broadly as mediation analysis, and in how to make methodological results of practical interest. This line of work has been recognized with the Arthur Linder Prize (2023), the Lambert Award (2023), and the Sverdrup Prize (2024). Some relevant articles are:
M J Stensrud, J G Young, V Didelez, J M Robins, M A Hernán. Separable effects for causal inference in the presence of competing events. Journal of the American Statistical Association, 2022
M J Stensrud, J M Robins, A Sarvet, E J Tchetgen Tchetgen, J G Young. Conditional separable effects. Journal of the American Statistical Association, 2023
L Wen, A Sarvet, M Stensrud. Causal effects of intervening variables in settings with unmeasured confounding. Journal of Machine Learning Research, 2024
M J Stensrud, M A Hernán. Why use methods that require proportional hazards. American Journal of Epidemiology, 2025
M J Stensrud, M A Hernán. Why test for proportional hazards. JAMA, 2020
Another line of research studies optimal decision making and precision medicine. I develop statistical theory for dynamic treatment regimes, including regimes that can outperform human experts and existing algorithms under explicit assumptions, and I work on formal definitions of harm and their implications for safe and value aligned decision systems. Some relevant articles are:
M J Stensrud, J D Laurendeau, A L Sarvet. Optimal regimes for algorithm assisted human decision making. Biometrika, 2024
J D Laurendeau, A L Sarvet, M J Stensrud. Improved bounds and inference on optimal regimes. Journal of the American Statistical Association, 2025
A L Sarvet, M J Stensrud. Perspectives on harm in personalized medicine. American Journal of Epidemiology, 2023
A L Sarvet, M J Stensrud. Rejoinder to Harm in personalized medicine — an alternative perspective. American Journal of Epidemiology, 2025
I also work on causal inference when conventional statistical assumptions fail, for example under interference in infectious disease and vaccine studies. My recent work includes methods to identify and estimate vaccine effects when exposure is unmeasured, to separate immunologic and behavioral effects, and to quantify vaccine waning. This research is part of my SNSF Starting Grant on infectious diseases and an SNSF Project Grant on causal inference under resource constraints. Some relevant articles are:
M Janvin, M J Stensrud. Quantification of vaccine waning as a challenge effect. Journal of the American Statistical Association, 2025
G Perényi, M Stensrud. Variant specific treatment effects with applications in vaccine studies. Biometrics, 2025
M J Stensrud, L Smith. Identification of vaccine effects when exposure status is unknown. Epidemiology, 2023
M J Stensrud, D Nevo, U Obolski. Distinguishing immunologic and behavioral effects of vaccination. Epidemiology, 2024
A fourth line of work studies practical problems in medicine and epidemiology, broadly speaking, including large scale applications. For example, I have been the senior author on large studies on the effect of treatments against breast cancer. I also actively engage in methodology for epidemiologists, providing theoretical arguments that justify different empirical strategies and developing new ones when needed. Some relevant publications are:
M Piccininni, M J Stensrud. Immune selection stability is a neglected property of the causal risk ratio. American Journal of Epidemiology, 2025
A N Sawant, M J Stensrud. Explaining the sharp decline in birth rates in Canada and the United States in 2020. American Journal of Epidemiology, 2025
M Piccininni, M J Stensrud. Using negative control populations to assess unmeasured confounding and direct effects. Epidemiology, 2024
E Dumas, F Jochum, F Coussy, A S Hamy, A Majdling, S Houzard, C Le Bihan Benjamin, F Reyal, P Gougis, M J Stensrud. Explaining the relationships between age, endocrine therapy persistence, and risk of recurrence in hormone receptor positive early breast cancer: a nationwide cohort study. Journal of Clinical Oncology, 2025
An updated list of my publications is found on Google scholar.