Spatial Sampling Framework for Dental Facilities — Riyadh

Data and analysis code for the study “Validating Dental Sampling Frames in Rapidly Expanding Cities: Riyadh Case” (JDR Clinical & Translational Research).

This repository accompanies a field-validated census of private dental facilities in Riyadh, Saudi Arabia. It provides the calibration, validation, and quality-control datasets from the field census; a de-identified facility-level dataset; district-level allocation and street-sampling data; the city district basemap; and the R pipeline used to produce the study’s results, figures, and appendix tables.

---

Overview

Administrative licensing registries substantially undercount private dental facilities in rapidly expanding cities, and are not designed to serve as research sampling frames. This project builds and validates a spatially explicit sampling frame for private dental facilities in Riyadh through a citywide field census, a calibration/validation exercise, and a coverage/misclassification correction, and maps the resulting spatial distribution.

Outputs that reproduce from the data in this repository include the facility enumeration (732 private facilities providing dental services among 876 private health facilities surveyed), the calibration/validation/quality-control metrics, the spatial join to city districts, facility counts by district and region, and the kernel-density maps. See the Reproducibility section for the one class of results that requires restricted data.

• Manuscript: Alshihayb TS, Aldossri M. Validating Dental Sampling Frames in Rapidly Expanding Cities: Riyadh Case. JDR Clinical & Translational Research (under review).
• Setting: Riyadh, Saudi Arabia · Unit: private (and Ministry of Health) dental facilities · Basemap: Riyadh administrative districts.

---

Repository contents

File	What it is
Analysis_shared.R	The full R analysis pipeline (calibration → validation → quality control → spatial join → counts, maps, and regional summaries). Documented, portable, and runnable from the repository root.
sampling_streets_per_round.sas	SAS script that drew the stratified random street sample used to plan each field round (10% SRS stratified by district, fixed seed).
Facility_data_stripped.xlsx	De-identified facility-level dataset (one row per enumerated facility).
District_population_street_data_stripped.xlsx	District list, region assignment, team/round allocation, and the sampled/de-duplicated street lists. District population has been removed (see below).
Calibration_data.xlsx	Calibration counts per field-team group per district (inter-rater calibration).
Validation_data.xlsx	Independent validation subsample re-checked against Google Maps.
Quality_control_data.xlsx	Quality-control re-survey comparing facilities found by the QC team vs. the original collectors.
districts.geojson	Riyadh administrative district polygons (basemap for the spatial join and maps).
DATA_DICTIONARY.md	Variable-level documentation for every file and sheet.
CITATION.cff	Machine-readable citation metadata.
LICENSE / LICENSE-data.txt	Code (MIT) and data (CC BY 4.0) licenses.

---

Data

• Formats: open where practical — .geojson (RFC 7946) for geometry, .R/.sas for code. Tabular data are provided as .xlsx; column-level documentation is in DATA_DICTIONARY.md.
• De-identification: the facility file is stripped of directly identifying free-text where not needed for analysis; coordinates are retained because facility locations are the object of study and are publicly observable.

⚠️ Restricted data notice (district population)

Under the data-governance rules of the Saudi General Authority for Statistics (GASTAT), district/neighbourhood-level population counts cannot be shared publicly — only region/governorate totals may be published. District population is therefore removed from District_population_street_data_stripped.xlsx.

Consequently:

• ✅ Reproduce from the public data: facility enumeration, calibration, validation, quality control, the spatial join, facility counts by district and region, and the kernel-density maps.
• 🔒 Require the restricted district-population file (available from the authors on reasonable request): the population-standardised results — facilities per 10,000 residents, Figure 3, Appendix Figure 5 panel B, and the per-10,000 regional means.

When Analysis_shared.R is run on the public data it completes every step above and then stops with an explanatory message at the population-standardisation step.

---

Reproducibility

Requirements

• R ≥ 4.3 (developed on R 4.5.2).
• R packages: tidyverse, readxl, janitor, labelled, irr, flextable, gtsummary, officer, scales, sf, ggrepel, RColorBrewer, viridis, ragg, patchwork, ggnewscale, ggtext.

  install.packages(c("tidyverse","readxl","janitor","labelled","irr","flextable",
                     "gtsummary","officer","scales","sf","ggrepel","RColorBrewer",
                     "viridis","ragg","patchwork","ggnewscale","ggtext"))

Run

1. Download or clone the repository.
2. Open the folder in R/RStudio so the working directory is the repository root (the data files are read by relative name; no absolute path is hard-coded).
3. Run Analysis_shared.R top to bottom.

Notes

• The street sample was drawn with a fixed seed (set.seed(123) in R; seed=100 in SAS), so the sampling steps are deterministic.
• The single administrative region variable used throughout is new_region, the official Riyadh Municipality five-region classification (North, East, Center, West, South).

---

How to cite

If you use these data or code, please cite both the software/dataset and the article. Citation metadata is in CITATION.cff; GitHub renders a “Cite this repository” button from it.

Alshihayb TS, Aldossri M, Alasmary SS, et al. Spatial Sampling Framework for Dental Facilities — Riyadh [data and code]. Zenodo; 2026. https://doi.org/10.5281/zenodo.21658240

Alshihayb TS, Aldossri M. Validating Dental Sampling Frames in Rapidly Expanding Cities: Riyadh Case. JDR Clinical & Translational Research. 2026.

This repository is archived on Zenodo with a persistent DOI: 10.5281/zenodo.21658240 — the concept DOI, which always resolves to the latest version. The current release (v1.0.1) is 10.5281/zenodo.21658241.

---

License

• Code (.R, .sas): MIT.
• Data (.xlsx, .geojson): Creative Commons Attribution 4.0 International (CC BY 4.0).

You may reuse either, including commercially, provided you give appropriate credit.

---

Data sources & acknowledgements

• Facility locations: original field census by the study team.
• District basemap and boundaries: Riyadh administrative districts.
• Population (region-level, in the manuscript): Saudi General Authority for Statistics (GASTAT), 2022 Census. District-level population is governed by GASTAT and not shared here.
• Region-level resident income (in the manuscript): derived from resident-income platforms as described in the paper.

---

Contact

Talal S. Alshihayb — corresponding author. Please open a GitHub issue for questions about the code or data, or contact the corresponding author for access to restricted data.