A global analysis of National Malaria Control Programme vector surveillance by elimination and control status in 2018

Thirty-four malaria endemic countries and one country recently certified as malaria-free participated in the survey [Africa (n = 18), Asia–Pacific (n = 14) and the Americas (n = 3)] between 1 November 2017 and 19 November 2018 (Fig. 1). Most surveys were completed by an individual for each country and the responses reported may have varied if completed by a different individual or by a group.

Data on active monitoring by NMCPs of the eight WHO recommended priority vector indicators [15] were documented (Fig. 2). The number of vector indicators that programmes monitored ranged from none to all eight with 86% of country NMCPs monitoring at least one indicator. Due to the limited number of participating countries in the Americas. Regional analyses focused on country programmes in Africa and the Asia-Pacific where 97% of malaria cases occurred in 2017 [1].

Overall, 91% of participating countries distribute LLINs, 31% implement IRS and 41% practice LSM as part of the NMCP strategies (Fig. 3). Integrated vector control was part of the national strategy in 58% of countries, most frequently manifested as sympatric distributions of LLINs with IRS in areas with higher malaria rates. Statistical differences in vector intervention use by countries in the Africa and Asia-Pacific regions were not found (χ² = 2.078, df = 3, p = 0.556).

Long-lasting insecticide-treated nets (LLINs) were distributed through focal responses and/or mass campaigns in the surveyed countries. Insecticides used in IRS differed significantly between the regions (χ² = 11.51, df = 3, p = 0.009). In Africa, all four main insecticide classes were widely used [64% of countries used organophosphates, followed by pyrethroids (50%), carbamates (43%), and the organochlorine, DDT (21%)]. Pyrethroids were used in IRS by 100% of Asia-Pacific responding countries implementing IRS. Regional differences in LSM methods were not found (χ² = 2.719, df = 3, p = 0.437) with larviciding being the dominant strategy, usually with an organophosphate (e.g., Temephos).

Other vector control methods were less frequently deployed (in 28% of countries) (Fig. 3). There was a greater tendency for NMCPs of the Asia-Pacific countries to incorporate other vector control interventions (Africa = 17%, Asia-Pacific = 43%; χ² = 1.302, df = 1, p = 0.254) (e.g., outdoor space spraying, topical repellents, hammock nets, coils and mosquito proof housing).

Regional differences in ITN usage and coverage indicators were found (χ² = 14.29, df = 6, p = 0.026; Fig. 4). In Africa, 94% of responding countries used the indicators “the proportion of the population having slept the previous night under an ITN/LLIN”), and “the proportion of households with at least one ITN/LLIN”. In the Asia-Pacific there was more diversity in indicators used: the proportion of households with at least one ITN/LLIN and the proportion of targeted risk groups receiving ITNs was used in 54% of countries while the proportion of the population with access to an ITN/LLIN within their household was used in 46% of countries.

Regional differences in IRS coverage indicators were not found (χ² = 2.53, df = 4, p = 0.639; Fig. 4); the proportion of structures or households sprayed with IRS was the most commonly used indicator of IRS coverage (Africa = 93%, Asia-Pacific = 67%). Other common indicators were the proportion of the population at risk protected by IRS (Africa = 71%, Asia-Pacific = 56%) and the proportion of target risk groups receiving IRS (Africa = 64%, Asia-Pacific = 56%).

Monitoring of LSM was undertaken by 75% of countries, usually as follow up surveys for the presence of larvae in treated habitats with survey frequency dependant on available resources.

Types of surveys to monitor LLIN usage and coverage or IRS coverage did not differ between Africa and Asia-Pacific countries (LLIN: χ² = 8.89, df = 5, p = 0.113; IRS: χ² = 3.262, df = 4, p = 0.515; Fig. 5). For LLIN use and coverage, countries in Africa tended to use the Malaria Indicator Survey (88%) or cross-sectional demographic and health surveys (63%), while Asian-Pacific countries used data collected as routine programme data (46%). For IRS, countries in both regions favoured routine programmatic data to determine coverage (Africa = 64%, Asia-Pacific = 78%).

Statistical differences in the survey frequency to monitor ITN use and coverage or IRS coverage between African and Asian-Pacific countries were not found (LLIN: χ² = 5.53, df = 5, p = 0.35; Additional file 1: Figure S1; IRS: χ² = 8.921, df = 6, p = 0.178; Additional file 1: Figure S2). Annual surveys for ITN and IRS coverage were most frequently conducted (ITN: Africa = 33%, Asia-Pacific = 31%; IRS: Africa = 62%, Asia-Pacific = 38%).

Most countries distributing LLINs assessed physical durability (Africa = 80%, Asia-Pacific = 69%) and the effective life of the nets (Africa = 60%, Asia-Pacific = 85%). The proportion of countries assessing durability and effective life did not vary significantly between Africa and Asia-Pacific countries (physical durability: χ² = 0.05, df = 1, p = 0.826; effective life: χ² = 1.04, df = 1, p = 0.308).

Most countries conducting IRS assessed the initial efficacy (Africa = 86%, Asia-Pacific = 89%) and the effective life (Africa = 93%, Asia-Pacific = 100%) of insecticides using cone tests or wall bioassays. Significant differences between the Africa and Asia-Pacific countries in the proportion of countries assessing IRS efficacy were not found (initial efficacy: χ² < 0.001, df = 1, p = 0.99; effective life: χ² < 0.001, df = 1, p = 0.99).

Monitoring adult densities is a fundamental surveillance activity needed to determine seasonality, indoor/outdoor biting, peak biting time and to map spatial distributions. Post-collection analysis is required to determine three indicators: sporozoite rates, survivorship, and human blood indices. Fifty-six per cent of countries measured adult vector densities at least once a year, with no difference between the Africa and Asia-Pacific regions (χ² = 0.074, df = 1, p = 0.786). The main adult collection techniques were human landing catches (indoors or outdoors) and indoor resting collections in both regions (Fig. 7). Adult densities were measured frequently enough to determine seasonality in 41% of countries, with no difference between the Africa and Asia-Pacific regions (χ² = 0.211, df = 1, p = 0.645). Peak biting times were measured in 38% of countries; and was more frequently measured in the Asia-Pacific (50%) compared with countries in Africa (28%), but this difference was not significantly different (χ² = 0.846, df = 1, p = 0.357). Indoor/outdoor biting rates were determined in 34% of countries, with no difference between the Africa and Asia-Pacific regions (χ² = 0.054, df = 1, p = 0.814). The sporozoite rate was measured in 38% of countries, with no difference between the Africa and Asia-Pacific regions (χ² = 305, df = 1, p = 0.581).

Larval surveys were conducted at least annually in 47% of countries [no difference between the Africa and Asia–Pacific regions (χ² = 1.914, df = 1, p = 0.166)].

Eighty-six percent of countries in the Asia-Pacific knew the distributions of their vectors compared with 61% of countries in Africa. The geographic scale at which country vector distributions were known differed significantly between the Africa and Asia-Pacific regions (χ² = 11.73, df = 4, p = 0.019; Fig. 8). Asia-Pacific countries knew their vector distributions at a finer scale (36% to the village level) compared to countries in Africa. However, the methods of identifying vector species differed between the regions (χ² = 6.4, df = 1, p = 0.011). Morphology only based identifications of mosquitoes was used in 92% of countries in the Asia-Pacific, while 67% of countries in Africa confirm morphological identifications with molecular techniques.

Of the 35 participating countries, six were classified as “eliminating” based on their inclusion in the E2020 and one (Sri Lanka) was certified as malaria-free in 2016 (Table 1, Fig. 2). The remaining 28 countries were in the malaria control phase. Interventions deployed and management strategies were compared between eliminating and control countries at a global scale due to the small number of participating eliminating/eliminated countries. Large differences were noted in the proportion of control versus eliminating countries using various vector control strategies. While all eliminating countries implemented IRS, only 57% of eliminating countries distributed LLINs. Conversely, all countries in the control phase distributed LLINs but only 61% included IRS in their national malaria control strategies. The difference in the proportion of control versus eliminating countries deploying LLINs was statistically significant (χ² = 8.226, df = 1, p = 0.004). Larval source management was almost twice as likely to be reported being used in eliminating compared to countries controlling malaria while non-recommended interventions were more than threefold more likely to be used in eliminating countries and this difference was statistically significant for the alternative (non-WHO recommended) interventions (χ² = 5.469, df = 1, p = 0.019).

Significant differences between the proportion of countries in the control and elimination phases that had strategies for IR and integrating multiple vector control strategies or that used vector data as part of their decision-making process were not found, although IR management strategies were more commonly reported in control compared to elimination countries (54% and 29%, respectively). Despite the WHO call for countries to have IR management plans and for implementing integrated vector management, only 49% of countries reported having an IR management plan with 66% deploying multiple sympatric vector control strategies.

Vector data was used in decision making by 64% of countries (Table 1). The most common techniques for storing data was using Excel in 70% (19 of 27) of responding countries or with paper-based methods as reported by 63% of countries (17 of 27 countries). Vector surveillance data was most frequently presented using Excel (38% of countries) followed by data dashboards (in 28% of countries). Spatial decision support systems were employed in 14% (3 of 21 countries).

Vector surveillance data informed seven categories of decisions (Table 2). The most commonly cited use for vector data was to stratify adult vector control strategies (24% of countries). The second and third most commonly identified roles for vector data was in selecting insecticides (18% of countries) and larval control strategies (12% of responding countries). Three countries each used data to define receptive areas and to identify sites for focus studies. While 69% of countries use IRS, data was used by 3% of countries to determine where IRS should be applied despite data on indoor/outdoor resting behaviours being collected in 34% of countries. Only one country used data on LLIN durability to select LLINs for distribution despite 91% of surveyed countries distributing ITNs as part of their national malaria control strategy. While 77% of countries monitor for IR, only 18% consider vector data when selecting insecticides for programmatic use. LSM is implemented in 43% of countries but only 12% select larval control strategies based on larval survey data (and only 3% of countries test for IR in larvae). Differences between eliminating and control countries in how data were used were not seen.

The countries in the elimination phase monitored a mean of 4.7 vector surveillance indicators whereas countries in the control phase monitor a mean of 3.8 indicators (Table 3). The WHO guidelines stratify the importance of vector surveillance indicators by transmission intensity. In eliminating countries, it is recommended to track receptivity (a function of vector distributions and population size) and to define IR phenotypes. The seven eliminating countries monitor these three priority indicators with greater frequency than countries controlling malaria; for eliminating countries, 86% monitor IR phenotypes, 86% conduct larval surveys and 100% knew their vector distributions compared to 75%, 39% and 64% of countries controlling malaria, respectively. Of the six other prioritized vector surveillance indicators for high transmission countries, the only indicator that was measured in more than half of countries was annual adult densities (57%). For the remaining indicators, the frequency at which they were measured was 43% for IR mechanisms, 32% for sporozoite rates, indoor/outdoor biting and resting behaviours with only 11% measuring the human blood index annually.