Activity-Based Contracting for Insecticide-Treated Net (ITN) Mass Distribution Campaigns in the Democratic Republic of Congo: A Multi-province Quasi-experimental Evaluation

Abstract

BackgroundOperational inefficiencies continue to limit the effectiveness of insecticide-treated net (ITN) mass distribution campaigns, particularly in challenging operating environments such as the Democratic Republic of the Congo (DRC).Results-based contracting has been proposed to improve campaign performance, but evidence regarding its effectiveness at scale remains limited. Materials and methodsWe conducted a quasi-experimental evaluation of ITN mass distribution campaigns implemented in 20 provinces of the DRC between 2022 and 2024, comparing four provinces applying an activity-based contracting (ABC) approach with 16 provinces applying the standard input-based approach.The primary outcome was compliance with ITN allocation norms, as verified by independent agencies.Secondary outcomes included independently verified household coverage, campaign preparation time, and unit cost per ITN distributed.Three comparison frameworks were combined: concurrent, within-province historical, and contextually matched.Because the ABC approach was assigned at the provincial level, we analyzed all indicators at both the household level (two-sample Z-tests for proportions with risk differences and 95% confidence intervals) and the provincial level (exact Wilcoxon rank-sum tests, with the province as the unit of observation). ResultsCompliance with ITN allocation norms was higher in ABC provinces than in non-ABC provinces (94.4% versus 85.6%; risk difference +8.8 percentage points, 95% CI +8.6 to +9.0).The difference persisted when province was the unit of analysis, with complete separation between the two groups: the four ABC provinces occupied the four highest ranks among the 20 (median 93.6% versus 85.6%; exact p = 0.0004).It was reproduced in the matched four-versus-four comparison (exact p = 0.029).All secondary outcomes were more favorable in ABC provinces: independently verified household coverage 98.6% versus 95.8%, adjusted campaign preparation time 14.6 versus 22.8 weeks, and unit cost USD 0.668 versus USD 0.773 per ITN distributed, but none reached statistical significance at the provincial level, and all are reported as exploratory.Measured ITN loss rates were higher under ABC (0.33% versus 0.17%); because losses were established by independent physical inventory under ABC and by administrative balance elsewhere, the two figures are not measurement-equivalent and were not compared statistically. ConclusionsScaling up ABC beyond a single pilot province was associated with higher compliance with ITN allocation norms, consistently across concurrent, historical, and matched comparison frameworks and at the level at which the intervention was assigned.Differences in coverage, timelines, and unit costs pointed in the same direction but were not statistically established.Results-based contracting merits consideration to improve the quality of large-scale distribution campaigns in resource-constrained settings, pending confirmation from evaluations conducted independently of program implementers and using designs that allow formal control of confounding.