Marginal structural models with Latent Class Growth Modeling of Treatment Trajectories

TL;DR

This paper introduces a novel method combining Latent Class Growth Modeling with Marginal Structural Models to better analyze treatment adherence patterns and control for time-dependent confounding in observational studies.

Contribution

It proposes integrating LCGM with MSM to classify treatment trajectories and accurately estimate effects, addressing limitations of standard methods.

Findings

Proposed method reduces bias in treatment effect estimation.

Simulation studies demonstrate improved accuracy over traditional models.

Effective control of time-dependent confounding in observational data.

Abstract

In a real-life setting, little is known regarding the effectiveness of statins for primary prevention among older adults, and analysis of observational data can add crucial information on the benefits of actual patterns of use. Latent class growth models (LCGM) are increasingly proposed as a solution to summarize the observed longitudinal treatment in a few distinct groups. When combined with standard approaches like Cox proportional hazards models, LCGM can fail to control time-dependent confounding bias because of time-varying covariates that have a double role of confounders and mediators. We propose to use LCGM to classify individuals into a few latent classes based on their medication adherence pattern, then choose a working marginal structural model (MSM) that relates the outcome to these groups. The parameter of interest is nonparametrically defined as the projection of the true MSM onto the chosen working model. The combination of LCGM with MSM is a convenient way to describe treatment adherence and can effectively control time-dependent confounding. Simulation studies were used to illustrate our approach and compare it with unadjusted, baseline covariates-adjusted, time-varying covariates adjusted and inverse probability of trajectory groups weighting adjusted models. We found that our proposed approach yielded estimators with little or no bias.