De-identified dataset and analysis code for: Nonlinear Association Between Fever-to-IVIG Interval and Clinical Outcomes in Kawasaki Disease — A Retrospective Cohort Study of 6,921 Children

Rigorous de-identification was applied: all direct identifiers and exact dates were removed; age is released in integer months; calendar year is released only as era categories; row order was randomly permuted; and extreme timing values were top-coded at 21 days. A full data dictionary, missing-data summary, de-identification protocol, and residual re-identification risk assessment are provided in the accompanying README file.An executable Python script (reproduce_analyses.py) is included, which reproduces all primary and sensitivity analyses reported in the article — including group comparisons (reference: 5–7 days), multivariable logistic regression with standardized covariates, nonlinearity likelihood-ratio tests, era-stratified analyses (2012–2018 vs. 2019–2025), and a sensitivity analysis restricted to complete KD — using only pandas, numpy, scipy, and statsmodels.Access: Controlled access under a data-use agreement (non-commercial academic research only; no re-identification attempts; no onward redistribution).

Requests should be sent to the corresponding author and will be acknowledged within 30 days.Ethics: Approved by the Institutional Review Board of Wuhan Children's Hospital (2024R034-E01), with a waiver of informed consent for de-identified retrospective data.