Statistical analysis plan for the “empirical treatment against cytomegalovirus and tuberculosis in HIV-infected infants with severe pneumonia” clinical trial

Background: The EMPIRICAL trial aims to assess safety and efficacy of an empirical treatment against cytomegalovirus (CMV) and tuberculosis (TB) compared to standard of care (SoC), on adverse events and 15-day and 1-year mortality among infants living with HIV hospitalized with severe pneumonia in Africa. 

Methods and design: The EMPIRICAL trial (NCT03915366) is an international multicenter phase II-III, open-label randomized factorial clinical trial conducted in six African countries. The trial has four randomization arms in a 1:1:1:1 fashion with patients allocated to (i) TB-Treatment plus SoC, (ii) valganciclovir plus SoC, (iii) both TB-Treatment and valganciclovir plus SoC, and (iv) SoC only. 

Discussion: This paper describes the statistical analysis plan (SAP) for the trial which, per the study publication plan, needs to be published prior to the database lock and final analysis results. The SAP includes details of the analyses to be undertaken and unpopulated tables that will be reported to address primary and secondary endpoints. The database will be locked on 31st January 2025.