Uganda HIE Synthetic FHIR R4 Dataset: 10,000 Patient Records Conforming to Uganda HMIS Data Standards

Uganda HIE Synthetic Dataset: 10,000 FHIR R4 Patient Records This dataset comprises 10,000 synthetic patient records conforming to HL7 FHIR R4 (4.0.1), generated for the functional evaluation of a national Health Information Exchange (HIE) platform designed for Uganda. The dataset includes 10,000 Patient resources, 19,945 Encounters, 19,945 Conditions, and 69,758 Observations, distributed across 20 FHIR R4 transaction bundles (uganda-hie-synthetic-bundle-001.json through 020.json). Records were generated using a purpose-built tool (uganda-fhir-synth v1.0.0) with a fixed random seed (42) for full reproducibility. Patient demographics reflect Uganda's age distribution and linguistic diversity (Acholi, Luganda, Runyankole, Swahili, and English name pools). Clinical content uses ICD-10 diagnosis codes drawn from Uganda HMIS disease surveillance categories (malaria, tuberculosis, HIV, non-communicable diseases, maternal and neonatal conditions) and LOINC-coded vital sign and laboratory observations. Facility references use Uganda Ministry of Health Facility List (MHFL) codes, and patient identifiers follow the structural format of Uganda's National Identification Number (NIN) — these are synthetic identifiers and do not correspond to real individuals. All resources conform to four custom FHIR profiles (UgandaPatient, UgandaEncounter, UgandaObservation, UgandaCondition) under the canonical base hin.go.ug. Ethics statement: This dataset is entirely synthetic. No real patient data was used at any stage of generation. All names, identifiers, and clinical values were generated algorithmically. No individual can be identified from this dataset. A machine-readable summary of generation parameters is provided in manifest.json. Licence: CC BY 4.0