Statistical Analysis Plan for a Cluster Randomised Trial of a Complex Intervention to Reduce Anaemia and Cardiometabolic Risk Amongst Pregnant and Postpartum Women in Rural India

1The SMART Health Pregnancy 2 trial aims to determine if the SMARThealth Pregnancy complex intervention can improve womens health in the year after pregnancy, specifically by decreasing the prevalence of anaemia by 9% and improving screening, referral and follow-up following a pregnancy affected by either anaemia, diabetes or hypertensive disorders of pregnancy. It is designed as a parallel-group cluster randomised superiority trial. The unit of randomisation was the Primary Health Centre (PHC). A matched-pair design, where clusters are paired before randomizing one to each trial arm, was followed. This statistical analysis plan pre-specifies the method of analysis for every outcome and key variable collected in the trial. The primary outcome is the difference in the proportion of participants with any degree of anaemia (Hb measured using a point-of-care test < 12g/dL) at 12 months after delivery between the intervention and control clusters. The primary analysis will consist of a mixed effects logistic regression with anaemia at 12 months as a binary outcome, treatment group, and baseline Hb as fixed effects and PHCs as random effects. The analysis plan also includes planned sensitivity analyses including covariate adjustments and subgroup analyses.