Statistical analysis plan for the Chinese Herbal medicine in Acute INtracerebral haemorrhage (CHAIN) trial

Introduction: The traditional Chinese medicine herbal compound FYTF-919 (Zhong Feng Xing Nao prescription) may improve outcome from acute intracerebral hemorrhage by reducing brain edema, hematoma absorption, and enhancement of the immune system. We outline the statistical analysis plan (SAP) for the Chinese Herbal medicine in Acute INtracerebral hemorrhage (CHAIN) study. CHAIN is a multicenter, prospective, randomized, double-blind, placebo-controlled trial being undertaken at 20–30 hospitals in China. After the completion of eligibility checks, patients are randomly allocated to FYTF-919 (100 mL per day, oral) or matching placebo over 28 days. A sample size of 1,504 patients is estimated to provide 90% power (α 0.05) for a 0.06 absolute improvement in the primary outcome of utility-weighted modified Rankin scale scores at 90 days, analyzed by general linear regression. Methods: The SAP was developed by the study statistician, principal investigators, international experts, and the study project manager. The plan provides details for analyzing baseline characteristics, patient management, and outcomes. It includes provisions for covariate adjustments, subgroup analysis, the handling missing data, and in the conduct of sensitivity analyses. Discussion: A predefined SAP was established for CHAIN, facilitating transparent and verifiable analysis. The CHAIN SAP was prospectively developed with a focus on maintaining high-quality standards of internal validity to minimize potential analysis biases.