MCDS Seminar Series: Latent-state models for precision medicine (VIDEO RECORDING AND SLIDES AVAILABLE)

Image for MCDS Seminar Series: Latent-state models for precision medicine (VIDEO RECORDING AND SLIDES AVAILABLE)

More Information

mcds@unimelb.edu.au

Our first virtual seminar in the 2020 series was held on Friday 29 May. A video recording and accompanying slides from the webinar can be viewed below.

Our centre was pleased to host Eric B. Laber, Goodnight Distinguished Professor of Statistics at NC State University and Faculty Scholar at Amazon. His research focuses on methods development for data-driven decision-making with applications to precision public health, defence, sports/e-sports, and inventory management. He is also passionate about STEM outreach in primary and secondary education. You can learn more about his research and outreach at: Laber-Labs.com

Seminar Title: Latent-state models for precision medicine

Webinar slides: Download

Abstract: Observational longitudinal studies are a common means to study treatment efficacy and safety in chronic mental illness.  In many such studies, treatment changes may be initiated by either the patient or by their clinician and can thus vary widely across patients in their timing, number, and type.

Indeed, as a motivational example, in the observational longitudinal pathway of the STEP-BD study of bipolar depression, no two patients have the same treatment history even after coarsening clinic visits to a weekly timescale.  Estimation of an optimal treatment regime using such data is challenging as one cannot naively pool together patients with the same treatment history, as is required by methods based on inverse probability weighting, nor is it possible to apply backwards induction over the decision points, as is done in Q-learning and its variants.  Thus, additional structure is needed to effectively pool information across patients and within a patient over time.

Current scientific theory for many chronic mental illnesses maintains that a patient's disease status can be conceptualized as transitioning among a small number of discrete states. This theory leads to the construction of a partially observable Markov decision process model of patient health trajectories wherein observed health outcomes are dictated by a patient's latent health state. This model is then used to estimate an optimal treatment regime under two common paradigms for quantifying long-term patient health. The latent-state modelling approach provides high-quality estimates of an optimal treatment strategy in settings where other approaches cannot be applied without ad hoc modifications.