Title: A Novel Method for Identifying a Parsimonious and Accurate Predictive Model for Multiple Clinical Outcomes
L. Grisell Diaz-Ramirez MS,a,b Sei J. Lee MD,a,b Alexander K. Smith MD,a,b Siqi Gan MS,a,b W. John Boscardin PhDa,b
a Division of Geriatrics, University of California, San Francisco 490 Illinois Street, Floor 08, Box 1265, San Francisco, CA 94143, United States b San Francisco Veterans Affairs (VA) Medical Center 4150 Clement Street, 181G, San Francisco, CA 94121, United States
File format: .csv
Description: HRS data with 39 predictors and 4 outcomes of 5,531 respondents
File format: .R
Description: R code to perform LASSO Selection based on Optimal λ at Minimum BIC for Individual Outcome and Union methods
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for LASSO Individual and Union methods using 500 bootstrap samples
File format: .R
Description: R code for Fig. 4: Selection with the baBIC method and Individual Outcome Methods in the Case-study Data
File format: .R
Description: R code for Fig. 7: Comparison of Mean Number of Predictors Across Case-study Bootstrap Data and Simulations with Case-study Censoring and 25% Censoring
File format: .R
Description: R code for Fig. 8: Comparison of Mean Harrell’s C-statistic Across Case-study Bootstrap Data and Simulations with Case-study Censoring and 25% Censoring
File format: .sas
Description: SAS code to perform BIC backward elimination by Outcome using HRS original data set. It uses Cox regression for Death, and Competing-risk regression for rest of outcomes.
File format: .sas
Description: SAS code to perform BIC backward elimination by Outcome using HRS original data set. It uses Cox regression for 4 outcomes and Wolbers et. al (2009) adaptation to the Competing-risk settings.
File format: .sas
Description: SAS code to perform best average BIC (baBIC) backward elimination using HRS original data set. It uses Cox regression for Death, and Competing-risk regression for rest of outcomes. baBIC=absolute(BICk-BICbest)/absolute(BICfull-BICbest): BICfull and BICbest are the BICs of the full and best individual models.
File format: .sas
Description: SAS code to perform best average BIC (baBIC) backward elimination using HRS original data set. It uses Cox regression for 4 outcomes and Wolbers et. al (2009) adaptation to the Competing-risk settings. baBIC=absolute(BICk-BICbest)/absolute(BICfull-BICbest): BICfull and BICbest are the BICs of the full and best individual models.
File format: .sas
Description: Compute Beta coefficients used in Scenarios 1, 2, and 3 of the simulation study
File format: .R
Description: Produce cleaner format of Beta coefficients used in Scenarios 1, 2, and 3 of the simulation study
File format: .R
Description: R code to generate simulated data sets (training/test) for Scenarios 1, 2, and 3 with case-study censoring and 25% censoring
File format: .sas
Description: SAS code to import training simulated data sets under Scenario 1 (baBIC method) with case-study censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import test simulated data sets under Scenario 1 (baBIC method) with case-study censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import training simulated data sets under Scenario 1 (baBIC method) with 25% censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import test simulated data sets under Scenario 1 (baBIC method) with 25% censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import training simulated data sets under Scenario 2 (Individual Outcome method) with case-study censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import test simulated data sets under Scenario 2 (Individual Outcome method) with case-study censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import training simulated data sets under Scenario 2 (Individual Outcome method) with 25% censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import test simulated data sets under Scenario 2 (Individual Outcome method) with 25% censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import training simulated data sets under Scenario 3 (Full method) with case-study censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import test simulated data sets under Scenario 3 (Full method) with case-study censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import training simulated data sets under Scenario 3 (Full method) with 25% censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to import test simulated data sets under Scenario 3 (Full method) with 25% censoring from R and obtain some general statistics
File format: .sas
Description: SAS code to perform BIC backward elimination for individual outcomes for simulations of Scenario 1 with case-study censoring and training data.
File format: .sas
Description: SAS code to perform BIC backward elimination for individual outcomes for simulations of Scenario 1 with 25% censoring and training data.
File format: .sas
Description: SAS code to perform baBIC backward elimination for simulations of Scenario 1 with case-study censoring and training data
File format: .sas
Description: SAS code to perform baBIC backward elimination for simulations of Scenario 1 with 25% censoring and training data
File format: .sas
Description: SAS code to perform BIC backward elimination for individual outcomes for simulations of Scenario 2 with case-study censoring and training data.
File format: .sas
Description: SAS code to perform BIC backward elimination for individual outcomes for simulations of Scenario 2 with 25% censoring and training data.
File format: .sas
Description: SAS code to perform baBIC backward elimination for simulations of Scenario 2 with case-study censoring and training data
File format: .sas
Description: SAS code to perform baBIC backward elimination for simulations of Scenario 2 with 25% censoring and training data
File format: .sas
Description: SAS code to perform BIC backward elimination for individual outcomes for simulations of Scenario 3 with case-study censoring and training data.
File format: .sas
Description: SAS code to perform BIC backward elimination for individual outcomes for simulations of Scenario 3 with 25% censoring and training data.
File format: .sas
Description: SAS code to perform baBIC backward elimination for simulations of Scenario 3 with case-study censoring and training data
File format: .sas
Description: SAS code to perform baBIC backward elimination for simulations of Scenario 3 with 25% censoring and training data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for Individual, Union, Intersection, and Full methods in simulations of Scenario 1 with case-study censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for Individual, Union, Intersection, and Full methods in simulations of Scenario 1 with 25% censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for baBIC method in simulations of Scenario 1 with case-study censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for baBIC method in simulations of Scenario 1 with 25% censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for Individual, Union, Intersection, and Full methods in simulations of Scenario 2 with case-study censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for Individual, Union, Intersection, and Full methods in simulations of Scenario 2 with 25% censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for baBIC method in simulations of Scenario 2 with case-study censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for baBIC method in simulations of Scenario 2 with 25% censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for Individual, Union, Intersection, and Full methods in simulations of Scenario 3 with case-study censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for Individual, Union, Intersection, and Full methods in simulations of Scenario 3 with 25% censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for baBIC method in simulations of Scenario 3 with case-study censoring and testing data
File format: .sas
Description: SAS code to compute summary statistics for number of predictors and C-statistic for baBIC method in simulations of Scenario 3 with 25% censoring and testing data
File format: .sas
Description: Generation of 500 bootstrap samples from HRS case-study data
File format: .sas
Description: Perform BIC backward elimination for individual outcomes in 500 bootstrap samples
File format: .sas
Description: Compute summary statistics for number of predictors and C-statistic for Individual, Union, Intersection, and Full methods in 500 bootstrap samples
File format: .sas
Description: Perform baBIC backward elimination in 500 bootstrap samples
File format: .sas
Description: Compute summary statistics for number of predictors and C-statistic for baBIC method in 500 bootstrap samples
File format: .sas
Description: Compute Cumulative Incidence for simulations of Scenario 1 with case-study censoring and training data
File format: .R
Description: Figure of Cumulative Incidence for simulations of Scenario 1 with case-study censoring and training data
File format: .sas
Description: Compute Cumulative Incidence for simulations of Scenario 1 with 25% censoring and training data
File format: .R
Description: Figure of Cumulative Incidence for simulations of Scenario 1 with 25% censoring and training data
File format: .sas
Description: Compute Cumulative Incidence for simulations of Scenario 2 with case-study censoring and training data
File format: .R
Description: Figure of Cumulative Incidence for simulations of Scenario 2 with case-study censoring and training data
File format: .sas
Description: Compute Cumulative Incidence for simulations of Scenario 2 with 25% censoring and training data
File format: .R
Description: Figure of Cumulative Incidence for simulations of Scenario 2 with 25% censoring and training data
File format: .sas
Description: Compute Cumulative Incidence for simulations of Scenario 3 with case-study censoring and training data
File format: .R
Description: Figure of Cumulative Incidence for simulations of Scenario 3 with case-study censoring and training data
File format: .sas
Description: Compute Cumulative Incidence for simulations of Scenario 3 with 25% censoring and training data
File format: .R
Description: Figure of Cumulative Incidence for simulations of Scenario 3 with 25% censoring and training data