ihmeuw/cost_effectiveness_hpv_vax_metareg

Cost Effectiveness Meta Regression of HPV Vaccination

This repository contains the code that performs the cost effectiveness meta-regression analysis of HPV vaccination programs. This analysis has been accepted for publication in PLOS ONE under the title, "Cost-effectiveness of HPV vaccination in 195 countries: A meta-regression analysis".

Purpose

The purpose for this repository is to make the analytic code for this project available as part of Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER) compliance.

Organization

All input, intermediate, and output files are defined in the file paths.py, and are replaced with "FILEPATH".

Scripts should be run in the following order:

1. create_paired_df.py

   Constructs a data set consisting of pairs of ratios from the same article and location that differ only in that they use different values of one covariate.

2. crosswalks.py

   Analyses of differences between ICERs for sensitivity-reference pairs of ratios. Uses functions defined in crosswalk_functions.py.

3. hpv_analysis.py

   Most of the analytic code as well as some code to generate plots, including those in Figure 2 of the publication. Uses functions defined in mr_functions.py and plotting_functions.py.

4. predictions.py

   Generates predictions and uncertainty intervals using model objects fitted in hpv_analysis.py.

5. logistic_regression.R

   Runs a logistic regression to predict the probability that our predicted ICERs are cost-saving and adjusts the predicted ICERs accordingly.

6. map_icers.R

   Creates the maps of the adjusted predicted ICERs and the upper bound of their uncertainty intervals in Figure 3 of the publication.

Inputs

Inputs required for the code to run are:

1. Valid paths to directories must be specified as ROOT_DIR, PLOT_DIR, and MODEL_RESULTS_DIR in paths.py.

2. A data file whose path is specified in paths.py as CLEANED_REG_DF.

3. A file specifying the values of all covariates for each prediction. Its path is specified in paths.py as CLEANED_PREDS_DF.

4. A shapefile in rds format that defines the boundaries of countries. This is used only in map_icers.R and is saved in a location specified as SHAPEFILE in paths.py.