Can repellents prevent malaria in Tanzania

Abstract

Background: Current malaria control tools, long lasting insecticidal nets and indoor residual spraying have had a significant impact on malaria transmission in sub-Saharan Africa. However these tools will not be able to eradicate malaria and there is need for complementary tools if this goal is to be attained. This work focused on evaluating and recommending tools that can be used to complement current control tools with emphasis on outdoor and early evening transmission. The main tool evaluated in this thesis was a topical repellent to be used in the early evening. Other tools recommended were spatial repellents and permethrin- impregnated clothing. Methods: A repellent efficacy trial was conducted in the semi-field and field setting to evaluate the protection from early evening biting given by a topical repellent lotion containing 15% N.N-diethyl-m-toluamide (DEET). A cluster randomized, placebo controlled clinical trial, designed to assess the effect of 15% DEET against malaria transmitted in the early evening was then conducted in a village in rural Tanzania. A total of 940 households were recruited and randomized, with 462 households randomized to the intervention arm and 462 households randomized to the control arm. The feasibility of lotion repellent use was assessed using entry and exit questionnaires. Focus group discussions were conducted 3 years after a clinical trial to assess the community knowledge, attitude and practice towards a different set of repellents to those used during the clinical trial. A systematic review was then conducted to put the results of this trial in context with other repellent trials. An attempt was made to design a clinical trial taking into account the shortcomings of the current and other repellent trials reviewed. Findings: Topical repellents containing 15% DEET provided >80% protection against early evening biting over four hours. According to protocol analysis of the cluster randomized trial found no difference between the intervention and control arms after accounting for socio-economic status, education of household head and household construction materials (Wilcoxon rank sum z = 0.529, p = 0.596). The most important predictor of malaria in this study was age, with younger age categories significantly associated with greater malaria risk. Socio-economic status was not associated with malaria. Compliance to repellent use was reported to be 80% during the study. From the FGDs, it emerged that community knowledge was the major barrier to repellent use, followed closely with availability. The community preferred using long lasting insecticidal nets (LLINs) because of their cost effectiveness. However, the community preferred using repellents in the early evening before employing LLINs. Interpretation: This study demonstrates that topical repellents have no effect against early evening malaria transmission in this community. However, shortcomings in the design and implementation might have masked the treatment effect and better-designed studies are required to establish repellents effect in this setting. Topical repellents provided protection against early evening biting and were readily accepted and used in this community, indicating the potential of using repellents complimentary to LLINs in this setting. The short-term duration of effect of this repellent, required frequent reapplication and therefore impacted compliance, emphasizing that future studies should consider using longer lasting tools such as spatial repellents.