Public repository of R code to reproduce analyses and figures in "Adherence to a priori-defined Diet Quality Indices Throughout the Early Disease Course is Associated with Survival in Head and Neck Cancer Survivors: An Application Involving Marginal Structural Models"
Program Descriptions:
diet_data_preproccess_score_calculations.R: Code for conversion of raw dietary intake data to servings per day, computation of food groups, and, finally, computation of index scores utilized in the analyses.
imputations.R: Code for imputing those with missing data. Imputed and non-imputed datasets are retained.
AC_mortality_analysis.R: Main analysis with marginal structural models for all-cause mortality.
CA_mortality_analysis.R: Main analysis with marginal structural models for cancer-specific mortality.
AC_interaction_effectMod_analysis.R: Effect-modification analysis (all-cause mortality).
CA_interaction_effectMod_analysis.R: Effect-modification analysis (cancer-specific mortality).
AC_sensitivity_analysis.R: Sensitivity analyses (all-cause mortality).
CA_sensitivity_analysis.R: Sensitivity analyses (cancer-specific mortality).
baseline_index__only_models.R: Analyses using time-invariant (with baseline fixed diet index scores) Cox Proportional Hazards Models (all-cause and cancer-specific mortality).
alluvial_plots.R: Code for generating alluvial plots in manuscript.
spline_plots.R: Code for conducting restricted cubic splines analyses and generating the plots in the manuscript.
table_1_epi_characteristics.R: Code for generating Table 1 in the manuscript.
quantfunction.R: Source code for function used in various analysis files (above).
surv_miner_bugfix_826: Bug-fix for survminer package (this is required and used in AC_mortality_analysis.R and CA_mortality_analysis.R to generate adjusted survival curves.
Contact Us:
Email: cam17@illinois.edu Email: aarthur4@kumc.edu