Statistical Analysis Plan for the Bedside Ultrasound Conducted in Kids with distal upper Limb fractures in the Emergency Department (BUCKLED) trial

Background The Bedside Ultrasound Conducted in Kids with distal upper Limb fractures in the Emergency Department (BUCKLED) trial is a multicentre, randomised, open-label, non-inferiority study comparing point-of-care ultrasound to x-ray as the initial imaging modality in children with clinically non-angulated distal forearm injuries. Objective The purpose of this document is to minimise bias by defining and making publicly available our analysis approach prior to completing the study and reviewing or analysing the trial data. Methods We developed a statistical analysis plan (SAP) for reporting and analysing the BUCKLED trial including our approach to expected protocol deviations, withdrawals, missing data, and loss to follow up. Study methods are described using a previously published study protocol. Broad statistical analysis principles are outlined. Primary and secondary trial outcomes are described, along with appropriate methods for statistical comparison. Results This report pre-specifies the SAP for the primary outcome of upper limb function at 4 weeks, using an intention-to-treat analysis with a hypothesis of non-inferiority. Results of a per-protocol analysis will also be reported. Subgroup analyses will include diagnostic category and age category and secondary outcomes will include cost, quality of life, satisfaction, pain, adverse events, rates of imaging, Emergency Department length of stay, treatment time and diagnostic accuracy. Methods of statistical comparison are outlined. Conclusion This report describes the SAP for the BUCKLED trial, to minimise analysis bias and ensure transparency and internal validity for the findings of this randomised controlled trial.