betaFeedback

WorksMathematics

Assessing Subgroup Heterogeneity in Carryover Effects: A Simulation Study and an Observational Alzheimer Disease Treatment-Transition Application

Abstract

Subgroup heterogeneity in carryover effects is not routinely evaluated in 2x2 crossover studies, and its inferential consequences remain insufficiently characterized. When the scientific question concerns subgroup-specific carryover, omission of the corresponding interaction causes the fitted model to target a collapsed population-average contrast rather than the subgroup difference of interest. We used simulation to evaluate the finite-sample bias, confidence-interval coverage, Type I error, and power of carryover-by-subgroup interaction inference under an independence-based analysis and REML-based MMRM analyses with CS and UN covariance specifications. Reduced mean models were included as population-average benchmarks. We also applied the models to observational Alzheimer disease treatment-transition sequences from NACC, where carryover was defined operationally as a model-based association with the preceding regimen rather than an identified pharmacologic persistence effect. Full models directly estimated subgroup-specific carryover heterogeneity, whereas reduced models estimated a collapsed carryover contrast that did not characterize this heterogeneity. Under the full mean model, point-estimate bias was small, whereas covariance misspecification affected empirical confidence-interval coverage and Type I error. Reliable inference on subgroup-specific carryover effects requires a mean model that explicitly represents the heterogeneity of interest and a covariance structure appropriate for the repeated-measures data. Failure to address either component can obscure subgroup differences or compromise statistical inference. These results provide practical guidance for evaluating heterogeneous carryover effects in crossover and treatment-transition studies.

Paper details

vlnc:W_xzfqbaz7
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
arXiv nonexclusive license
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…