# Azithromycin mass drug administration to reduce child mortality in Niger (AVENIR II): a master protocol for a cluster-randomized adaptive platform trial to evaluate community-based health interventions

## TL;DR

This paper outlines a trial in Niger to study the effects of giving azithromycin to children to reduce deaths while monitoring antibiotic resistance and other health outcomes.

## Contribution

The paper introduces a flexible, adaptive trial design to evaluate azithromycin's impact and other health interventions in real-world settings.

## Key findings

- Azithromycin MDA reduces child mortality but raises antimicrobial resistance concerns.
- The trial design allows for ongoing monitoring and re-randomization every 2 years.
- The platform trial can adapt to new evidence and evaluate additional interventions.

## Abstract

Trials have demonstrated that azithromycin mass drug administration (MDA) to children 1–59 months old reduces mortality but increases antimicrobial resistance (AMR). The World Health Organization recommends that programs include mortality and AMR monitoring. Niger is expanding the azithromycin MDA for child survival program nationwide.

To establish program monitoring and leverage the infrastructure to evaluate other community health interventions, AVENIR II is designed as a cluster-randomized adaptive platform trial with monitoring and re-randomization every 2 years. The initial focus is to monitor under-5 mortality, AMR, implementation, and safety as the azithromycin program expands in Niger. All eligible primary health center catchment areas (Centre de Santé Intégrés, CSIs) will be included in biannual oral azithromycin MDA to children 1–59 months old. A subset will be randomized to delay MDA for the first 2 years, after which they will receive MDA. Another subset will then be randomized to stop MDA for the next 2 years. The proportion randomized to delay or stop will be determined using an adaptive algorithm including (1) results of prior azithromycin MDA mortality trials, (2) expert opinion on the appropriate ethical balance between delivering the program and monitoring AMR, and (3) statistical power to detect a programmatically relevant difference between arms. We anticipate 5–10% of CSIs will be randomized to delay or stop at each randomization. Mortality and AMR will be monitored at baseline and every 2 years. Implementation and safety outcomes will be monitored continuously. To enable ongoing monitoring while ensuring program access, CSIs receiving MDA will be re-randomized using the adaptive algorithm updated with new mortality results, and no CSI will go without MDA for more than 2 years. In this platform design, additional arms may be added or dropped based on information from other studies, updates to guidelines, or preferences of Niger policymakers, and other interventions may be evaluated.

The risk of AMR has led to caution in the implementation of azithromycin MDA. We present a design that enables continued rigorous evaluation of program impact on key outcomes, with flexibility to evaluate other interventions as well.

ClinicalTrials.gov NCT06358872. Registered on April 2024.

The online version contains supplementary material available at 10.1186/s13063-026-09561-2.