Environmental data and aggregate results for short-term weather changes before epistaxis presentation in Xining

We provide environmental data and aggregate results supporting our time-stratified case-crossover study of short-term weather changes before first hospital presentation for epistaxis in Xining, China. The study includes 370 patients whose current episode required management through the otolaryngology emergency service. The collection includes daily environmental series, station and reanalysis comparisons using matched observation windows, clinical interval summaries, local weather-change models, lag-window and weather-trajectory analyses, and spatial and concurrent-weather sensitivity analyses. Supporting Data Sets S1-S4 contain the numerical results and source data underlying the manuscript and supplementary analyses. The supporting materials distinguish statistical support for signed temperature change from inconclusive evidence for maximum absolute change. Weather-state associations, their interquartile-range scales, hypothesis families and corresponding source tables are documented. Detailed source registers describe station coordinates, data availability and nearby oxygen field observations. The collection also includes simulation results for our Fixed-Sample Power Assessment. Power was evaluated after retrospective record collection and identification of all 370 eligible patients; this complete cohort determined the sample size. No a priori power calculation was used to determine sample size. Environmental inputs and aggregate results are available under CC BY 4.0, with source-specific terms documented in the data guide and licensing statement. Access to individual clinical records is administered by the responsible institution through the corresponding author. Companion software: https://doi.org/10.5281/zenodo.23813519. Code repository: https://github.com/qudadd/xining-weather-epistaxis.