A shared parameter model for the estimation of longitudinal concomitant intervention effects.

Area of Science:

• Biostatistics
• Clinical Trials
• Longitudinal Data Analysis

Background:

• Concomitant interventions are often introduced during undesirable patient health trends.
• Standard statistical models may produce biased intervention effect estimates if the intervention timing is not incorporated.
• Accurate modeling is crucial for evaluating treatment efficacy in clinical settings.

Purpose of the Study:

• To propose a novel change-point approach for modeling concomitant interventions in longitudinal studies.
• To develop a robust method for estimating intervention effects and associated parameters.
• To evaluate the performance of the proposed method using real-world clinical trial data and simulations.

Main Methods:

• A shared parameter change-point model was developed to analyze pre- and postintervention trends.
• A likelihood-based estimation technique was employed for parameter estimation.
• The method was applied to a clinical trial dataset and validated through simulation studies.

Main Results:

• The proposed change-point model effectively evaluates pre- and postintervention time trends.
• The likelihood-based method provides reliable estimates of intervention effects.
• Demonstrated accuracy in a clinical trial for depression and heart disease, and simulation studies.

Conclusions:

• The developed change-point approach offers an accurate way to model and estimate intervention effects in longitudinal studies.
• This method mitigates bias introduced by unmodeled intervention timing.
• The approach is applicable to various longitudinal clinical trial settings.