Skip to content

lbenz730/calendar_time

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Calendar Time-Varying Effects

Benz, L., Mukherjee, R., Wang, R., Arterburn, D., Fischer, H., Lee, C., Shortreed, S.M., Levis, A.W., and Haneuse, S. "A Statistical Framework for Understanding Causal Effects that Vary by Treatment Initiation Time in EHR-based Studies" Under Review (Pre-Print)

R Scripts (scripts/)

  • helpers.R: File of helper functions

Data (scripts/data)

  • build_weight_trials.R: Build target trials for TTE dataset w/ relative % weight change as the outcome
  • process_medications.R: Script to process some hypertension + antilipemic medications for use in the weight loss trials dataset.

Analysis (scripts/analysis)

  • EIF_helpers.R: Helper function for EIF for $\beta$ and $\chi_m$
  • EIF_weight_analysis.R: EIF applied to weight outcomes (estimation of $\widehat\chi_m$)
  • EIF_weight_projections.R: Projection step (estimation of $\widehat\beta$)
  • standardization_EIF_weight.R: Standardization analysis to compute $\widehat\chi_{j,m}$
  • hypothesis_testing.R: Run hypothesis testing for $\theta$ statistic
  • plot_EIF_weight_results.R: Plot results from data application (Figure 5, Table 2, Table S1)
  • plot_confounder_distributions.R: Look at patterns of covariates over time (Figures 1, S1, S2)

Simulations (scripts/simulations)

  • inform_sims.R: Get realistic range of coefficients for generating simulated data
  • generate_data.R: Functions to generate simulated data sets and compute true effects for simulation scenarios
  • specify_inputs.R: Specify and visualize structure of treatment effects across simulation scenarios (Figure 3)
  • compute_true_sigma_gamma_ratios.R: Script to compute and visualize the true values of $\frac{\sigma_m^2}{\sigma_m^2 + \gamma_m^2}$ for simulation data generating process.
  • run_simulation_pipeline.R: Wrapper to run simulation iterations
  • analyze_sim_results.R: Analysis of simulation results (Figure 4, S3)
  • latex_table.R: Generate summary table of sim parameters (Table S2)

Simulation Inputs (sim_inputs/)

.rds files containing the simulation parameters for each of the 18 simulation scenarios

Figures (figures/)

Figures saved out from various analyses

Shell (shell/)

.sh files for transferring files to and from the cluster.

Jobs (jobs/)

.sh files for batch jobs on the cluster

  • build_weight_trials.sh: Prep data for analyses w/ weight loss trials
  • EIF_weight.sh: Wrapper to run EIF weight analysis
  • EIF_transport.sh: Job array for EIF weight analysis transported to various populations
  • run_sims.sh: Wrapper to run job arrays for simulations
    • sims_n100.sh: $n = 100$ subjects
    • sims_n500.sh: $n = 500$ subjects
    • sims_n1000.sh: $n = 1,000$ subjects
    • sims_n5000.sh: $n = 5,000$ subjects
    • sims_n10000.sh: $n = 10,000$ subjects
    • sims_n25000.sh: $n = 25,000$ subjects

About

Causal inference method for estimation and inference of treatment effects which vary over treatment initiation time (calendar time) in electronic health record-based observational studies.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors