No abstract

Summary New tools to prevent malaria morbidity and mortality are needed to improve child survival in sub‐Saharan Africa. Insecticide treated bednets (ITBN) have been shown, in one setting (The Gambia, West Africa), to reduce childhood mortality. To assess the impact of ITBN on child survival under different epidemiological and cultural conditions we conducted a community randomized, controlled trial of permethrin treated bednets (0.5 g/m2) among a rural population on the Kenyan Coast. Between 1991 and 1993 continuous community‐based demographic surveillance linked to hospital‐based in‐patient surveillance identified all mortality and severe malaria morbidity events during a 2‐year period among a population of over 11 000 children under 5 years of age. In July 1993, 28 randomly selected communities were issued ITBN, instructed in their use and the nets re‐impregnated every 6 months. The remaining 28 communities served as contemporaneous controls for the following 2 years, during which continuous demographic and hospital surveillance was maintained until the end of July 1995. The introduction of ITBN led to significant reductions in childhood mortality (PE 33%, CI 7–51%) and severe, life‐threatening malaria among children aged 1–59 months (PE 44%, CI 19–62). These findings confirm the value of ITBN in improving child survival and provide the first evidence of their specific role in reducing severe morbidity from malaria.

The use of insecticide-treated bednets (ITBNs) has been shown to be effective in reducing mortality and morbidity from malaria. However, there is mixed evidence as to whether or not community-wide use of ITBNs engenders a 'mass effect', such that those not sleeping under bednets are offered protection from widespread ITBN use in the area in which they live. We have analysed data collected in Kilifi, Kenya, from a cohort of children followed from birth to investigate how the degree of net usage in the locality of a child affects the risk of developing malaria. This effect was explored using a Cox proportional hazards model. For those not using ITBNs, we found that an increasing level of ITBN usage within the area surrounding each child was associated with a decreasing risk of developing malaria, thus providing evidence in support of a mass community effect. The size and significance of this effect were found to decrease as non-overlapping areas of increasing distance away from a child's home were considered. The effect was significant for areas at distances of up to 1.5 km away from each child.

The results of recently completed trials in Africa of insecticide-treated bed nets (ITBN) offer new possibilities for malaria control. These experimental trials aimed for high ITBN coverage combined with high re-treatment rates. Whilst necessary to understand protective efficacy, the approaches used to deliver the intervention provide few indications of what coverage of net re-treatment would be under operational conditions. Varied delivery and financing strategies have been proposed for the sustainable delivery of ITBNs and re-treatment programmes. Following the completion of a randomized, controlled trial on the Kenyan coast, a series of suitable delivery strategies were used to continue net re-treatment in the area. The trial adopted a bi-annual, house-to-house re-treatment schedule free of charge using research project staff and resulted in over 95% coverage of nets issued to children. During the year following the trial, sentinel dipping stations were situated throughout the community and household members informed of their position and opening times. This free re-treatment service achieved between 61-67% coverage of nets used by children for three years. In 1997 a social marketing approach, that introduced cost-retrieval, was used to deliver the net re-treatment services. The immediate result of this transition was that significantly fewer of the mothers who had used the previous re-treatment services adopted this revised approach and coverage declined to 7%. The future of new delivery services and their financing are discussed in the context of their likely impact upon previously defined protective efficacy and cost-effectiveness estimates.