Dynamic treatment regimes: a new statistical toolkit for economic evaluation?

Thursday 31 August 2017, 12.00PM to 1.00pm

Speaker(s): Dr Noemi Kreif, Centre for Health Economics, University of York

Abstract:  Clinicians typically make a series of treatment decisions over the course of a patient's condition.  Examples include the treatment of chronic diseases such as diabetes and hypertension, where decisions on initiating treatment, changing the dose or introducing a new medication are updated over time, informed by the patient’s evolving health status.  This talk aims to introduce a growing literature in biostatistics that formalises such realistic, adaptive treatment decisions, and refers to these decision rules as “dynamic treatment regimes” (DTRs).  Through three case studies from different areas of clinical decision making (oncology, nutrition and diabetes), I will provide a conceptual overview of DTRs.  I will focus on the settings when patient-level observational data is available, introduce the concept of time-dependent confounding, and demonstrate statistical approaches that can be used to estimate the causal effects of such treatment regimes.

The second half of the seminar will be an interactive discussion on the potential role of these methods in economic evaluation.  Questions discussed will include: Could decision models of treatment pathways be improved by longitudinal patient-level information, such as cancer registry data?  Are effectiveness parameters estimated from longitudinal observational data credible alternatives of those obtained from trials, adjusted for treatment switching?  Are the causal parameters introduced in the talk directly applicable to provide input parameters for decision models?

Location: Alcuin A Block A019/20

Who to contact

For more information on these seminars, contact:

Thomas Patton
Dina Jankovic

Economic evaluation seminar dates

  • TBA