Methods and indicators for measuring patterns of human exposure to malaria vectors

Abstract

Background: Effective targeting and evaluation of interventions that protect against adult malaria vectors requires an understanding of how gaps in personal protection arise. An improved understanding of human and mosquito behaviour, and how they overlap in time and space, is critical to estimating the impact of insecticide-treated nets (ITNs) and determining when and where supplemental personal protection tools are needed. Methods for weighting estimates of human exposure to biting Anopheles mosquitoes according to where people spend their time were first developed over half a century ago. However, crude indoor and outdoor biting rates are still commonly interpreted as indicative of human-vector contact patterns without any adjustment for human behaviour or the personal protection effects of ITNs.

Main text: A small number of human behavioural variables capturing the distribution of human populations indoors and outdoors, whether they are awake or asleep, and if and when they use an ITN over the course of the night, can enable a more accurate representation of human biting exposure patterns. However, to date no clear guidance is available on what data should be collected, what indicators should be reported, or how they should be calculated. This article presents an integrated perspective on relevant indicators of human-vector interactions, the critical entomological and human behavioural data elements required to quantify human-vector interactions, and recommendations for collecting and analysing such data.

Conclusions: If collected and used consistently, this information can contribute to an improved understanding of how malaria transmission persists in the context of current intervention tools, how exposure patterns may change as new vector control tools are introduced, and the potential impact and limitations of these tools. This article is intended to consolidate understanding around work on this topic to date and provide a consistent framework for building upon it. Additional work is needed to address remaining questions, including further development and validation of methods for entomological and human behavioural data collection and analysis.

Keywords: Exposure; Human-vector contact; Human-vector interaction; Insecticide-treated nets; Outdoor biting; Outdoor transmission; Residual malaria transmission.

Fig. 1

Example of directly measured and behaviour-adjusted estimates of human exposure to malaria vectors from Asembo, western Kenya in 2011. Mosquito biting data was collected using human landing catches from June through July 2011 and human behavioural data was collected using a cross-sectional survey conducted from July through August 2011 [52]. Series A shows the proportion of the human population (1) outdoors (2) indoors and awake, and (3) indoors and asleep throughout the night, overlaid with directly measured indoor and outdoor biting rates for Anopheles arabiensis. Based on biting density alone, the estimated percentage of vector bites occurring indoors = 63%. Series B integrates vector and human behaviour data to show behaviour-adjusted cumulative exposure to vector bites for an unprotected individual. The percentage of vector bites occurring indoors for an unprotected individual $\left({\pi }_{I,u}\right)$ = 97% and the percentage occurring while asleep indoors for an unprotected individual $\left({\pi }_{S,u}\right)$ = 84%

Fig. 2

Example of indicators calculated using vector and human behaviour data from Asembo, western Kenya in 2011 [52]. Series A shows the behaviour-adjusted estimates of exposure to Anopheles arabiensis bites for an unprotected individual. Series B shows behaviour-adjusted estimates of exposure to vector bites for an ITN user. ITNs were assumed to prevent approximately 94% of bites while in use based on reference estimates from experimental hut trials. The percentage of all vector bites prevented by using an ITN $\left(P_S^{\ast }\right)$ = 79% and the proportion of remaining exposure occurring indoors for a protected user of an ITN $\left({\pi }_{I,p}\right)$ = 87%. Series C shows the population-wide mean exposure to vector bites. In this site the proportion of the population that reported using an ITN while asleep the previous night was 74% (arrows). Therefore, the population-wide mean personal protection against biting exposure given the reported community-level coverage of people using an ITN $\left(P_{S,C}^{\ast }\right)$ is 58%. An Excel file demonstrating how these indicators were calculated is included as Additional file 1