Implications of bio-efficacy and persistence of insecticides when indoor residual spraying and long-lasting insecticide nets are combined for malaria prevention

Abstract

Background: Bio-efficacy and residual activity of insecticides used for indoor residual spraying (IRS) and long-lasting insecticide nets (LLINs) were assessed against laboratory-reared and wild populations of the malaria vector, Anopheles arabiensis in south eastern Tanzania. Implications of the findings are examined in the context of potential synergies and redundancies where IRS and LLINs are combined.

Methods: Bioassays were conducted monthly for six months on three LLIN types (Olyset® PermaNet 2.0®,and Icon Life®) and three IRS treatments (2 g/m2 pirimiphos-methyl, 2 g/m2 DDT and 0.03 g/m2 lambda-cyhalothrin, sprayed on mud walls and palm ceilings of experimental huts). Tests used susceptible laboratory-reared An. arabiensis exposed in cones (nets and IRS) or wire balls (nets only). Susceptibility of wild populations was assessed using WHO diagnostic concentrations and PCR for knock-down resistance (kdr) genes.

Results: IRS treatments killed ≥ 85% of mosquitoes exposed on palm ceilings and ≥ 90% of those exposed on mud walls, but up to 50% of this toxicity decayed within 1-3 months, except for DDT. By 6th month, only 7.5%, 42.5% and 30.0% of mosquitoes died when exposed to ceilings sprayed with pirimiphos-methyl, DDT or lambda-cyhalothrin respectively, while 12.5%, 36.0% and 27.5% died after exposure to mud walls sprayed with the same insecticides. In wire-ball assays, mortality decreased from 98.1% in 1st month to 92.6% in 6th month in tests on PermaNet 2.0®, from 100% to 61.1% on Icon Life® and from 93.2% to 33.3% on Olyset® nets. In cone bioassays, mortality reduced from 92.8% in 1st month to 83.3% in 6th month on PermaNet 2.0®, from 96.9% to 43.80% on Icon Life® and from 85.6% to 14.6% on Olyset®. Wild An. arabiensis were 100% susceptible to DDT, 95.8% to deltamethrin, 90.2% to lambda cyhalothrin and 95.2% susceptible to permethrin. No kdr gene mutations were detected.

Conclusions: In bioassays where sufficient contact with treated surfaces is assured, LLINs and IRS kill high proportions of susceptible An. arabiensis mosquitoes, though these efficacies decay gradually for LLINs and rapidly for IRS. It is, therefore, important to always add intact nets in sprayed houses, guaranteeing protection even after the IRS decays, and to ensure accurate timing, quality control and regular re-spraying in IRS programmes. By contrast, adding IRS in houses with intact LLINs is unlikely to improve protection relative to LLINs alone, since there is no guarantee that unfed vectors would rest long enough on the sprayed surfaces, and because of the rapid IRS decay. However, there is need to clarify these effects using data from observations of free flying mosquitoes in huts. Physiological susceptibility of An. arabiensis in the area remains 100% against DDT, but is slightly reduced against pyrethroids, necessitating caution over possible spread of resistance. The loss of LLIN toxicity, particularly Olyset® nets suggests that protection offered by these nets against An. arabiensis may be primarily due to physical bite prevention rather than insecticidal efficacy.

Figure 1

Results of monthly bioassays showing residual activity of various IRS compounds sprayed on mud walls and Mikeka ceilings of experimental huts. The controls in this assays consisted of bioassays conducted walls and ceilings of unsprayed experimental huts.

Figure 2

Results of monthly bioassays showing residual activity of various IRS compounds sprayed on panels (1 sq metre each) lined with Mikeka (i.e. Mikeka panels) or mud (i.e. mud panels). The data were corrected using Abbot’s formula [27] to account for unexpectedly high mortalities in the controls. The controls consisted of panels lined with unsprayed mud or Mikeka.

Figure 3

Results of monthly bioassays showing residual activity of various LLINs when tested using either the standard WHO cone assays or the wire ball method (two intersecting wire cycles, each measuring 15 cm diameter, around which the test net is wrapped to form a netting ball). The controls in these assays consisted of a non-insecticidal mosquito net. No assays were conducted on the third month due to lack of mosquitoes.