Code and Aggregate Data for "Resurging Human Brucellosis despite Strengthened Control in a Semi-Arid Pastoral County of Xinjiang, China: Disentangling Case Detection from Transmission and Identifying Cost-Saving One Health Interventions"

This dataset contains all materials required to reproduce the study "Resurging Human Brucellosis despite Strengthened Control in a Semi-Arid Pastoral County of Xinjiang, China: Disentangling Case Detection from Transmission and Identifying Cost-Saving One Health Interventions." It comprises: (1) de-identified aggregate data — monthly and annual (2016–2023) admission counts for human brucellosis from the county-designated referral hospital (June 2014–June 2024) and township-level case distribution (townships pseudonymized with letter codes A–O); (2) Python source code — a deterministic sheep–environment–human compartmental transmission model, an approximate Bayesian computation (ABC) calibration routine, 10-year projections of eight intervention scenarios (S0–S7) with a double-ledger (human plus livestock) economic evaluation, and scripts reproducing Figures 1–4; (3) model outputs — posterior parameter draws, scenario comparison results, and all figures.

Individual-level hospital records are not included, in compliance with Chinese infectious-disease data regulations, and the county identity is anonymized as in the manuscript. Requirements: Python 3.9+ with NumPy, SciPy, pandas, and matplotlib. To reproduce all results, run run_calibration.py, run_scenarios.py, and make_figures.py in sequence from the src/ directory; the full pipeline takes approximately 1–2 minutes.