Using Pilot Data to Size Observational Studies for the Estimation of Dynamic Treatment Regimes

TL;DR

This paper introduces a method to determine the appropriate sample size for observational studies estimating dynamic treatment regimes, using pilot data to ensure sufficient power and accuracy in treatment effect estimation.

Contribution

It develops a novel sample size calculation procedure for observational studies of dynamic treatment regimes based on pilot data, addressing a gap in current methodologies.

Findings

The proposed method achieves high probability of accurately estimating the optimal regime's value.

Simulation studies demonstrate the method's effectiveness in various scenarios.

Application to electronic health records data successfully sized a study for depression treatment.

Abstract

There has been significant attention given to developing data-driven methods for tailoring patient care based on individual patient characteristics. Dynamic treatment regimes formalize this through a sequence of decision rules that map patient information to a suggested treatment. The data for estimating and evaluating treatment regimes are ideally gathered through the use of Sequential Multiple Assignment Randomized Trials (SMARTs) though longitudinal observational studies are commonly used due to the potentially prohibitive costs of conducting a SMART. These studies are typically sized for simple comparisons of fixed treatment sequences or, in the case of observational studies, a priori sample size calculations are often not performed. We develop sample size procedures for the estimation of dynamic treatment regimes from observational studies. Our approach uses pilot data to ensure a study will have sufficient power for comparing the value of the optimal regime, i.e. the expected outcome if all patients in the population were treated by following the optimal regime, with a known comparison mean. Our approach also ensures the value of the estimated optimal treatment regime is within an a priori set range of the value of the true optimal regime with a high probability. We examine the performance of the proposed procedure with a simulation study and use it to size a study for reducing depressive symptoms using data from electronic health records.