Supplementary appendix
This append ix formed part of the original submission and has been peer reviewed. We post it as supplied by the authors.
Supplement to: Pinder M, Jawara M, Jarju LBS, et al. Efficacy of indoor residual spraying with dichlorodiphenyltrichloroethane against malaria in Gambian communities
with high usage of long-lasting incticidal mosquito nets: a cluster-randomid controlled trial. Lancet 2014; published online Dec 9. dx.doi/10.1016/S0140-6736(14)61007-2.
Web Extra material
Supplementary material
Here we describe in more detail the study area, method of randomisation, collection of incidence data, finalising the data and the major entomological findings.
Study area and participants
The study was carried out in the Upper River Region, the far eastern region of The Gambia, and was bad in the MRC Unit’s field station in Bas (13.3167o N, -14.2167o W). This is a rural area of open Sudanian savannah with a single rainy ason from May to October, followed by a long dry ason. Malaria is highly asonal with most malaria episodes experienced during or immediately following the rainy ason; rainfall was above average in 2010 (1116 mm), and about average in 2011 (890 mm). This region is bicted by the river into the north and south banks. Houhold demographic data including bed net usage and hou construction was collected before enrolment of the children and the rollout of the interventions. Almost half the residents enrolled into the study lived in hous with thatched roofs (49.31%, 1685/3417), the remainder were metal, and the most common inner wall surfaces was bare mud (50.48%, 1725/3417) and matt paint (41.41%, 1415/3417). Randomisation
见人见事
The 97 villages were grouped into 70 clusters and the clusters were initially stratified into small and large bad on the median of their population, as larger clusters may favour the effectiveness of indoor residual spraying (IRS). The were then further divided into the four geographical areas, two on each bank of the river to allow for differences in rainfall from east to west and distance from Bas, the main centre of trade and development. Each of the clusters in each category of size and l
裤子搭配
ocation was then randomly assigned to IRS and long-lasting incticidal nets (LLINs) or LLINs alone, assuming equal allocation to each arm and constraining the allocation to 35 clusters in each arm by DJ. A Matlab (R2010, Math Works, Natick, Massachutts, U.S.A.) program was then ud to randomly repeat this allocation 100,000 times. The balance of each allocation was defined as the absolute difference between the total population in each arm. Fifty random allocations with the lowest population difference of five, were then randomly lected and assigned numbers 1 to 50. An independent obrver, who did not have access to the randomisations and was not involved in the study, then randomly lected a number from 1 to 50 and the corresponding cluster allocation was ud and a total of 35 clusters (47 villages) randomized to receive IRS with DDT, and 35 clusters (49 villages) receiving only LLIN. Balanced randomization was ud to enrol children of similar ages in each cluster with the target number enrolled increasing with village size (population <500: 75-115 enrolled, 500-1499:120-170, 1500 - 2258:180-190). Entomological sampling was conducted in a sub-t of clusters chon from tho nearer to the field station for logistical reasons and the houholds within each cluster were randomly lected by MP.
Obrver bias was reduced where feasible. Slide microscopists and their supervisors were blinded to the identity and intervention status of the subjects. Mosquito collector bias was reduced by using s
tandardized traps, which do not rely on the ability of the fieldworker to collect specimens. Trap catches were examined by a different person to the trap collector and blinded to the trap location. Apart from data on IRS, no data forms or samples carried the group allocation and this was only added to the datats after final cleaning.
Clinical evaluations
Twelve field assistants/nurs were posted to nine health clinics and three key health posts where they were responsible for working with government staff/village health workers to record cas of malaria in the cohort. Each staff member was responsible for five to eight clusters and their health posts, visiting the health posts at least once a week to collect data on cas and replenish supplies.
Entomology
Sample size rationale
Demonstrating a 60% reduction in hou entering mosquitoes (Anopheles gambiae) associated with IRS-LLIN, with 90% power and at the 5% level of significance, required six hous in each cluster and 16 clusters in each arm of the trial over two years 1.
平行四边形练习题
Incticide persistence
Persistence of incticides on walls was measured using World Health Organisation (WHO) cone tests (WHO, 2006) in six hous, stratified by wall surface at one, three and six weeks post-IRS using triplicate tests with an
average of 21 DDT susceptible A gambiae s.s. M form per test 2. Persistence of incticide on LLIN was determined in triplicate using WHO cone tests on six LLIN from a randomly lected cluster in each arm after they had been in u for 16 months.
Field collections
Mosquito exposure was measured using standardized light and exit traps, which respectively estimate indoor-resting and exiting mosquitoes. Six ntinel rooms in 32 clusters, where a connting adult slept under a bednet, were sampled monthly in both transmission asons. Potential risk factors known to affect mosquito densities in The Gambia 3 were recorded at each collection. Mosquitoes were killed by freezing before morphological identification by standard keys 4,5 and unfed and blood-fed mosquitoes were discted to determine whether they were parous. All female A gambiae s.l. were procesd for species determination using PCR 6 and Plasmodium falci
parum infection using an enzyme-linked immunosorbent assay 7.
Incticide susceptibility
Wild caught A gambiae s.l. larvae from Area 3 in 2010 and Area 1, 2, and 4 in 2011 were reared to adults and their susceptibility to permethrin and DDT was assd using WHO tube bioassays 8.
Statistical analysis
Differences in malaria transmission experienced in the two study arms were examined by comparing the proportion of sporozoite positive A gambiae s.l. The mean number of vector mosquitoes caught indoors in ntinel rooms adjusted for clustering was also examined. Differences in number of mosquitoes caught in light-traps between study arms was estimated by multilevel analysis using a mixed effect model on square root transformed data with cluster as a random effect and the intervention and covariates as fixed effects. The sporozoite rate, with a 95% confidence interval, was estimated for each arm of the study and estimates of entomological inoculation rate (EIR) were calculated as the mean number of A gambiae s.l. /hou/night multiplied by the sporozoite rate and the number of nights during the entomological survey period. The proportion of sporozoite infected mosquitoes was compared between the intervention arms using logistic regression. The effect of IRS
蛋白质的水解with DDT on mosquitoes leaving sleeping rooms was quantified as the percentage of A gambiae s.l. caught in exit traps among the total mosquitoes caught in both traps using data from both asons and the intervention arms were compared using Wilcoxon’s rank test.
Results
Incticide persistence
学生装扮>年华似流水Residual activity of DDT, estimated by WHO cone tests in 2011, was high with 99.2% mortality (95% CI 97.2 to 100) week one post-IRS and 94.3% (95% CI 89.3 to 99.3) after six weeks. Estimations of DDT residual activity in a non-study village within the Upper River Region in 2011 using the same batch of DDT showed high levels five months post-IRS on both mud and matt painted walls, mean mortality of 92.5 and 94.7%, respectively 25. The residual activity of permethrin estimated by WHO cone tests at 16 months post-donation was high with 91.0% mortality (95% CI 87.8 - 93.6) in a LLIN cluster and 89.3% (95% CI 85.8- 92.2) in an IRS-LLIN cluster; mortality on new LLIN was 92.8% (90.9 -94.5) and on an untreated net was 10.4% (7.4 - 14.1).
Field collections
All entomological collections were successful apart from one where the hou was locked. All successful catches had covariate data. A gambiae s.l. were prent in 36.4% of light traps (839/2303) and 9.0% of exit traps (207/2303). Over 94%, of mosquitoes were collected in light traps, the remainder from exit traps, and overall 37.1% (10,601/28,607) of tho collected were anophelines of which 72.3% (7664/10601) were A gambiae s.l.; all the rest were culicines. Over 99% of the A gambiae s.l. were identified to species, 70.7% (5372/7596) were A arabiensis and the rest A gambiae s.s.
Densities of anophelines, A gambiae s.l. and the sibling species of this complex varied by year and were slightly lower in the intervention arm (table 1), but there was no significant difference between the intervention arms (linear regression allowing for clustering, Year 1, p=0.299; Year 2, p =0.341). The proportion of sporozoite positive A gambiae s.l. was low (table 1) with a significant interaction between study arms and year (logistic regression, P=0.039) but within year there was no significant difference between the arms (sporozoite rate difference Year 1= 0.0013, P=0.38; Year 2 = - 0.0056, P=0.06). There was also no
significant difference in EIR between the two arms of the study (table 1). The influence of covariates on the primary entomological outcome (EIR) could not be examined due to the low sporozoite infecti
on rates but their influence on mosquito catch size was possible. Linear regression on numbers of A gambiae s.l. caught in light traps including year, the prence of open eaves, a tethered hor, matt painted walls and more than one sleeper in the room, showed no significant difference (P=0.281) between the study arms (adjusted mean caught over two years was 6.7, 95% CI 4.0 - 10.1, in the LLIN arm and 4.5, 95% CI 2.4 – 7.4, in the IRS-LLIN arm). IRS with DDT did not significantly influence the proportion of A gambiae s.l. leaving hous. The mean proportion exiting in 2010 was 12.81 (95% CI 9.47 – 16.15) in the LLIN arm and 10.34 (95% CI 6.69-14.00) in the IRS-LLIN arm, in 2011 the values were 9.62 (95% CI 4.56-14.68) and 6.24 (95% CI 1.58-10.92), respectively. The mean proportion exiting over both asons in the LLIN arm was 11.66 (95% CI 9.21-14.12) and in the IRS- arm this was 8.80 (95% CI 6.00-11.81; P=0.087 Wilcoxon rank test). In both arms of the study, the parity rate among A gambiae s.l. caught in light traps was high (table 1) and without statistical difference between the arms (linear regression allowing for clustering P=0.779).
Incticide susceptibility
To examine the susceptibility of the local vectors to the study incticides, in area 4 in 2010 and in areas 1, 2 and 3 in 2011, A gambiae s.l. larvae collected from breeding sites clo to enrolled villages were raid to adults and expod to papers impregnated with permethrin, DDT or solvent a
t an average of 21.3 mosquitoes / tube. In both years, and from all sites, mortality in the controls tubes was always less than 4%. In 2010, in area 4, mortality to DDT and permethrin was 100% (97.5% CI 89.7-100), but few A gambiae s.l. were examined (table 2). In 2011, in areas 1-3, larger numbers of mosquitoes were examined and mortality rates to DDT and permethrin were less than 100% (mean range 88.3-94.8%, table 2).
职业管理
Discussion
Our results support the notion that the DDT sprayed on the walls and permethrin on the nets were at a concentration which provided high killing rates in expod mosquitoes. The residual activity of DDT lasted for at least 5 months, to provide protection over the rainy ason, and for permethrin the residual activity was at least 16 months.
A subt of 32 clusters was sampled for the entomological endpoints as this was sufficient to detect a 60% reduction in hou entering mosquitoes. The entomology clusters were also well balanced for cluster size and distribution over the study area and also had proportional more Fula. Over both years of the study, there were slightly fewer A gambiae s.l. entering hous in the IRS-LLIN arm, but the differences were not statistically significant either with unadjusted and adjusted analys. This,
together with similar entomological inoculation rates and the long-lived vector population, indicated by the high parity rates in both study arms, supports the clinical data and the conclusion that IRS with DDT offered no additional protection in the prence of high LLIN coverage.
There is growing evidence that malaria vector control in sub-Saharan Africa is threatened by the spread of incticide resistance in mosquitoes , both against the pyrethroids ud for treating bednets 9 and all class of incticide ud for IRS 10. Twenty five years ago in The Gambia, shortly after permethrin-treated nets were introduced, little or no resistance to either DDT or permethrin was detected 11. In 2008, two years before the prent study, no resistance to either DDT (100.0% mortality, CI 82.4-100%) nor permethrin (100.0% mortality, CI 84.6-100%) was found in samples from around Bas town, located in the centre of the current study area, although only 19-22 mosquitoes were tested for each incticide 12. In 2010 we found similar results with adult A gambiae s.l. raid from larvae caught near study clusters east of Bas town (area 4). In 2011, larvae were caught in study areas 1-3, and the tube test results indicate low-level resistance to both incticides ud, with mean mortality of 88%, 89% and 91% to DDT and 93%, 93% and 94% to permethrin. In a pilot study which examined the possibility of using alternative incticides to DDT for IRS, larvae were collected in two villages in study area 4 which were not enrolled in the prent stud
y. Here we found high levels of resistance to DDT and permethrin 2. Overall the mosquitoes in our study were susceptible to both permethrin and DDT, with a few anomalous areas of high resistance.
高谈阔论造句References
1. Hayes RJ, Bennett S. Simple sample size calculation for cluster-randomized trials. Int J Epidemiol
1999; 28: 319-26.
2. Tangena J-AA, Majida A, D’Alessandro U, et al. Alternative treatments for indoor residual spraying
for malaria control in a village with pyrethroid- and DDT-resistant vectors in The Gambia. PLoS ONE 2013; 8: e74351.
3. Kirby MJ, Green C, Milligan PM, et al. Risk factors for hou-entry by malaria vectors in a rural town
and satellite villages in The Gambia. Malaria J 2008; 7: e2.
4. Gillies MT, Coetzee M. A supplement to the Anophelinae of Africa South of the Sahara. Johannesburg:
The South African Institute for Medical Rearch; 1987.
5. Gillies MT, DeMeillon B. The Anophelinae of Africa south of the Sahara (Ethiopian zoogeographical
region). 1968.
6. Scott JA, Brogdon WG, Collins FH. Identification of single specimens of the Anopheles gambiae
complex by the polymera chain reaction. Am J Trop Med Hyg 1993; 49: 520-9.
7. Wirtz RA, Duncan JF, Njelesani EK, et al. ELISA method for detecting Plasmodium falciparum
circumsporozoite antibody. Bull World Health Organ 1989; 67: 535 - 42.
8. WHO. Guidelines for testing mosquito adulticides for indoor residual spraying and treatment of
mosquito nets. WHO/CDS/WHOPES/GCDPP/2006.3 ed. Geneva: WHO; 2006.
9. Ranson H, N'Guessan R, LInes J, Moiroux N, Nkuni Z, Corbel V. Pyrethroid resistance in African
anopheline mosquitoes: what are the implications for malaria control? Trends Parasitol 2011; 27: 91-8.
10. Edi C, Koudou B, Jones C, Weetman D, H R. Multiple-incticide resistance in Anopheles gambiae
mosquitoes, Southern Côte d'Ivoire. Emerg Infect Dis 2012 18: 1508-11.
11. Hemingway J. Efficacy of etofenprox against incticide susceptible and resistant mosquito strains
containing characterized resistance mechanisms. Med Vet Entomol 1995; 9: 423-6.
12. Betson M, Jawara M, Awolola T. Status of incticide susceptibility in Anopheles gambiae s.l. from
malaria surveillance sites in The Gambia. Malaria J 2009 8: e187.
Table 1. Entomological characteristics by study arm
are prented with the total infected mosquitoes / total assayed
a) Each arm and year there were 576 trapping collections, except in the IRS-LLIN in year 1 where there were 575.
b) mean differences are prented for all variables but for the percentage of A gambiae s.l. with sporozoites risk and the entomological inoculation rate where risk ratio and rate ratio is given, respectively. All p values were >0.05
