sonath0427/sri-lankan-medical-institutional-data

Sri Lankan Medical Institutional Data (2024)

Dataset Description

This dataset accompanies the research paper:

Golden Hour Divide: Trauma Care Accessibility and Resource Vulnerability in Sri Lanka

The dataset consolidates district-level healthcare infrastructure, disease burden, demographic statistics, and hospital geospatial information collected from official Sri Lankan government publications for the year 2024. It was developed to support reproducible research on emergency healthcare accessibility, trauma systems, healthcare resource allocation, and spatial epidemiology.

The dataset forms the foundation of the analyses presented in the accompanying research paper, including accessibility modelling, Need-Gap Index (NGI) computation, H3 hexagonal spatial analysis, and district-level clustering.

Dataset Summary

The dataset includes information on:

• —Population distribution
• —Administrative land area
• —Healthcare institutions
• —Hospital bed capacity
• —ICU bed availability
• —Operating theatre availability
• —Specialist workforce
• —Disease prevalence
• —Geographic coordinates of ICU-equipped hospitals

The repository is intended for researchers working in:

• —Healthcare accessibility
• —Medical GIS
• —Spatial epidemiology
• —Public health
• —Disaster response
• —Emergency medicine
• —Health policy
• —Healthcare resource optimization

Data Sources

The dataset was compiled entirely from publicly available official government publications.

Primary sources include:

• —Annual Health Bulletin 2024
• —Medical Statistics Unit
• —Ministry of Health, Sri Lanka

All original statistics remain the intellectual property of the Ministry of Health, Sri Lanka.

Dataset Structure

The dataset is distributed in two complementary formats to support different research workflows.

1. Microsoft Excel Workbook

The complete dataset is available as a single Microsoft Excel workbook:

Sri_Lankan_Medical_Institutional_Data.xlsx

This workbook contains all twelve worksheets used throughout the accompanying research and is suitable for users who prefer working within spreadsheet software.

2. Individual CSV Files

For improved interoperability, reproducibility, and compatibility with data science tools, each worksheet is also provided as an individual CSV file within the data/ directory.

data/
├── 01_land_area_population.csv
├── 02_institutions_beds.csv
├── 03_icu_beds_distribution.csv
├── 04_operating_theatres.csv
├── 05_surgeons_physicians.csv
├── 06_asthma_heart_cvd.csv
├── 07_poisoning.csv
├── 08_pneumonia.csv
├── 09_traumatic_injuries.csv
├── 10_snake_bites.csv
├── 11_leading_causes.csv
└── 12_icu_locations.csv

Dataset Contents

Intended Uses

This dataset is intended for:

• —Healthcare accessibility modelling
• —Emergency medical service research
• —Spatial accessibility analysis
• —Health systems engineering
• —Medical GIS research
• —Public health analytics
• —Machine learning on healthcare infrastructure
• —Healthcare resource allocation
• —Policy simulation
• —Reproducibility of the accompanying publication

Research Paper

This dataset accompanies:

Golden Hour Divide: Trauma Care Accessibility and Resource Vulnerability in Sri Lanka

The paper introduces:

• —General Need-Gap Index (NGI)
• —Disease-specific Need-Gap Index
• —Terrain-aware H3 accessibility analysis
• —Resource-weighted vulnerability modelling
• —K-Means district archetype clustering
• —Policy simulation for specialist redistribution

Repository Structure

The dataset is provided in two formats for convenience.

Excel Workbook

Sri_Lankan_Medical_Institutional_Data.xlsx

Contains all twelve worksheets within a single Microsoft Excel workbook.

Individual CSV Files

The data/ directory contains each worksheet exported as an individual CSV file.

data/
├── 01_land_area_population.csv
├── 02_institutions_beds.csv
├── 03_icu_beds_distribution.csv
├── 04_operating_theatres.csv
├── 05_surgeons_physicians.csv
├── 06_asthma_heart_cvd.csv
├── 07_poisoning.csv
├── 08_pneumonia.csv
├── 09_traumatic_injuries.csv
├── 10_snake_bites.csv
├── 11_leading_causes.csv
└── 12_icu_locations.csv

Users may either use the Excel workbook or load the individual CSV files directly for analysis and machine learning workflows.

Citation

If you use this dataset, please cite:

@article{kirindage2026goldenhourdivide,
  title={Golden Hour Divide: Trauma Care Accessibility and Resource Vulnerability in Sri Lanka},
  author={Kirindage, Sonath and Nimsara, Vihanga and Rajapaksa, Sakindu and Hathurusinghe, Kavyanga and Dilshan, Lahiru and Arumugam, Subavarshana and Athukorala, Nathali and Wickramanayake, Sandareka and de Silva, Nisansa},
  year={2026},
  note={Under review}
}

Usage

Install the Hugging Face Datasets library:

pip install datasets

Load any dataset configuration:

from datasets import load_dataset

dataset = load_dataset(
    "sonath0427/sri-lankan-medical-institutional-data",
    "land_area_population"
)

print(dataset["train"])

Available Configurations

Licensing

This repository contains processed and reorganized data derived from publicly available publications of the Ministry of Health, Sri Lanka.

Users should appropriately acknowledge the Ministry of Health, Sri Lanka, when using the dataset.

The dataset itself is distributed under the license specified in this repository.

Limitations

• —Data represent official statistics for the year 2024.
• —Hospital accessibility is based on static road-network analysis.
• —Real-time traffic conditions are not included.
• —Disease counts represent reported healthcare utilization and may not reflect true incidence.
• —Updates published after the 2024 Annual Health Bulletin are not incorporated.

Ethical Considerations

This dataset contains no personally identifiable information (PII).

All data are aggregated at the hospital or district level and originate from publicly available government reports.

No patient-level records are included.

Acknowledgements

The authors acknowledge:

• —Ministry of Health, Sri Lanka
• —Medical Statistics Unit
• —Annual Health Bulletin 2024

for making the underlying data publicly available.

Contact

For questions regarding this dataset or the accompanying research, please contact the repository maintainers through GitHub.