Robust analyzes for longitudinal clinical trials with missing and non-normal continuous outcomes.

Area of Science:

• Biostatistics
• Clinical Trials
• Longitudinal Data Analysis

Background:

• Missing data is common in longitudinal clinical trials.
• Non-normal outcome distributions and outliers can bias traditional analysis methods.
• Existing methods like multiple imputation with mixed models may fail under these conditions.

Purpose of the Study:

• To develop a robust framework for handling missing data and non-normal outcomes in clinical trials.
• To improve the accuracy of average treatment effect (ATE) estimation.
• To provide a method that is robust to model misspecification.

Main Methods:

• Developed a robust framework based on Control-Based Imputation (CBI).
• Utilized sequential weighted robust regressions to address non-normality in covariates and response variables.
• Employed mean imputation and robust model analysis for ATE estimation.

Main Results:

• The proposed method provides consistent and asymptotically normal ATE estimators.
• The framework ensures robustness even when the analysis model is misspecified.
• Demonstrated superiority through simulations and an AIDS clinical trial application.

Conclusions:

• The robust CBI framework effectively handles missing data and non-normal outcomes in clinical trials.
• This approach offers improved accuracy and reliability for ATE estimation.
• The method is valuable for complex clinical trial data analysis.