PhD student: Tomas Maiuri
Title: Causal evaluation of digital medical devices under temporal confusion: Targeted learning, incremental interventions, and direct interventional effects
Supervisors: François Petit, Antoine Chambaz
Doctoral school: ED 393 Pierre Louis of Public Health
Promotion: 2025
Funding: ANR
Thesis abstract
Digital medical devices (DMDs), such as smartphone apps, have become an important part of monitoring and managing chronic diseases. Due in particular to the widespread use of digital medical data, they are expected to see significant growth in the near future. To support this trend, it is essential to conduct in-depth research into the complex issues raised by the development and evaluation, based on observational data, of time-dependent treatment rules. Mathematically, these can be modeled as dynamic treatment regimes, and their effects can be analyzed using treatment rule evaluation tools. Nevertheless, the context in which DMDs are used requires further theoretical study: a single DMD often combines several features, patients frequently use several applications with overlapping functions, and a DMD may be co-prescribed with a pharmacological treatment aimed at the same goal. This thesis project focuses on the development of statistical methods for evaluating DMDs based on observational data.
