De-identified outpatient prescription dataset of a renowned TCM practitioner (Prof. ZHAO Hanqing M.D., Ph.D., 2025): 335 prescriptions with 5,951 herb-level records

Background: Complete outpatient herbal prescription records of Professor Zhao Hanqing , January–December 2025, TCM Internal Medicine department, Heniantang Clinic. Contents: 5,951 herb-level records; 335 prescriptions; 122 de-identified patients; 438 unique herb strings with doses; 90 substantive disease labels (91 values including "Not recorded") and 101 substantive syndrome-pattern labels (102 values including "Not recorded"). Diagnoses are preserved exactly as recorded in the source clinic records; no terminology mapping was applied. Direct identifiers removed: sequential patient codes (P001–P122), prescription codes (RX001–RX335), ages generalised to ten-year bands, visit dates generalised to half-year periods (minimum k-anonymity = 3 at prescription level, 24 sex × age-band × half-year classes). What is new in version 1.2: for seasonal-prescribing research the solar term (jieqi, ~15-day window) of each visit, the administration route, and the dosing frequency. The solar-term field narrows visit windows to approximately 15 days (minimum k = 1; 116 of 161 sex × age-band × solar-term classes fall below k = 3, covering 160 of 335 prescriptions). Users requiring the stronger anonymity guarantee should analyse at half-year granularity. Files: prescription-level table (17 columns), prescription summary (15 columns), data dictionary, herb name lookup, QA records, Frictionless datapackage, README, CC BY 4.0 licence. Analysis code: https://github.com/zhhqtt/zhao-hanqing-tcm-dataset-2025 (commit fff843c)