Long-lasting insecticide-treated net usage in eastern Sierra Leone - the success of free distribution

Objective Médecins Sans Frontières (MSF) runs a malaria control project in Bo and Pujehun districts (population 158 000) that includes the mass distribution, routine delivery and demonstration of correct use of free, long-lasting insecticide–treated nets (LLINs). In 2006/2007, around 65 000 LLINs were distributed. The aim of this follow-up study was to measure LLIN usage and ownership in the project area. Methods Heads of 900 randomly selected households in 30 clusters were interviewed, using a standardized questionnaire, about household use of LLINs. The condition of any LLIN was physically assessed. Results Of the 900 households reported, 83.4% owning at least one LLIN. Of the 16.6% without an LLIN, 91.9% had not participated in the MSF mass distribution. In 94.1% of the households reporting LLINs, the nets were observed hanging correctly over the beds. Of the 1135 hanging LLINs, 75.2% had no holes or 10 or fewer finger-size holes. The most common source of LLINs was MSF (75.2%). Of the 4997 household members, 67.2% reported sleeping under an LLIN the night before the study, including 76.8% of children under 5 years and 73.0% of pregnant women. Conclusion Our results show that MSF achieved good usage with freely distributed LLINs. It is one of the few areas where results almost achieve the new targets set in 2005 by Roll Back Malaria to have at least 80% of pregnant women and children under 5 years using LLINs by 2010. Utilisation de moustiquaires imprégnées d’insecticide durable dans l’est de la Sierra Leone - Le succès de la distribution gratuite Objectif: Médecins Sans Frontières (MSF) mène un projet de lutte antimalarique dans les districts de Bo et Pujehun (population 158,000) qui comprend la distribution de masse, la livraison en routine et la démonstration de la correcte utilisation de moustiquaires gratuites imprégnées d’insecticide durable (LLINs). En 2006-2007, près de 65.000 moustiquaires ont été distribuées. L’objectif de cette étude de suivi était de mesurer l’utilisation et la possession de moustiquaires LLINs dans le sud-est de la Sierra Leone. Méthodes: Les chefs de 900 ménages choisis aléatoirement dans 30 groupes ont été interviewés, en utilisant un questionnaire standardisé, sur l’utilisation de LLINs par les ménages. L’état de toute LLIN a été physiquement examiné. Résultats: 83,4% des 900 ménages ont déclaré posséder au moins un LLIN. Des 16,6% de ménages sans LLIN, 91,9% n’avaient pas participéà la distribution en masse effectuée par MSF. Dans 94,1% des ménages possédant des LLINs, les moustiquaires observées pendaient correctement au-dessus des lits. Sur les 1135 LLINs accrochées, 75,2%) n’avaient pas de trous ou avaient moins de dix trous de la taille d’un doigt. La source la plus commune des moustiquaires était MSF (75,2%). Sur les 4997 membres du ménage, 67,2% ont déclaré dormir sous une moustiquaire imprégnée d’insecticide durable la nuit précédant l’étude, y compris 76,8% des enfants de moins de 5 ans et 73,0% des femmes enceintes. Conclusion: Nos résultats montrent que MSF a atteint le bon usage des moustiquaires distribuées gratuitement. Il s’agit d’un des rares domaines où les résultats ont presque atteint les nouveaux objectifs fixés en 2005 par ‘Roll Back Malaria’ d’avoir au moins 80% des femmes enceintes et des enfants de moins de 5 ans utilisant des moustiquaires en 2010. Uso de mosquiteras impregnadas con insecticida de larga duración en Sierra Leona del este – el éxito de la distribución gratuita Objetivo: Médicos sin Fronteras (MSF) tiene un proyecto de control de la malaria en los distritos de Bo y Pujehun (población 158.000) que incluye la distribución masiva y gratuita, entrega rutinaria, y demostración del uso correcto de mosquiteras impregnadas con insecticida de larga duración. En 2006/2007, se distribuyeron alrededor de 65,000 LLINs. El objetivo de este estudio de seguimiento fue medir el uso y tenencia de LLIN en el sudeste de Sierra Leona. Métodos: Se entrevistó a los jefes de 900 hogares elegidos al azar en 30 conglomerados, utilizando un cuestionario estandarizado, sobre el uso de LLINs en sus hogares. La condición de cualquier LLIN fue evaluada físicamente. Resultados: Un 83.4% de los 900 hogares reportaron tener al menos una LLIN. De los 16.6% sin una LLIN, un 91.9% no habían participado en la distribución masiva de MSF. En un 94.1% de los hogares que reportaron poseer una LLINs, se observó que las redes estaban correctamente colgadas sobre las camas. De las 1135 LLINs colgadas, un 75.2% o no tenían huecos o tenían diez o menos huecos del tamaño de un dedo. La fuente más común de LLINs era MSF (75.2%). De los 4997 miembros de los hogares incluidos en el estudio, un 67.2% reportaron haber dormido bajo una LLIN la noche antes de realizar la encuesta, incluyendo un 76.8% de los niños menores de 5 años y un 73.0% de las mujeres embarazadas. Conclusión: Nuestros resultados demuestras que MSF logró un buen uso de las LLINs con su distribución gratuita. Esta es una de las pocas áreas en donde los resultados obtenidos casi han alcanzado los objetivos propuestos en el 2005 por Roll Back Malaria de tener al menos un 80% de las mujeres embarazadas y niños menores de 5 años utiliznado LLINs antes del 2010. Malaria still threatens the lives of millions, particularly in lower-income countries where it is endemic. Approximately half the world’s population is at risk from this preventable, treatable and curable disease. In 2006, 247 million malaria cases caused nearly 1 million deaths, mostly in children under 5 years (WHO 2008). A major interdisciplinary strategy to control malaria is underway, based on prevention and prompt and effective treatment (Anonymous 2008; WHO 2008). Long-lasting insecticide–treated nets (LLINs) are part of the prevention strategy. Across a range of transmission settings in Africa, high levels of LLIN use have been shown to reduce malaria-related mortality, especially in children under 5 years (Lengeler 2004). Achieving such levels is a goal in the malaria control efforts of the Roll Back Malaria (RBM) Partnership (WHO 2008). The 2000 RBM Summit in Abuja, Nigeria, set a target for 2005: 60% of those most vulnerable to malaria (children under 5 years and pregnant women) should have access to and sleep under LLINs (WHO 2000; Rowe et al. 2006). In 2005, RBM raised this target to 80% to be reached by 2010 (RBM 2005). There is no clear consensus on the most suitable and effective way of achieving socio-economic equity in distribution and full population ownership (possession of an LLIN) and usage (sleeping under an LLIN). Opinions differ on the benefits of mass distribution versus routine delivery and free distribution versus cost-sharing (Curtis et al. 2003; Cohen Khatib et al. 2008). To determine the effectiveness of distribution channels, assessments of LLIN ownership are important. However, usage rather than ownership is the crucial indicator for whether distribution will lower the burden of malaria (Baume a maximum of two LLINs was given to households with three or more children under 5 years. In each village, a reference person was chosen from the community and trained by the distribution team to work as a volunteer to help calculate the number of LLINs needed for the village, facilitate the distribution process and support villagers in correctly hanging and using LLINs. Before the mass distribution, at least two education and awareness meetings were held with the head of the village and village opinion leaders with the help of a health educator. A theatre performance took place in each village demonstrating how to hang and use LLINs. Malaria education sessions were held at least weekly in the primary and secondary health structures to coincide with antenatal clinics. The aims of this follow-up study were to measure the usage and ownership of LLINs and to see whether these results met the RBM Abuja targets. We used a three-stage cluster sampling method with a probability proportional to the estimated population adapted from the method recommended by WHO (Henderson approximately 357 villages with a total population of 158 000 in a radius of around 10 km around five community health centres. The study population included everyone living in the catchment area. The basis for the population and village estimation was a 2007 MSF mapping (MSF in-house mapping 2007). Each of the five community health centre catchment areas contained between 22 580 and 39 208 people living in 43–93 villages. Most houses were widely scattered around the village centre, usually the market place, and accessible only by footpath. The population size per village averaged 440 (minimum 16, maximum 3431 persons). Half the population are subsistence farmers and a quarter are diamond diggers (Gerstl 2009). Fewer than 20% of adults are able to read and write (Gerstl 2008). The average household size was between six and seven household members (MSF in-house mapping 2007, unpublished; MSF internal report 2008, unpublished; Gerstl 2008). We took six household members as the conservative average, so with an estimated population size of 158 000, there were about 26 300 households in the catchment area. From previous studies (Gerstl 2008), we also estimated that at least 50% of households would own at least one LLIN. With a precision of 10%, α-error of 5% and design effect of 2, 758 households were required. A sample size of 760 children under 5 years, an at-risk group for malaria, was estimated as necessary for an expected use of LLINs of 50%, precision of 10%, α-error of 5% and design effect of 2. As this study was part of a mortality and nutritional study in which 900 households and 900 children under 5 years were required, the minimum sample size was exceeded. A three-stage cluster sampling methodology was used. In the first stage, 30 clusters were selected from a list of all five community health centre catchment areas using a probability of allocation proportional to the respective population size of each area. In the second stage, the selected number of clusters per catchment area (between 4 and 7) was allocated to villages within this area by systematic sampling. The probability of allocation was proportional to the respective population size of each village. In the third stage, 30 households were randomly selected within a village (=cluster). A pen was thrown on the ground in the centre of the village, and a line drawn in the direction it pointed, towards the edge of the village. Households were counted along this line by walking to the edge of the village. With the use of a random number chosen from a random number table, one of these households was selected as the first to be interviewed. The next closest household was then interviewed until 30 had been included. If the village had fewer than 30 households, the cluster was continued by selecting the (geographically) closest village where the same methodology was used to select the first household. Heads of households were interviewed at their homes in one of the local languages (Mende or Creole) or English using a standardized, pre-piloted questionnaire. The study was anonymous. Five teams of three interviewers completed one cluster in 1 day. Household members were asked about the presence and quantity of LLINs in the household, details and quality of existing LLINs and usage of LLINs. If there were no nets in the household, household members were asked the reason why. In net-owning households, interviewers asked permission to enter and count LLINs and establish whether they were hanging correctly over sleeping places. If LLINs were tied up over the sleeping place, household members were asked to demonstrate correct usage. Interviewers noted the brand and assessed the condition of LLINs by counting holes graded in three sizes: finger (no finger-size holes, 1–10 finger-size holes, >10 finger-size holes), fist (likewise) and head (likewise). The head of each household was asked why LLINs were not hanging correctly over sleeping places (if applicable), the source of each LLIN, the year of acquisition and the frequency with which it was washed. Data were entered into EpiData 3.0 software (The EpiData Association, Odense, Denmark). Data cleaning checked for inconsistencies in data entry and responses. Data analysis used stata 8.1 (StataCorp, College Station, TX, USA) and spss 11.0 (SPSS, Chicago, IL, USA). All indicators (e.g. sex and of the study LLIN usage and were as and with of design effect were also all the design effect was to we not report the was received from the of MSF and the and of the of Health and Sanitation of Sierra Leone. was from the of households before the and care was to that all household members that household was the and 2008, 900 households with a total of 4997 people were and interviewed. one to 1 of the study Of the 900 households interviewed, 83.4% reported owning at least one LLIN; 16.6% not own an LLIN 91.9% one of these households they had not participated in any MSF LLIN distribution. for all given for not owning an LLIN. Of the households with no were not for the mass distribution not a under 5 years or a pregnant included at least one under 5 years, included at least one under 5 years and at least one pregnant and included at least one pregnant and households were for the mass of these not own an LLIN. of houses not for the mass distribution not own an LLIN. Of the households that reported owning at least one LLIN, 94.1% had the correctly hanging over the had at least one LLIN, not hanging over the and in the with the were and not be The main given by the households for hanging not used and still in used to as and at of Of these included at least one under 5 not a under 5 years or a pregnant The households that LLINs included household In these households, interviewers counted 1135 were correctly hanging over the were one LLIN correctly or was between household members and one correctly hanging LLIN was between three household members The main given for the hanging LLINs were not used and observed to be in and used to as for all 4 the year of distribution and frequency of correctly hanging LLINs. Of the 1135 correctly hanging LLINs, not have any holes, and a had at most 10 holes. by year of the number of LLINs without holes was of the LLINs received 2008, of the LLINs received 2007 and of the LLINs received before 2007 PermaNet® was the most common brand in The most common source of LLINs was were MSF mass and in the antenatal or from the referral In 2007 or before 2007 of LLINs were and of LLINs had been up to 4 by year of the number of LLINs that had not been was of those in 2008, of those in 2007 and of those before 2007 Of the 4997 people in the study, had under an LLIN the night before the study, as reported by the head of the household. and were not with usage. Of children under 5 years, 76.8% were reported as sleeping under an LLIN. Of the women were more than had under an LLIN. Of those were reported as sleeping under an LLIN as were of year 73.0% of year and of those and In a and distribution strategy in high ownership and usage of of households at least one LLIN, almost all these households had their LLINs correctly and almost had under an LLIN the night before the was for the most – children under 5 years and pregnant women of households for mass distribution not own an LLIN. to the an LLIN should for at least and years of use (WHO 2007). Most LLINs were to be still households had received between 2006 and 2008 and only around had been more than were in good condition – had no holes or fewer than 10 holes. Our results the 2005 RBM target of at least 60% of pregnant women and children under 5 years using LLINs (WHO and were to the 2010 target of 80% (RBM 2005). The on malaria prevention is LLIN usage not LLIN In to a of between ownership and usage of LLINs has been in et al. et al. et al. and the et al. 2008). reason for this be the of LLIN et al. 2008; et al. 2008; et al. 2009). The of the MSF distribution have usage of LLINs in reason for high usage be since people to use LLINs more the However, as Sierra Leone has a perennial with this would not between LLIN usage and ownership will ownership of at least one LLIN per household is et al. 2009). distribution achieve a households access be In study, on average, three household members one correctly hanging LLIN. studies a maximum of two people per LLIN per household as to reported usage were good et al. 2003; et al. et al. et al. 2009). In Nigeria, the LLIN ownership et al. 2008). In free LLIN distribution was to et al. and in free LLIN distribution was to lower mortality and morbidity et al. et al. 2007). In people received free LLINs were no to use than those had for & Dupas 2008). A between LLIN ownership and free distribution was in a of data from countries in et al. free distribution the only way to in ownership and to achieve high LLIN usage. countries such as and have achieved in LLIN usage (Baume & Marin 2008). there areas where usage is In 2007, only of children living in areas of malaria transmission were by an LLIN et al. 2007, million LLINs would have been to 80% in population et al. 2007). However, at the of 2006, only million effective LLINs were in in (WHO 2008). Long-lasting insecticide–treated nets at high ownership and usage levels population and those not sleeping under a will achieving mass et al. A usage of benefits et al. 2007). Our LLIN usage and ownership should to the There are in usage and ownership We not a control area. it is to results to areas and malaria and we results with However, a indicator by in Sierra Leone in 2005 that only 5% of children under 5 years had under an LLIN a lower than we The study was in an area where for years MSF has free malaria and treatment in a of free primary and the population is used to high quality free The for the mass distribution were around per LLIN. We are that in settings and for and these be However, LLIN distribution is In distribution of LLINs within the Health in in of and years within set by malaria prevention studies et al. 2008). It has been estimated that with LLINs in is by at the of per year et al. 2007). In ownership and usage of LLINs in study population almost achieved the 2010 RBM target of 80% LLIN usage in vulnerable population To the 2010 RBM we the use of mass distribution and routine delivery of LLINs with an We the study teams for their and the We also the Médecins Sans Frontières teams in and for their to Sans for the this study would not have been without the and of the people of the villages of Bo and Pujehun