Age-based targeting of biannual azithromycin distribution for child survival in Niger: an adaptive cluster-randomized trial protocol (AVENIR)

Abstract

Background: Biannual distribution of azithromycin to children 1-59 months old reduced mortality by 14% in a cluster-randomized trial. The World Health Organization has proposed targeting this intervention to the subgroup of children 1-11 months old to reduce selection for antimicrobial resistance. Here, we describe a trial designed to determine the impact of age-based targeting of biannual azithromycin on mortality and antimicrobial resistance.

Methods: AVENIR is a cluster-randomized, placebo-controlled, double-masked, response-adaptive large simple trial in Niger. During the 2.5-year study period, 3350 communities are targeted for enrollment. In the first year, communities in the Dosso region will be randomized 1:1:1 to 1) azithromycin 1-11: biannual azithromycin to children 1-11 months old with placebo to children 12-59 months old, 2) azithromycin 1-59: biannual azithromycin to children 1-59 months old, or 3) placebo: biannual placebo to children 1-59 months old. Regions enrolled after the first year will be randomized with an updated allocation based on the probability of mortality in children 1-59 months in each arm during the preceding study period. A biannual door-to-door census will be conducted to enumerate the population, distribute azithromycin and placebo, and monitor vital status. Primary mortality outcomes are defined as all-cause mortality rate (deaths per 1000 person-years) after 2.5 years from the first enrollment in 1) children 1-59 months old comparing the azithromycin 1-59 and placebo arms, 2) children 1-11 months old comparing the azithromycin 1-11 and placebo arm, and 3) children 12-59 months in the azithromycin 1-11 and azithromycin 1-59 arms. In the Dosso region, 50 communities from each arm will be followed to monitor antimicrobial resistance. Primary resistance outcomes will be assessed after 2 years of distributions and include 1) prevalence of genetic determinants of macrolide resistance in nasopharyngeal samples from children 1-59 months old, and 2) load of genetic determinants of macrolide resistance in rectal samples from children 1-59 months old.

Discussion: As high-mortality settings consider this intervention, the results of this trial will provide evidence to support programmatic and policy decision-making on age-based strategies for azithromycin distribution to promote child survival.

Trial registration: This trial was registered on January 13, 2020 (clinicaltrials.gov: NCT04224987 ).

Keywords: Adaptive trial; Azithromycin; Cluster-randomized trial; Mass drug administration; Mortality.

Fig. 1

AVENIR trial design summary for communities included in primary outcomes during Stage I. Communities randomized to the azithromycin 1–11 arm will receive biannual distribution of azithromycin to children 1–11 months of age with placebo to children 12–59 months of age. Communities randomized to the azithromycin 1–59 arm will receive biannual distribution of azithromycin to children 1–59 months of age. Children randomized to the placebo arm will receive biannual distribution of placebo to children 1–59 months of age

Fig. 2

Eligible regions in Niger with and AVENIR study timeline. Stage I includes implementation in the Dosso, Tahoua, and first half of the Maradi regions, with communities from Dosso and Tahoua contributing to the primary outcomes described in this protocol. Continuation to Stage II depends on the primary outcome results as well as logistical, financial, and security concerns. Stage II would include implementation in the remaining half of Maradi as well as the Zinder and Tillabéri regions. The map was created by the authors in R (R Foundation for Statistical Computing, Vienna, Austria)