Extending propensity score adjustment and marginal structural modelling

Using the large clinician owned and operated MSBase international registry of multiple sclerosis outcomes, this seminar will provide worked examples of how to apply and evaluate propensity score matching as a means of re-balancing comparison study arms with regards to the baseline distribution of prognostic end-point correlates and confounders.

This will include head-to-head treatment comparisons between competing disease modifying drugs with regards to their long term influence on relapse and disability progression. The seminar will also introduce marginal structural models for the longitudinal management of bias both at baseline and during the study period.