Insecticide-treated bed nets in rural Bangladesh: their potential role in the visceral leishmaniasis elimination programme

Objective To analyse the feasibility, acceptability and effectiveness of insecticide-treated bed nets with slow-release insecticides (KO Tab 123) as an option for kala-azar vector management in Bangladesh. Methods Intervention study involving an insecticide dipping programme through village health workers supervised by public health officers covering 6967 households in Mymensingh and 8287 in Rajshahi district. In a subsample of households, sandfly densities at baseline, 1, 12 and 18 months were measured with CDC light traps both in intervention and control areas. Bioassays were performed for determining the bioavailability of the insecticide and tests of chemical residues in the treated bed nets were undertaken. Satisfaction surveys and direct observation of use of treated bed net use were conducted. Results The dipping programme was feasible with the help of communities and public health staff, was well accepted, reached a coverage of 98.2% and 96.2% in the two study sites within 4 weeks and was effective in terms of a significant reduction in sandfly densities (approximately 60%) for a period of 18 months. Bioassay results were satisfactory (>80% sandfly mortality) and the average chemical content of the treated bed nets was sufficient for killing sand flies at the end of the observation period. Conclusion Bed nets treated with slow-release insecticides can be an important complementary measure for sandfly control in the visceral leishmaniasis elimination programme. Moustiquaires imprégnées d’insecticide dans le Bangladesh rural: leur rôle potentiel dans le programme d’élimination de la leishmaniose viscérale Objectif: Analyser l’applicabilité, l’acceptabilité et l’efficacité des moustiquaires imprégnées d’insecticide à libération lente (KO Tab 123) en tant qu’option pour le contrôle du vecteur du kala-azar au Bangladesh. Méthodes: Etude d’intervention impliquant un programme d’imprégnation d’insecticide par le biais des agents de santé du village, supervisés par des agents de la santé publique, portant sur 6967 ménages à Mymensingh et 8287 dans le district de Rajshahi. Dans un sous-échantillon de ménages, les densités de phlébotomes au départ, à 1 mois, à 12 mois et à 18 mois ont été mesurées au moyen de pièges lumineux CDC à la fois dans les zones témoins et d’intervention. Des tests biologiques ont été effectués pour déterminer la biodisponibilité de l’insecticide et des tests sur les résidus chimiques dans les moustiquaires imprégnées ont été effectués. Des surveillances de satisfaction et l’observation directe de l’utilisation des moustiquaires traitées ont été menées. Résultats: Le programme d’imprégnation a été possible avec l’aide des communautés et du personnel de la santé publique, a été bien accepté, a atteint une couverture de 98,2% et 96,2% dans les deux sites d’étude endéans 4 semaines et a été efficace en termes d’une réduction significative de la densité des phlébotomes (∼60%) sur une période de 18 mois. Les résultats des essais biologiques ont été satisfaisants (> 80% de mortalité des phlébotomes) et la composition chimique moyenne des moustiquaires imprégnées a été suffisante pour tuer les phlébotomes à la fin de la période d’observation. Conclusion: Les moustiquaires traitées aux insecticides à libération lente peuvent être un moyen complémentaire important pour le contrôle des phlébotomes dans le programme d’élimination de la LV. Mosquiteras impregnadas con insecticida en Bangladesh rural: Su papel potencial en el programa de eliminación de la leishmaniasis visceral Objetivo: Analizar la factibilidad, aceptación y efectividad de las redes mosquiteras impregnadas con insecticidas de liberación lenta (KO Tab 123) como una opción para el manejo del vector de kala-azar en Bangladesh. Métodos: Estudio de intervención involucrando un programa de impregnación con insecticida realizado por trabajadores comunitarios supervisados por oficiales sanitarios y que cubrió 6967 hogares en Mymensingh y 8287 en el distrito de Rajshahi. En una submuestra de hogares se midieron las densidades de la mosca de arena al inicio del estudio, en los meses 1, 12 y 18 mediante trampas de luz CDC tanto en áreas de intervención como de control. Se realizaron bioensayos para determinar la biodisponibilidad del insecticida así como pruebas de residuos químicos en las mosquiteras impregnadas. Se llevaron a cabo encuestas de satisfacción y observación directa del uso de las mosquiteras impregnadas. Resultados: El programa de impregnación con la ayuda de trabajadores comunitarios y sanitarios del servicio público era factible, fue bien aceptado, alcanzó una cobertura del 98.2% y 96.2% en los dos lugares de estudio dentro de las 4 semanas y fue efectivo en términos de una reducción significativa en las densidades de moscas de arena (∼60%) durante un período de 18 meses. Los resultados de los bioensayos fueron satisfactorios (>80% de mortalidad de las moscas de arena) y el contenido promedio de químicos en las mosquiteras impregnadas era suficiente para matar las moscas de arena al final del período de observación. Conclusión: Las mosquiteras impregnadas con insecticidas de liberación lenta pueden ser una medida complementaria importante para el control de las moscas de arena en el programa de eliminación de la LV. Kala-azar (KA) or visceral leishmaniasis (VL), a potentially fatal parasitic disease transmitted by sand flies, is a serious public health problem in the Indian sub-continent (WHO, 2005a; Bhattacharia et al. 2006). In Bangladesh, where VL is largely a neglected disease, 34 of 64 districts are currently affected by VL (Bern et al. 2005; Bern & Chowdhury 2006; Mondal et al. 2008, 2009). Unique features of VL in Bangladesh, India and Nepal make the disease a candidate for elimination including: Phlebotomus argentipes being the only vector, humans are the only reservoir; there is a defined geographical area; availability of a highly sensitive and specific rapid diagnostic test (rK39) and there is an increasing number of treatment options (WHO 2005b; Sundar et al. 2010). This inspired national and international policymakers in 2005 to initiate a regional VL elimination initiative which includes Bangladesh, India and Nepal (WHO 2005a,b). Integrated vector management is one of the main strategic pillars of the VL elimination programme resting on indoor residual spraying (IRS) with insecticides, use long-lasting insecticide-treated bed net (LNs) and environmental vector management (WHO 2005a,b). Since 1985, there have been no vector control activities in Bangladesh because of the lack of human resources, logistics and funds (Bhattacharia et al. 2006; Mondal et al. 2008). Recent research has shown that IRS is most effective in reducing sandfly densities, but also that LNs were effective and to a lesser extent, lime plastering of walls (Joshi et al. 2009). A systematic literature review on the efficacy of LNs in reducing sandfly densities and VL incidence as well as recent intervention studies in India and Nepal have shown the potential of LNs in reducing sandfly vectors as an important complementary measure to IRS (Jalouk et al. 2007; Ostyn et al. 2008; Picado et al. 2010). Thus, bed net treatment with recently developed slow-release insecticides may be a cost-effective measure against VL vectors in places, where the use of un-impregnated bed nets against mosquitos is widespread. We aimed to investigate the feasibility, acceptability and effectiveness of a dipping programme of existing bed nets with slow-release insecticide tablets (KO Tab 123) in VL endemic areas of Bangladesh, where a 95% household coverage was found (Mondal et al. 2008). This was a non-randomized controlled trial and was a part of an operational research programme which included an evaluation of a community-based intervention with 18 months follow-up in intervention and control areas. The study sites were selected in the two most affected districts in Bangladesh based on past reports of high VL incidence: Mymensingh (Putijana union within the sub-district Fulbaria) and Rajshahi (Deopara union within the sub-district Godagari). The numbers of households in the intervention areas of Putijana and Deopara were 6967 and 31 280, and the numbers of inhabitants were 8287 and 35 442, respectively. The control areas for assessing sandfly densities were Gogram, Pakuria, Rishikul and Mohanpur unions in Rajshahi. The study was conducted from March 2008 to August 2009. The implementation of a dipping programme of the existing un-treated bed nets was performed through local public health workers and community volunteers: Trained field research assistants numbered all households and collected household information including the numbers of the existing bed nets by visiting each household in the study areas at the beginning of the study. The research team trained the local public health staff including community health workers (CHWs) and community volunteers in applying the dipping programme. A Kala-azar Nirmul Committee’ (village committee for VL elimination) of 2–3 members was formed in each of the study villages. The committee selected the community volunteers, identified suitable dipping points and dates and informed the community through local mosques about the dipping programme. Inhabitants were informed about the need for washing the nets before bringing them for dipping, procedures of dipping, safety measures and subsequent drying of the dipped nets in a horizontal position in a shaded area. CHWs did the same during their routine household visit. They also distributed informative leaflets to the households. On the dipping day, community people brought their recently washed nets to the dipping point. Community volunteers dipped the nets in a freshly prepared solution of KO-Tab 123 containing 0.4 g deltamethrin in a 1.6 g Tablet and a chemical binder [Bayer (Ply) Ltd. Reg. No. 1968/011192/07, Isando, South Africa] according to manufacture instructions. After dipping the net, owners dried their nets by putting them in a horizontal position in a shaded area. CHWs, village committee members and the research team observed and supported the dipping procedure. A few households without bed nets bought new ones to dip and those who were unable to buy new nets received commercially insecticide-treated nets after the dipping programme at no cost from the research team. In both study sites, trained field research assistants applied a structured questionnaire to heads of HHs (households) 1 and 6 months after the start of the dipping programme to assess acceptance and satisfaction regarding the treated bed nets. For sample size calculation, we hypothesized that the percentage of HH heads who might report adverse effect of treated nets or their dissatisfaction would be 5%. Requiring a maximum 95% confidence interval ±2% units the number of HHs needed to be surveyed was calculated to be 456. Thus, the interviews were carried out in 500 houses, selected by simple random sampling in Putijana/Mymensingh and in Deopara/Rajshahi, respectively, at 1 and 6 months after bed net impregnation. In addition, night observations (after 8:00 pm at night) were made in a 10% sub-sample of households to determine whether people used the treated nets and to validate interview answers. Monitoring of sandfly density with CDC Miniature Light Trap model 512 was carried in the Rajshahi site only and the traps was developed by the United States Centers for Disease Control (CDC). We calculated the sample size for sandfly measurement based on the following assumptions: (i) At baseline, the sandfly density in households of the intervention and control arm are the same but after intervention the density will be reduced by about 50% in the intervention arm and 10% in the control arm. (ii) The sample size ratio of intervention arm: control arm is 2:1 because it was necessary to include more households in the intervention arm than the control arm for a proper follow-up of sandfly densities. (iii) With an 80% power and alpha = 0.05, the minimum required sample size was determined to be 18 in the control arm and 37 HHs in the intervention arm. We randomly selected 50 HHs from the intervention area (Deopara/Rajshahi) and 26 HHs from the control areas for monitoring the sandfly densities. Sand flies were collected from each house by a CDC light trap which was set up between 6 pm and 6 am in the corner of the main room. The captured sand flies were killed with chloroform-soaked cotton before counting them. Sandfly collection was performed at baseline (before the intervention) as well as at 4 weeks, 12 months and 18 months after the intervention (dipping programme). Information which might be relevant for sandfly density – such as household type, having domestic animal and cattle shed – was also collected. Bioassays were performed using the WHO cone method (WHOPES 2005.11) on treated and non-treated bed nets with wild caught (manually collected) sand flies at 4 weeks, 12 months and 18 months after the dipping programme. In both study sites, 10 treated and 10 untreated nets were collected from 10 randomly selected HHs from intervention and control. Bioassays were carried out at room temperature at 25–29 °C and 75–85% humidity. Resting sand flies were collected manually by aspirator during early morning and evening. In each net, 10–12 female sand flies were introduced in each of the cones placed in five places (four on four sides and one on the top) of the net and exposed for 3 min. The sand flies were then transferred from each cone to a paper cup for observation. They were observed for 24 h to calculate the mortality. In the same way, control cones were placed on non-impregnated nets, and sandfly mortality was determined. The mortality rates of sand flies were corrected with Abbot’s formula (WHOPES 2005.11). Chemical analysis of the treated nets was performed at 1, 12 and 18 months after dipping. Impregnated polyester bed nets collected from randomly selected HHs in the Mymensigh study site were subjected to chemical analysis of the residual insecticide concentration in the nets. The chemical analysis was performed at Bio Genius Gmbh, Germany, where the deltamethrin content was estimated by a standard analytical method using high-pressure liquid chromatography (HPLC). Five pieces (four sides + top) of 25 × 25 cm from each of 10 nets were sampled. Deltamethrin was extracted into HPLC grade acetonitrile solvent (75 ml). Fixed quantities (25 μl) of the test samples were analysed on HPLC (HP 1100 with Auto sampler) system using reverse phase column (MERCK, LiChroCART® 125-4 LiChrospher® 100 RP-18, 5 μm) with acetonitrile/water as mobile phase (70:30 v/v and binary pump) and UV detector (diode array detector) at 230 nm at a flow rate of 1.5 ml/min. The content of deltamethrin active ingredient from the test samples was computed as mg/m2 from the average peak areas of duplicate injections by external standardization method and by The standard concentration was 25 mg/m2 et al. Five in each study site – and two – with 10–12 between and of were conducted 1 after the programme. The on the following of community and net of insecticide-treated nets, acceptance or of the programme and treated nets, net washing of increasing community and of bed net dipping. were by and were conducted in The were using a trained were to the the were into and analysed The information and test results were into the after been were used to the of of and analysis were performed in a because of the highly were calculated using as well as on the of the The analysis was performed using 10 To investigate the effect of intervention on vector we carried out the following The intervention effect was calculated by the following We the model to determine whether the intervention significant on the sandfly because the sandfly the the model the following = + + Intervention + intervention was 1 it was in the intervention and it was in the control Intervention effect and were from the of intervention and respectively. were and significant at analysis was performed for 1, 12 and 18 months of follow-up without and with for found during intervention at 1, 12 and 18 months were by test for both The study was by the at and Bangladesh informed was from all HH heads for observation and for informed was also from the as well as from village and public health workers for their in the study. Community in bed net dipping was as is in public health programme. In Putijana and respectively, 98.2% and 96.2% of the number of households in the bed net dipping programme. household on average and bed nets and and of existing bed nets were dipped in Putijana/Mymensingh and Deopara/Rajshahi, respectively. The results of the household interview surveys at 1 after dipping that of in Putijana/Mymensingh and in Deopara them to be against sand flies and high of acceptance 6 months to in Putijana/Mymensingh and in as a in efficacy against was by the of the adverse which might be to the treated bed nets. In it was found that the interview to bed net use high direct observation that in and of households in Putijana/Mymensingh and Deopara/Rajshahi, respectively, one or more were in the treated bed to the of the the most for a dipping programme was and They also the programme to the district. The average that each household would be and to on the insecticide was and in Putijana/Mymensingh and Deopara/Rajshahi, respectively. about of the in both the sites that community the insecticide tablets for from the Sandfly density did significant with of of and number of in the or with of domestic and of households with cattle more sand flies to households without cattle and for household with and without cattle respectively, = The number of and the also with the sandfly density = = Sandfly densities at points before and after net treatment that the dipping programme reduced the sandfly density in the intervention areas to the control areas in sandfly density by the intervention was about for 18 months period The intervention effect on sandfly densities at points in as estimated by the was significant The effect was significant after for and of the Intervention at points did = and = for simple and The corrected sandfly mortality at 24 h after been exposed to the treated nets was high at 1 and at 12 months after the dipping. the mortality rate but at about 80% at 18 months after dipping the two sites Sandfly to treated bed nets. The chemical analysis that the average concentration of deltamethrin in the treated nets 18 months from 24 to mg/m2 of deltamethrin in the treated bed nets. The the dipping programme and were after dipping no from and was that after 1 the nets, no at night was would to have IRS because nets them at night all that received the to nets more than a and to them after dipping in A few that the insecticide tablets were in the local and would to their nets We found that treatment of existing bed nets with a slow-release insecticide through local public health personnel and communities is feasible within a period of weeks to well and We did measure the sandfly reduction the nets – which is more than the nets – or in in intervention the effect of the programme on the sandfly We a significant reduction in sandfly densities for at 18 months in with nets to in untreated control The of sand flies, through was also satisfactory for 18 in the chemical the average deltamethrin concentration from to after 18 months. a concentration of 3 mg/m2 on bed nets is effective against vectors et al. studies to significant reduction in sandfly density by a effect et al. et al. 2006; et al. 2007; Ostyn et al. 2008; Picado et al. 2010). household coverage with LNs to be as the intervention in households in an community did sandfly densities in India et al. 2008). trial in Bangladesh, India and Nepal an average sandfly reduction of in in communities to in control communities (Joshi et al. recent in India and Nepal a vector reduction by or by spraying activities into in intervention with in control et al. 2010). The results in study may be because of insecticide of the vector, because of the of chemical in the study area the of the study. The reduction in and in VL incidence by effective vector control is an from is it a reduction in VL incidence through of LNs et al. et al. LNs may have a and a more of the nets. in with a high coverage of non-impregnated nets, the dipping programme with slow-release insecticides may be more LNs have a number of operational such as on community and on IRS by Thus, LNs be an option for sandfly for the national VL elimination programme in Bangladesh, where no vector control method is currently in We are to all study community volunteers and community in the both study areas. We with the of and local public health of of of Bangladesh. We for Tab 123 The study was supported by the for and in in The study was conducted in and with the of to the research