Cost-effectiveness analysis of interventions to improve diagnosis and preventive therapy for paediatric tuberculosis in 9 sub-Saharan African countries: A modelling study
# tippi
## License
This work is licensed under a Creative Commons Attribution 4.0 International License
creativecommons.org
## Structure & analysis order
The repo is structured as:
```
.
├── dataprep
│ ├── graphs
│ └── outdata
├── inference
│ ├── data
│ ├── graphs
│ └── outdata
└── model
├── data
├── graphs
├── indata
└── outdata
```
Analyses should be run in this order:
- *dataprep* : preparation of pre/post data for inference. Also contains some utility functions.
- *inference* : statistical analyses of pre/post data - can all be run and post-processed using the bash script.
- *model* : the main modelling analysis
- *modeldata.R* : prepares effect estimates from the inference analysis and makes parts of Table 1
- *modeloutcomes.R* : main results for each sensitivity scenario
- *runmodel.sh* : bash script to run all modelling analyses at once (on multicore machine with plenty memory)
*resultsupload.R* in the top level parses and uploads results to googlesheets for inclusion in the article. The uploads require authetication (authors only).
### Required packages ###
The following R packages are required:
- inference: rstanarm
- plotting: ggplot2, scales, ggpubr, ggthemes
- modelling: HEdtree, discly (install using `devtools::install_github('petedodd/packagename')` )
- data manipulation: data.table
- utilities: glue, here,
- data I/O: googlesheets4, readxl