harshvh/tbtriage-cost-analysis

R code for the cost analysis of the TB TRIAGE+ community-based tuberculosis screening trial (CAD4TBv7 vs CAD4TBv7-CRP), South Africa and Lesotho. # TB TRIAGE+ comparative cost analysis R code for the prospective two-country comparative cost analysis of two community-based TB active case-finding screening approaches (CAD4TBv7 alone versus CAD4TBv7 combined with point-of-care CRP testing) in South Africa and Lesotho. Perspective: programme/provider. Currency: 2023 USD. Financial and economic costs reported separately; capital costs annualised at 3%. Probabilistic sensitivity analysis (10,000 simulations) with an assumed 20% coefficient of variation (WHO-CHOICE). ## Requirements - R (>= 4.1 recommended) - Packages: tidyverse, data.table, reshape2, gridExtra, ggrepel, treemap, plotly, crosstalk, scales, corrplot, patchwork Install packages once: ```r install.packages(c("tidyverse", "data.table", "reshape2", "gridExtra", "ggrepel", "treemap", "plotly", "crosstalk", "scales", "corrplot", "patchwork")) ``` ## How to run Open `triage-costing.Rproj` in RStudio first, so the working directory is the repository root. All paths in the scripts are relative to that root. The `output/` and `Figures/` directories are created automatically on first run. Run the whole pipeline: ```r source("run-all.R") ``` Or run scripts individually, in this order: | Order | Script | Produces | Depends on | |-------|--------|----------|------------| | 1 | `00-base-functions.R` | packages, constants, helper functions | (sourced by every other script) | | 2 | `01a-costs-southafrica.R` | `output/zaf_costs_*.RDS`, `output/cost_corr_sa_*`, `output/cost_breakdown_sa_*` | — | | 3 | `01b-costs-lesotho.R` | `output/ls_costs_*.RDS`, `output/cost_corr_ls_*`, `output/cost_breakdown_ls_*` | — | | 4 | `02a-aggregate-bycountry.R` | country-level tables and figures | 01a, 01b | | 5 | `02b-aggregate-overall.R` | overall tables, capital/recurrent split, per-person costs | 01a, 01b | | 6 | `03a-plots-components.R` | cost-component and cost-driver figures; `output/total_costs_*` | 01a, 01b | | 7 | `03b-plots-sensitivity.R` | combined component figure …