Does washing insecticide-treated nets 20 times for experimental hut evaluations provide a suitable proxy for their end-of-life performance under household conditions?

Background Insecticide-treated nets (ITNs) are washed 20 times as part of experimental hut trials to simulate the loss of active ingredient (AI) occurring over their intended 3-year lifespan and estimate insecticidal durability. The ability of the 20-wash method to predict the end-of-life performance of ITNs has not been empirically validated. Methods We performed an experimental hut trial to compare the efficacy of new ITNs unwashed and washed 20 times to field-aged ITNs withdrawn from households three years post-distribution against a pyrethroid-resistant vector population in Covè, Benin. Four products from pyrethroid-only (Interceptor®), pyrethroid-piperonyl butoxide (PermaNet® 3.0), pyrethroid-pyriproxyfen (Royal Guard®) and pyrethroid-chlorfenapyr (Interceptor® G2) ITN types were tested. Net pieces were tested in bioassays and sent for chemical analysis to assess differences in surface AI bioavailability and total chemical content between washed and field-aged nets. Susceptibility bioassays were also performed to assess insecticide resistance in the Covè vector population. Results Mosquito mortality in experimental huts was similar or slightly higher with field-aged nets than washed nets with Interceptor® (11% vs 10% p = 0.339, OR = 1.19, 95% CIs: [0.84,1.69]), PermaNet® 3.0 (12% vs 18% p < 0.001, OR = 1.78, 95% CIs: [1.34,2.38]) and Royal Guard® (9% vs 14% p = 0.076, OR = 1.33, 95% CIs: [0.97,1.83]). Likewise, field-aged Royal Guard® induced a similar reduction in fertility to washed Royal Guard® (22% vs 29%, p = 0.066). In contrast, mortality was significantly lower with field-aged nets Interceptor® G2 compared to washed nets (54% vs 19%, p < 0.001, OR = 0.18, 95% CIs: [0.14,0.24]). Blood-feeding inhibition was higher with field-aged nets than washed nets across all ITN types. Retention of non-pyrethroid AIs was lower than the pyrethroid, particularly with field-aged nets (PermaNet® 3.0 (roof): 25% vs 68% p < 0.001, Royal Guard®: 27% vs 53% p < 0.001, Interceptor® G2: 14% vs 39% p < 0.001). Conclusion In this setting, the 20-wash method provided a suitable proxy for the end-of-life killing and sterilising performance of Interceptor®, PermaNet® 3.0 and Royal Guard® in experimental huts. In contrast, washing overestimated the end-of-life performance of Interceptor® G2 for mortality and underestimated the personal protection of all field-aged ITNs.