This repository contains the analysis code for CAR T-cell timing and patient outcomes project.
The project consists of five main scripts:
- Prepares baseline tables and descriptive summaries for the study cohort.
- Includes summary statistics for demographics, clinical variables, and treatment characteristics.
- Includes outcomes tables.
- Runs the main logistic regression models for primary and secondary binary outcomes.
- Includes model-based marginal effect estimates.
- Includes sensitivity analyses to test the robustness of findings.
- Performs KM analysis for overall and event-free survival.
- Calculates adjusted restricted mean survival time (RMST) at 2 years.
- Conducts subgroup analyses for logistic and survival outcomes.
- Stratifies by sex, cancer type, hospital, and other prespecified factors.
- Includes interaction terms and marginal effect plots for interpretation.
- Analyzes key laboratory values in relation to timing of CAR T-cell infusion.
- Compares distributions and trends across early and late infusion groups.
The analysis was developed using R (≥ 4.0) and the following key packages:
tidytablecollapselme4,lmerTestbroom.mixedsurvival,rmspseudobootgtsummary,gtggplot2