The excito-repellency effect of all the treated nets was consistently high (≥ 50%) compared to the control. Mosquitoes were affected by pyrethroid on the nets, corroborating previous findings that pyrethroid-impregnated nets provide an irritant barrier against susceptible and resistant An. gambiae s.s mosquitoes, due to the excito-repellency and knock-down effects of pyrethroids . The absence of holes in the nets, for ethical reasons, could have impacted the excito-repellency effect of all the treated nets, as sleepers would be accessible to An. gambiae s.s, Culex sp, Mansonia sp and other mosquitoes when under holed nets. Access to a blood meal may then reduce the excito-repellency rate recorded with each net. However, in areas with pyrethroid resistance, such as Benin , Burkina Faso and Cameroon , even when holes were present in the nets, the mortality rates recorded were low (30-40%) whereas in susceptible area such as Tanzania, mortality were above 80%  confirming the fact that mosquitoes resistance to pyrethroid reduced efficacy of nets. Low blood feeding rates were recorded for all treatment arms because nets were tested intact.
Cone bioassays against resistant An. gambiae s.s showed that the KD rate was > 95% and the mortality was > 80%, only with unwashed PermaNet® 3.0, although with washed 20x PermaNet® 3.0 a mortality rate > 80% was recorded against resistant An. gambiae s.s. Four out of five of the cone bioassays were performed on the polyester sides of washed PermaNet® 3.0 coated with high dose of deltamethrin (85 mg/m2). Previous studies conducted in sub-Saharan Africa have shown that the deltamethrin content remained high in the polyester sides of washed PermaNet® 3.0 even after 20 washes [6, 17]. Chemical analysis of deltamethrin and PBO content on nets used in this study was not conducted.
The highest mortality rates against An. gambiae s.s (55.0%) and Culex sp (51.6%) were recorded with unwashed PermaNet® 3.0; mortality rates recorded with the other treatment arms were low (less than 40%) and did not differ significantly. The high dose of deltamethrin alone in the side panels of unwashed PermaNet® 3.0 could explain the significant difference in mortality of resistant An. gambiae s.s with this net compared to the other treatment arms. A previous study showed that the loss of deltamethrin after washing is very low in the side panels of PermaNet® 2.0  where the dose of deltamethrin was initially equal to 55 mg/m2 (±25%) before washing. That could partially explain why in these studies unwashed and washed PermaNet® 2.0 performed equally  confirming results reported in this study even if none measurement related to the deltamethrin and PBO doses was done before and after nets washing in the study.
The loss of efficacy of washed PermaNet® in this study could be due to the resistance of An. gambiae s.s to pyrethroids, as in susceptible area such as Tanzania, the mortality rates recorded are usually above 80% even with washed PermaNet® 2.0 . Consequently, washing is considered to be a more important loss mechanism for coated nets than for incorporated nets. Over the course of 20 washes, there was a weak partial loss of activity against the resistant An. gambiae s.s and Culex spp with PermaNet® 3.0 in the experimental hut trial, which was supported by bio-efficacy data indicating a reduction in mortality rates detected via cone bioassays on washed PermaNet 3.0. Probably the loss of deltamethrin after 20 washings in the polyester sides of PermaNet® 3.0, combined with the high pyrethroid resistance rates in An. gambiae s.s and Culex spp explains the low mortality rate recorded with washed PermaNet® 2.0 and PermaNet® 3.0. Results from cone bio-assays showed that the mortality rate recorded with 20x washed PermaNet 3.0 in the roof and sides and 20x washed PermaNet 2.0 against resistant An. gambiae was equal to 86% and 57%, respectively. Similar bioassays results were reported in Ladji, in southern Benin where An. gambiae s.s M form species were resistant to pyrethroids and DDT, with high frequency of kdr and metabolic resistance. In Benin, similar mortality rates were observed with An. gambiae tested in WHO susceptibility kits with deltamethrin (43.9% vs. 45.2%) . The current study demonstrated that washed and unwashed PermaNet 2.0, washed PermaNet 3.0 and the conventionally-treated washed net (CTN) performed equally, while unwashed PermaNet 3.0 was associated with increased mortality. This is consistent with a recent study carried out in Benin, West Africa, which showed that although unwashed PermaNet® 3.0 was associated with a higher mortality than unwashed PermaNet 2.0 or unwashed Olyset, once washed, PermaNet 2.0 and 3.0 performed similarly . In Vietnam where resistance did not seem to be affecting net efficacy, all PermaNet arms were performing slightly better than conventionally treated nets washed until just before exhaustion . In Burkina Faso, in Vallee du Kou where the kdr mutation frequency was high (> 80%), results showed a strong reduction of LLIN efficacy. In this area, a significantly higher mortality and blood feeding inhibition was associated with unwashed and washed PermaNet® 3.0 compared to unwashed and washed PermaNet® 2.0 .
This study raises concerns that resistance appears to be having an impact on the efficacy of the nets in Ivory Coast (<40% mortality), in southern Benin (40 to < 30% mortality) with deltamethrin lambda-cyhalothrin [5, 15] treated nets and in Burkina Faso (<40% mortality)  compared to >80% mortality in Tanzania where the vectors are susceptible . Resistance mechanisms in Anopheles gambiae from the village of Ladji in Benin include kdr and metabolic mechanisms [5, 19]. Resistance due to the involvement of P450s and kdr [15, 20] has undermined attempts at malaria control with deltamethrin residual spraying in southern Africa caused by An. funestus . Elevated P450 activity in a strain of An. gambiae from Cameroon has been found to reduce the efficacy of permethrin-treated netting in laboratory tests [22, 23]. However, recent cone bio-assays performed on PermaNet 2.0 and PermaNet 3.0 after eight months use with metabolic resistant An. gambiae and Anopheles arabiensis from northern Cameroon showed very high mortality rates (Koudou et al, unpublished data). Enzyme-based pyrethroid resistance mechanisms, such as elevated esterases and/or P450's, combined mechanisms such as kdr and metabolic resistance, and other mechanisms such as reduced penetration of insecticide, may be more of an obstacle to the control of malaria vectors in Cote d'Ivoire .
Washing of PermaNet® 2.0 and PermaNet® 3.0 did not affect their deterrent effects against resistant An. gambiae s.s or Culex sp. However, the deterrent effect, i.e. the reduction in the number of mosquitoes entering the hut, may not be a reliable indicator of ITN efficacy, as within the same geographical area this index has varied considerably with different vectors and nets (including PermaNet® 2.0) between experiments, e.g. from zero to 70.0% against kdr-based pyrethroid resistant An. gambiae s.s [25, 26].