**1. Introduction**

Test procedures are published by the World Health Organization (WHO) on the use of the WHO insecticide susceptibility bioassay (or WHO tube bioassay) to monitor the resistance in adult mosquitoes to a range of insecticides commonly used for mosquito control [1]. Resistance monitoring using this approach relies on the collection of wild female adult mosquitoes or wild larvae, which are then reared to adulthood in a test facility. They are then exposed to a discriminating concentration (DC) of insecticide on a treated filter paper so that their knockdown and mortality can be scored.

The WHO tube bioassay is a simple direct response-to-exposure test. The test kit was developed in 1958 to test for the emergence of resistance to organochlorine and organophosphate insecticides following widespread resistance to organochlorine insecticides [2]. It was designed to expose a defined number of adult mosquitoes of known ages and physiological statuses to an insecticide impregnated on a filter paper for a standard exposure time (1 h).

While the WHO test procedures provide parameters for some key environmental conditions which should be kept constant while carrying out the bioassay, during insecticide resistance monitoring there are multiple potential sources of (non-resistance-associated) variability, which can influence the result of the bioassay. During a recently concluded formal WHO multicenter study to establish species-specific discriminating concentrations and procedures for new and existing insecticides, WHO tube and WHO bottle assays (an adaptation of the Centre for Disease Control (CDC) bottle bioassay developed to align end points with those of the WHO tube test) were used to generate concentration response data in multiple testing centers for a range of insecticides and multiple *Anopheles* and *Aedes* species, so that DCs could be established and validated. Within this extensive dataset, a substantial degree of variation was seen both within and between centers performing replicate assays using the same standardized methodologies [3].

One of the sources of variability in data generated when using the WHO tube bioassay methodology is the mosquitoes being tested. When using the method in the field to screen for resistance in the target population of an intervention, wild-caught adult mosquitoes should be used so that any differences in susceptibility may more closely reflect the changes in intrinsic resistance level seen for a particular intervention, and the sampled populations will be representative samples of the wild vector population in terms of age distribution and genetic variability. However, the age distribution, blood feeding status, nutritional status, and gravidity will vary between samples and potentially reduce the comparability of the results between tests and between sites. This differs from the mosquito populations that DCs are established on. Moreover, while operators using this methodology should ensure that they follow the guidance for the selection of mosquitoes for testing, it is still possible to accidentally include mosquitoes that fall outside these parameters (e.g., older than 2–5 days, males, partially blood-fed) when testing with wild-caught mosquitoes. Instead of wild-caught mosquitoes, F1 progeny of wild-caught mosquitoes can be used; although this requires facilities to rear and test the mosquitoes, there is greater control over the rearing conditions than for wild-caught mosquitoes.

Larval rearing conditions in laboratories have been shown to have an impact on bioassay results. Overcrowding or poor diet reduce insecticide tolerance by reducing size and fitness, for example [4,5]. Any impact on longevity because of larval rearing conditions could impact the outcome on mosquito survival. Poor mosquito survival could lead to high control mortality and more discarded tests (control mortality > 20%); this will affect the feasibility of testing. Larval rearing may also be more important for insecticides such as chlorfenapyr, where the effect of the compound is impacted by metabolism, though the correlation between longevity and size is not always positive [6]. In contrast, when the WHO tube bioassay is employed for research, a well-characterized (and ideally susceptible, so that resistance levels do not have to be maintained) laboratory strain can be used alongside the mosquito strain of interest. The benefit of this is that the researcher knows the rearing conditions of their laboratory strain and the resistance status and background

of the strain, while this is not the case for field-caught mosquitoes, this well-characterized susceptible reference strain can be tested alongside as a comparator.

The effect of time-of-day of testing on bioassay results is not well-explored in the literature, but *Anopheles* typically bite at night when they may be more metabolically active, and so susceptibility testing may yield differential results if conducted during the day than during the night. Most testing is carried out during the day, so this is unlikely to be a significant source of variability of the data currently being generated using the WHO tube bioassay method. It is good practice, however, to conduct resistance-monitoring assays at the same time of day each time and report the testing time alongside the data to aid interpretation, as it has been shown that time of day can impact metabolic detoxification and insecticide resistance in *Anopheles gambiae* [7]. Chlorfenapyr (while not validated for use in the WHO tube bioassay) is also strongly affected by a temperature of <25 ◦C [3].

Mosquito age has been shown to affect insecticide resistance, with mosquitoes older than 10 days post-emergence showing increased susceptibility to insecticides [8]. As well as age of mosquito, the nutritional status of adults can also affect the response to insecticide exposure. Machani et al. showed that the ingestion of a blood meal increased insecticide susceptibility [9]. Further to this, it has also been shown that lowering the temperature during insecticide susceptibility testing below the recommended 27 ± 2 ◦C can strongly affect insecticide tolerance [10]. As part of a study conducted in a Ugandan field insectary, which lacked environmental controls but where temperature and humidity were monitored, a strong and highly statistically significant decline in *A. gambiae* mortality was detected as humidity increased [11]. In light of this, it is important to be as consistent as possible when performing susceptibility bioassays and, where it is not possible to control the conditions fully, at least to understand the effect external factors can have on the outputs from this testing and report the environmental conditions alongside the data so that the results can be interpreted accordingly.

The WHO test procedures for monitoring Insecticide susceptibility have been reviewed and updated multiple times since their original publication in 1958, and some methodological details have changed between versions. We therefore set out to review the current literature to identify which test procedures are being referenced when using this method, what data and methodological detail is being reported when this method is used, and where data gaps lie for this methodology. We aimed to achieve this by looking for parameters in the test procedures which leave room for interpretation and using a literature search to explore how these parameters can influence the results of the bioassay. Parameters which were not clearly defined or supported with evidence and where evidence is not already available in the literature were chosen to be explored experimentally. In doing this, we hope to sugges<sup>t</sup> additional guidance on the optimum method for performing the WHO tube bioassay, as well as the key information required for the reporting of insecticide resistance data, thus producing more robust data and reporting it in a way that supports more meaningful interpretations.

#### **2. Materials and Methods**

#### *2.1. Test Procedures Review*

Thirteen WHO documents containing details outlining how to perform the WHO tube bioassay or the rationale behind the bioassay parameters, both published test procedures and meeting reports were reviewed to extract specific methodological details outlined in the test procedures. For each of these documents, the specifics and justifications for the methodological details outlined below were noted for each document and then compared.


