SoeHtutAung/MCDA-Spatial-DHS-Rshiny-Nigeria

Multi-criteria decision analysis tool to support strengthening community health workforce using a blended approach, including data extractions, spatial optimization, and R Shiny dashboard. # MCDA-Spatial-DHS-Rshiny-Nigeria Multi-criteria decision analysis tool **(MCDA)**, using R Shiny app, to support strengthening community health workforce using a blended approach, including data extractions, spatial optimization, and R Shiny dashboard. This project used both ***public and private*** datasets. # 1. Step-wise approach used in the project At first step, information on disease burden, population distribution and driving distance from the capital city is used to support selection of states. At next step, the dashboard allows customized ward selection where user can select different set of parameters and apply different level of importance on each parameter to calculate vulnerability score of each ward. Through this process, number of community health workers (CHWs) needed could be estimated and priority list of wards for CHW expansion could be generated. In each ward, spatial optimization model is used to quantify exact number of CHWs needed and identify where to assign them. # 2. Creating ward-level datasets To enable this approach, various spatial datasets and survey datasets are used to generate parameters at granular (ward) level. ## 2.1 ETL Refer to *Chapter 3.4 Data sources and extraction* of `docs/MSc project report_20250311.pdf` for data sources and extraction methods. Raw data files are kept in `data/` folder, but they are not uploaded in this public repository. ## 2.2 Preparing dataset Refer to *Chapter 3.4 Data manipulation* of `docs/MSc project report_20250311.pdf` for data management steps. Scripts for ETL processed are kept in `scripts/` folder. From spatial and other datasets, ward-level final datasets are created and kept as shape file in `data/shp/NGA_wards_dashboard`. Other outputs from data processing are saved in `outputs/` folder. ### Travel time to nearest health facility Percentage of population (left) and population size (right) at wards, who are living beyond 1 hour distance from the nearest PHC facilities (GRID3 – Health fa …