Back

Deriving Greater Value from Malaria Bed Nets through Extending Net Retention: A Modelling Study

Floyd, J. R.; Trett, A.; Golumbeanu, M.; Champagne, C.; Pothin, E.; Ward, A.; Cohen, J. M.; Seethaler, T.

2025-08-06 epidemiology
10.1101/2025.08.02.25332507 medRxiv
Show abstract

Insecticide-treated nets (ITNs) are central to malaria prevention in sub-Saharan Africa, but their effectiveness depends on how long households retain and use them. Although ITNs are designed to last three years, recent evidence suggests median retention times are significantly shorter in most countries, creating gaps in protection between mass distribution campaigns. In this study, we model the potential impact of improving ITN retention by enhancing physical durability on malaria burden across 43 countries in sub-Saharan Africa using an individual-based malaria transmission model. We simulate scenarios in which ITN retention times are extended by 1 to 8 months beyond current country-specific baselines and compare averted cases relative to a business-as-usual scenario. Results show that a 4-month increase in retention could avert 6% of malaria cases over three years; and that to get the same amount of impact from increasing ITN access would require an additional 171 million nets. Gains from improved retention are greatest in the first few months of extension and accumulate over time, making durability enhancements a cost-effective strategy for maintaining protection as funding constraints grow. These findings support the integration of predictive durability metrics, such as the Resistance to Damage (RD) score, into ITN procurement decisions and highlights the importance of selecting well performing, high-value ITNs to strengthen malaria control without deploying more ITNs.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.