Standardizing operational vector sampling techniques for measuring malaria transmission intensity: evaluation of six mosquito collection methods in western Kenya

Abstract

BackgroundOperational vector sampling methods lack standardization, making quantitative comparisons of malaria transmission across different settings difficult. Human landing catch (HLC) is considered the research gold standard for measuring human-mosquito contact, but is unsuitable for large-scale sampling. This study assessed mosquito catch rates of CDC light trap (CDC-LT), Ifakara tent trap (ITT), window exit trap (WET), pot resting trap (PRT), and box resting trap (BRT) relative to HLC in western Kenya to 1) identify appropriate methods for operational sampling in this region, and 2) contribute to a larger, overarching project comparing standardized evaluations of vector trapping methods across multiple countries.MethodsMosquitoes were collected from June to July 2009 in four districts: Rarieda, Kisumu West, Nyando, and Rachuonyo. In each district, all trapping methods were rotated 10 times through three houses in a 3 × 3 Latin Square design. Anophelines were identified by morphology and females classified as fed or non-fed. Anopheles gambiae s.l. were further identified as Anopheles gambiae s.s. or Anopheles arabiensis by PCR. Relative catch rates were estimated by negative binomial regression.ResultsWhen data were pooled across all four districts, catch rates (relative to HLC indoor) for An. gambiae s.l (95.6% An. arabiensis, 4.4% An. gambiae s.s) were high for HLC outdoor (RR = 1.01), CDC-LT (RR = 1.18), and ITT (RR = 1.39); moderate for WET (RR = 0.52) and PRT outdoor (RR = 0.32); and low for all remaining types of resting traps (PRT indoor, BRT indoor, and BRT outdoor; RR < 0.08 for all). For Anopheles funestus, relative catch rates were high for ITT (RR = 1.21); moderate for HLC outdoor (RR = 0.47), CDC-LT (RR = 0.69), and WET (RR = 0.49); and low for all resting traps (RR < 0.02 for all). At finer geographic scales, however, efficacy of each trap type varied from district to district.ConclusionsITT, CDC-LT, and WET appear to be effective methods for large-scale vector sampling in western Kenya. Ultimately, choice of collection method for operational surveillance should be driven by trap efficacy and scalability, rather than fine-scale precision with respect to HLC. When compared with recent, similar trap evaluations in Tanzania and Zambia, these data suggest that traps which actively lure host-seeking females will be most useful for surveillance in the face of declining vector densities.

Reemergence of Anopheles funestus as a Vector of Plasmodium falciparum in Western Kenya after Long-Term Implementation of Insecticide-Treated Bed Nets

“…was 0.76 (95% CI: 0.61-0.96). 25 These relative collection rates were used to calibrate the 2002-2008 LT to HLC for comparison to the 2011 HLC ( Figure 4B). …”

“…The traps were set in 30-60 houses each month and run for two consecutive nights in each house. Collection rates of LT relative to HLC from Wong and colleagues 25 were used to compare the LT data from 2002 through 2008 to the HLC data from 2011. Collection rates of BNT relative to HLC, described previously, were used to compare the BNT data from 1993 through 1997 to the HLC data from 2011.…”

“…Although LT undersampled An. funestus relative to HLC, 25 the different sensitivities of these collection methods are unlikely to account for the differences observed between these years. Thus, the data strongly suggest that An.…”

“…Calibrating the relative rate of other methods to HLC is limited in part by the population targeted by the sampling method (e.g., hostseeking versus blood-fed females), but the three sampling methods used in this study to compare Anopheles female abundances across time (BNT, LT, and HLC) all target hostseeking females at night. Calibration of other methods to HLC is also limited by spatial variation in mosquito populations' response to different sampling methods, even among districts in western Kenya, 25 and possibly by temporal changes in mosquito behavior. However, all of our collections were done in the Asembo area of western Kenya, and we limited our comparisons to the months of the year for which we had data collected by all three collection methods.…”

Novel serologic biomarkers provide accurate estimates of recent Plasmodium falciparum exposure for individuals and communities

“…Existing metrics have varying intrinsic levels of precision and accuracy and are subject to a variety of extrinsic factors, such as cost, time, and availability of trained personnel (19). For example, entomological measurements provide information on mosquito to human transmission for a community but are expensive, require specially trained staff, and lack standardized procedures, all of which reduce precision and/or make interpretation difficult (19)(20)(21)(22). Parasite prevalence can be measured by detecting parasites in the blood of individuals from a cross-sectional sample of a community and is, therefore, relatively simple and inexpensive to perform, but results may be imprecise, especially in areas of low transmission (19,23), and biased by a number of factors, including immunity and access to antimalarial treatment (5,6,19,(23)(24)(25).…”