Analysis of stepped wedge cluster randomized trials in the presence of a time-varying treatment effect

TL;DR

This paper develops an analytical framework for analyzing stepped wedge cluster randomized trials with time-varying treatment effects, highlighting potential biases in traditional models and proposing new methods to accurately estimate treatment effects over time.

Contribution

It introduces a novel analytical approach to model and estimate time-varying treatment effects in stepped wedge trials, addressing limitations of existing models assuming immediate effects.

Findings

Traditional models can be severely biased when effects vary over time.

Some estimators may converge to the opposite sign of the true effect.

Proposed models can accurately estimate the entire effect curve.

Abstract

Stepped wedge cluster randomized controlled trials are typically analyzed using models that assume the full effect of the treatment is achieved instantaneously. We provide an analytical framework for scenarios in which the treatment effect varies as a function of exposure time (time since the start of treatment) and define the "effect curve" as the magnitude of the treatment effect on the linear predictor scale as a function of exposure time. The "time-averaged treatment effect", (TATE) and "long-term treatment effect" (LTE) are summaries of this curve. We analytically derive the expectation of the estimator resulting from a model that assumes an immediate treatment effect and show that it can be expressed as a weighted sum of the time-specific treatment effects corresponding to the observed exposure times. Surprisingly, although the weights sum to one, some of the weights can be negative. This implies that the estimator may be severely misleading and can even converge to a value of the opposite sign of the true TATE or LTE. We describe several models that can be used to simultaneously estimate the entire effect curve, the TATE, and the LTE, some of which make assumptions about the shape of the effect curve. We evaluate these models in a simulation study to examine the operating characteristics of the resulting estimators and apply them to two real datasets.