Implementation, coverage and impact of a programmatic mass drug administration campaign for malaria in southern...

Between 2022 and 2023, the Mozambican National Malaria Control Programme developed a locally-tailored implementation strategy for programmatic mass drug administration (pMDA), which was piloted in Chidenguele (Manjacaze District), southern Mozambique. Two rounds of door-to-door pMDA with dihydroartemisinin-piperaquine were conducted, targeting 59,271 individuals in 14,818 households, according to administrative data. Satellite imagery was used to support household enumeration and field navigation.

An evaluation of the pMDA was conducted. Coverage was assessed using programmatic data and a community household survey (n = 770 individuals). A quasi-experimental design, using neighboring areas not receiving pMDA as a comparison group, was used to evaluate the impact of the pMDA on clinical malaria incidence during the ensuing 22 months through a controlled interrupted time-series analysis of routine surveillance data, adjusting for covariates.

According to programmatic data, household availability coverage (households reached/target households) increased from 59.4% (8,796/14,818) in the first round to 94.3% (13,972/14,818) in the second, following optimization of the implementation strategy. Programmatic or contact coverage (individuals treated/target population) increased from 41.2% (24,437/59,271) to 69.7% (41,320/59,271). In the second round, 8% of the target population was not reached, 8.7% were absent during visits, 6.9% were ineligible, and 6.7% refused participation. The household survey showed similar coverages and estimated that 81.6% of respondents (628/770, 95% CI 78.6-84.2) were treated in at least one of the rounds. We found weak evidence of a larger decrease in malaria incidence immediately following the pMDA implementation in the pMDA group versus the comparison group (level change incidence risk ratio [IRR] = 0.66, p = 0.074), and no evidence of a difference in the malaria trend over time between the two groups (trend change IRR = 0.96, p = 0.104).

Although more than 90% of target households and individuals were reached in the second round, achieving the recommended 80% programmatic coverage remained challenging. This target was reached only when considering participation in either of the two rounds, highlighting the importance of conducting multiple rounds. High coverage requires strong community engagement, household revisits and/or fixed points, and substantial human and logistical resources. Using satellite imagery and triangulating with census data allowed us to estimate denominators, though challenges remain in this process. We found weak evidence of an impact on malaria incidence immediately following the pMDA and no evidence of a difference in the incidence trend over time. Overall, pMDA can be implemented under programmatic conditions, but is resource-intensive and should be reserved for specific contexts alongside other core malaria interventions.