**4. Addressing data deficits through programmatic entomological surveillance platforms**

**Figure 3.** The global distribution of reported measurements for (A) insecticide resistance, (B) the proportions of bloodmeals obtained from humans, (C) the proportion of human exposure to bites occurring indoors, for *Anopheles* vectors of malaria, and (D) the only four locations, we are aware of, where estimates for the mean adult biting density, survival and human blood index for even a single vector, as well as human population size and infectiousness are all available, so that malaria transmission and control can be modelled in a site-specific manner based on local estimates of these parameters. Panels A and B respectively represent the species-specific subset of all insecticide susceptibility bioassay [45] and human blood index [7, 49] data collated by the time they were most recently published as dataset summary reports. The studies represented in Panel C include only three reports of species-specific estimates [46–48] for the proportion of human biting exposure occurring indoors, with the remaining handful all relating to undifferentiated mixtures of species in a complex or group [34, 50, 51]. Panel D represents only four locations in only three countries (Nigeria, Tanzania, Papua New Guinea), all of which were small rural villages with intense transmission and anthropophagic vectors (otherwise at least some of these parameters would probably have been impossible to measure), but nevertheless yielded remarkably different vector-parasite demography patterns and suitability for various complementary vector

control measures beyond LLINs and IRS [26, 52].

408 Towards Malaria Elimination - A Leap Forward

The latest guidelines from WHO for entomological surveillance, monitoring and evaluation offer an excellent new framework for comprehensively applying existing field and analytical methods, and for conducting operational research into improving their use practices in the future [53]. In order for national programmes to make evidence-based decisions about what vector control measures to deploy, and evaluate their ongoing impact, the remarkably diverse arsenal of entomological methods already at our disposal [54] now need to be adapted to programmatic surveillance platforms that are nationally representative of all major ecological and epidemiological strata in the country [6–8, 11]. Such programmatic platforms should emphasize the absolute minimum number of essential entomological metrics, with strong data quality control and assurance processes to maximize confidence and minimize ambiguity of interpretation (**Figure 4**).

Data quantity and quality obviously trade off against each other, especially when working across very large geographic areas, so it is often necessary to select the smallest number of surveillance sites required to adequately represent all major epidemiological and ecological strata in the country. Based on our experience, a minimum of five surveillance sites per stratum is suggested to adequately capture variation within each stratum. The recent WHO guideline of one sentinel site per million people [53] also represents a good benchmark that

operational coverage gaps as soon as they arise. Entomological monitoring to inform daily operations of such area-wide insecticide applications therefore need to be repeated on a correspondingly regular basis, and on geographic scales approximately matched to the scale at which insecticide application succeeds or fails. While active larval searches can be useful for monitoring the effectiveness of insecticide applications, they are prone to large observational biases, especially towards over-representing the most obvious and accessible habitats most likely to be effectively covered. Adult mosquitoes, however, fly and gravitate towards stimuli so they can be passively monitored with traps. Adult mosquito density measurements are therefore a far more reliable way to verify impact and inform operational implementation of larviciding or space-spraying programmes that actively deliver relatively short-lived insecticides across large areas. For such pro-active area-wide interventions, much higher spatial and temporal resolution is required than for more passive human-targeted approaches like LLINs and IRS, which rely on mosquitoes being attracted to protected individuals and households. National surveillance platforms need to not only monitor the most useful predictors and indicators of successful vector control, but also the products and delivery processes that are essential to achieving impact in practice. Biologically-rational selection of an optimal overall intervention approach, such as LLINs, IRS or larviciding, does not in itself guarantee success. First, it is essential that the most efficacious products within that class are procured, and that they are then effectively delivered. WHO pre-qualification and centralized procurement help target investments towards reliable products. Subsequently, laboratory-based efficacy testing of products sampled from various stages of the supply chain are an invaluable means to quality assure, and even quality control, product supplies. However, beyond successful procurement and supply, it is also essential to ensure that vector control products are satisfactorily delivered and remain effective over the lifetimes required of them. While the coverage of areawide, frequently re-applied products like larvicides can be difficult to quantify objectively [55, 56], coverage indicators for human-targeted measures like LLINs and IRS can be readily incorporated into questionnaires for routine epidemiological surveys. Reliable, standardized methodology has long been available for measuring the extent and durability insecticidal activity of walls, roofs and ceilings treated with IRS, and have been applied to great effect across multi-country scales, to demonstrate programmatically-relevant variations in product performance [57]. Standardized methods for assessing the physical and insecticidal durability of LLINs reveal similarly important variations in performance, including some notable shortfalls relative to the requirements for recommendation of a '3-year net' [58]. Regardless of how well they are developed, manufactured and tested before they are delivered, products do not always 'do what they say on the tin', so their effectiveness in the field needs to be regularly

Entomological Surveillance as a Cornerstone of Malaria Elimination: A Critical Appraisal

http://dx.doi.org/10.5772/intechopen.78007

411

assessed and re-assessed on nationally representative scales.

**and intervention impact**

**5. Directly interpretable entomological metrics of mosquito** 

**behaviours, insecticide susceptibility, intervention effectiveness,** 

Entomological monitoring to inform routine programme implementation needs to yield measurements that can be directly and informatively interpreted, so that those collecting the

**Figure 4.** A suggested flow diagram for cyclical collection and assessment of programmatic surveillance data to inform malaria vector control on national scales. Formal annual review cycles may be supplemented by *ad hoc* review meetings at short notice, whenever surprizing or alarming trajectories are observed through ongoing data monitoring.

can be modified according to need by countries that are particularly large, small or diverse. In addition to routine surveys of established surveillance sites, *ad hoc* spot checks and focus investigations are also recommended as ways to further improve vector surveillance and control [53]. Additionally, much more intensive, finer-scale surveillance of vector population dynamics is required wherever pro-active mosquito abatement methods, specifically larvicide application or space spraying, are deployed. These vertically-managed methods for delivering insecticides across large areas need to be repeated on a regular basis, often as frequently as every week. In order for mosquito population density measurements to be useful for monitoring purposes, they need be collected at spatial scales fine enough to identify operational coverage gaps as soon as they arise. Entomological monitoring to inform daily operations of such area-wide insecticide applications therefore need to be repeated on a correspondingly regular basis, and on geographic scales approximately matched to the scale at which insecticide application succeeds or fails. While active larval searches can be useful for monitoring the effectiveness of insecticide applications, they are prone to large observational biases, especially towards over-representing the most obvious and accessible habitats most likely to be effectively covered. Adult mosquitoes, however, fly and gravitate towards stimuli so they can be passively monitored with traps. Adult mosquito density measurements are therefore a far more reliable way to verify impact and inform operational implementation of larviciding or space-spraying programmes that actively deliver relatively short-lived insecticides across large areas. For such pro-active area-wide interventions, much higher spatial and temporal resolution is required than for more passive human-targeted approaches like LLINs and IRS, which rely on mosquitoes being attracted to protected individuals and households.

National surveillance platforms need to not only monitor the most useful predictors and indicators of successful vector control, but also the products and delivery processes that are essential to achieving impact in practice. Biologically-rational selection of an optimal overall intervention approach, such as LLINs, IRS or larviciding, does not in itself guarantee success. First, it is essential that the most efficacious products within that class are procured, and that they are then effectively delivered. WHO pre-qualification and centralized procurement help target investments towards reliable products. Subsequently, laboratory-based efficacy testing of products sampled from various stages of the supply chain are an invaluable means to quality assure, and even quality control, product supplies. However, beyond successful procurement and supply, it is also essential to ensure that vector control products are satisfactorily delivered and remain effective over the lifetimes required of them. While the coverage of areawide, frequently re-applied products like larvicides can be difficult to quantify objectively [55, 56], coverage indicators for human-targeted measures like LLINs and IRS can be readily incorporated into questionnaires for routine epidemiological surveys. Reliable, standardized methodology has long been available for measuring the extent and durability insecticidal activity of walls, roofs and ceilings treated with IRS, and have been applied to great effect across multi-country scales, to demonstrate programmatically-relevant variations in product performance [57]. Standardized methods for assessing the physical and insecticidal durability of LLINs reveal similarly important variations in performance, including some notable shortfalls relative to the requirements for recommendation of a '3-year net' [58]. Regardless of how well they are developed, manufactured and tested before they are delivered, products do not always 'do what they say on the tin', so their effectiveness in the field needs to be regularly assessed and re-assessed on nationally representative scales.
