Efficacy of Wolbachia-mediated sterility for control of dengue: emulation of a cluster randomized target trial

BackgroundMatings between male Aedes aegypti mosquitoes infected with wAlbB strain of Wolbachia and wildtype females yield non-viable eggs. We evaluated the efficacy of releasing wAlbB-infected Ae. aegypti male mosquitoes to suppress dengue. MethodsWe specified the protocol of a two-arm cluster-randomised test-negative controlled trial (cRCT) and emulated it using a nationally representative test-negative/positive database of individuals reporting for febrile illness to any public hospital, general practitioner or polyclinic. We built a cohort of individuals who reside in Wolbachia locations versus a comparator control group who do not reside in Wolbachia locations. We emulated a constrained randomisation protocol used in cRCTs to balance dengue risk between intervention and control arms in the pre-intervention period. We used the inverse-probability weighting approach to further balance the intervention and control groups using a battery of algorithmically selected sociodemographic, environmental and anthropogenic variables. Intention-to-treat analyses was conducted to estimate the risk reduction of dengue given Wolbachia exposure. FindingsThe final cohort consisted of 7,049 individuals residing in areas treated by Wolbachia interventions for at least 3 months and 69,216 individuals residing in non-treated areas in the same time period. Intention-to-treat analyses revealed that, compared with controls, Wolbachia releases for 3, 6, 12 or more months was associated to a 47% (Odds ratio (OR): 0.53 [0.45-0.62]), 47% (OR: 0.53 [0.50-0.65]) and 59% (OR: 0.41 [0.39-0.50]) protective efficacy against dengue respectively. When exposed to 12 or more months of Wolbachia releases, protective efficacy ranged from 36% (OR: 0.64 [0.58-0.96]) to 77% (OR: 0.23 [0.22-0.33]) dependent on township, and from 48% (OR: 0.52 [0.48-0.7]) to 78% (OR: 0.22 [0.09-0.32]) across years. The proportion of virologically confirmed dengue cases was lower overall in the intervention arm, and across each subgroup. Protective efficacies were found across all townships, years, age and sex subgroups, with higher durations of Wolbachia exposure similarly associated to greater risk reductions of dengue. InterpretationOur results demonstrated the potential of Wolbachia-mediated sterility for strengthening dengue control in tropical cities, where dengue burden is the greatest. FundingThis study was supported by funding from Singapores Ministry of Finance, Ministry of Sustainability and the Environment, National Environment Agency, and National Robotics Program. JTL is supported by the Ministry of Education (MOE), Singapore Start-up Grant. SB is supported by an MOE Tier 2 grant.