Long lasting insecticidal nets (LLIN) are a key vector control tool for the prevention of malaria in endemic countries. They are estimated to have saved millions of lives [1, 2]. LLIN procurement volumes have continuously increased since widespread mass distributions began in the early 2000s. Annual LLIN deliveries are now approaching 300 million pieces, while the cost per net has more than halved in the last 10 years [3]. The increasing demand for cheap LLINs, the rapid spread of pyrethroid resistance and competition for LLIN market share between manufacturers has diversified the product palette in recent years [4–6]. Alongside much needed progress (e.g., with new and innovative active ingredient formulations), this increased market complexity also led to challenges for LLIN product quality assurance, and the risk and occurrence of substandard products [4, 7–12]. Recipient countries and donors increasingly recognize that LLINs are not just ‘mosquito nets’ but complex public health commodities that require stringent quality assurance and performance monitoring [4, 12, 13].
This issue is not new and the WHO published guidelines for testing LLINs, initially in 2005 [14]. In 2013, these guidelines were updated and augmented with additional guidance for monitoring the durability of LLINs under ‘operational conditions’ [15]. The declared aims of the WHO guidelines are (1) to provide specific standardized procedures for testing of LLINs and (2) to harmonize standard testing procedures [15].
Tests for LLIN bioefficacy under laboratory conditions, i.e., the ability of LLIN products to kill mosquitoes in cone bioassays, are an integral part of the WHO testing guidelines and play a crucial role in the LLIN product prequalification process and, increasingly, in post market surveillance [4, 7, 12, 16–18]. Given that the market fate of LLIN products hinges on their performance in these bioassays, it is important that they are harmonized and transparent [4, 13].
Bioefficacy tests usually involve the exposure of living mosquitoes to LLIN product material, and thus, even in their simplest form, these tests require a complex experimental setup including the maintenance of a mosquito colony [19, 20].
The simplest bioassay recommended to evaluate LLIN products for bioefficacy is the WHO cone bioassay [15, 21]. WHO guidelines outline in detail how cone bioassays should be performed, providing important parameters such as sample size, number of mosquitoes to be exposed, number of replicates to be performed, exposure time, and acceptable temperature and humidity ranges. However, other key parameters that may also affect the key cone bioassay endpoints of 60 min knockdown (KD60) and 24h mortality (M24) are either undefined or only partially defined by WHO. An obvious example for this is that the mosquito strain to be used in WHO cone bioassays is currently not further specified. The WHO guidelines require ‘susceptible’ (female) Anopheles mosquitoes [15, 21]. Yet, it is evident that ‘susceptibility’ does not mean the same (or even similar) response of different mosquito strains to standardized exposure to insecticides. In other words, using different mosquito strains in bioassays is synonymous with systematic bias [22].
This situation can be likened to the known species-dependent mosquito responses observed in WHO tube bioassays used for insecticide resistance monitoring and the recognized need for species-specific, WHO-recommended, discriminatory insecticide concentrations to be used in these assays [22, 23]. It is thus surprising that species-specific guidance for each prequalified LLIN product is lacking, as the declared priorities of harmonization and standardization of bioassays would clearly require this level of rigor.
Besides the use of different mosquito strains, other important bioassay parameters are also left open for interpretation or are not standardized even if the relevant guidelines are explicit. As a result, testing laboratories may be inclined to ‘optimize’ these undefined or less stringently enforced bioassay parameters to achieve the highest bioefficacy outcomes (i.e., high KD60 and M24). This stands in contrast to the overarching aim of ‘harmonization’ and ‘standardization’ as it may further increase systematic bias between testing facilities and lead to situations where LLIN products may routinely ‘pass’ testing criteria in some settings but not in others.
One such undefined (or partially defined) set of parameters is the configuration of the bioassay board to which the LLIN samples are pinned for the duration of the test. While this may seem trivial, even here multiple parameters may crucially influence bioassay results. For example, while WHO guidelines state that these boards should be placed at a 45 degree angle, Owusu et al., 2016, compared different angles for positioning of the cone bioassay board set up and found that mosquitoes (pyrethroid susceptible An. gambiae Kisumu-1 and pyrethroid resistant An. stephensi STI) spent more time on the nets at a 60 degree angle [24]. As a result, some studies are now reconfiguring this experimental detail in order to maximize the measured bioefficacy indicators of M24 and KD60 [20, 25]. Another such modification is the use of bioassay boards with circular holes i.e., the material behind the tested sample being removed, again with the intention to ‘maximize exposure’ and thus increase bioefficacy endpoints (Fig. 1). While WHO guidelines do not specify whether this should be done or not and we are not aware of studies having systematically quantified the resulting effect. Several laboratories are conducting WHO cone bioassays on boards with holes cut [20, 24] while others are not [12, 20].
Therefore, in this study, we investigated whether circular holes in the bioassay boards intended to ‘force the mosquitoes to stand on the net surface’ (as shown in Fig. 1) lead to systematic bias on the KD60 and M24 key bioassay endpoints.