Integrated vector management (IVM) as a tool for community empowerment towards malaria elimination in Rwanda

Most malaria control interventions are designed to target the life cycle of the malaria vector or the malaria parasite with the purpose to interrupt malaria transmission. However, the interruption of malaria transmission and eventual elimination of the disease remain one of the hardest challenges in malaria control, particularly in countries with low income-economies. Based on lessons from past malaria control strategies, and the promising results from insecticide treated nets (ITNs), long-lasting insecticide treated nets (LLINs) and new insecticides for indoor residual spraying (IRS), the global public health community was mobilized to aim for malaria elimination. Thus, the outcomes obtained on malaria reduction encouraged some countries, including Rwanda, to move forward to a malaria elimination phase. At the same time, malaria endemic countries have been encouraged to adopt the concept of integrated vector management (IVM) and, consequently, to rationalize the usage of limited resources. This approach intends to transform the existing vertical system of vector control by making it more evidence-based, and by enhancing the participation of decentralized communities. For the current thesis, the concept of IVM, with an emphasis on empowerment of the local community, was implemented for malaria control in Rwanda. The goal was to investigate the feasibility of malaria elimination in Rwanda and evaluate potential hindrances in rolling out IVM. Key questions therefore were: (1) what are the key entomological and environmental determinants which characterize malaria transmission in Rwanda? (2) what are the behavioral characteristics of vector populations over time and space in response to existing vector control interventions? and (3) will community based know-how and larval source management interventions next to the National Malaria Control Programme (NMCP) policy on LLINs, IRS and case management, achieve and sustain malaria control and contribute to elimination? The study began in the context of a lack of accurate entomological information as previous studies were carried out during the malaria eradication programme (1955-1969). Chapter 2 therefore provides updated entomological knowledge on malaria vector species composition, their major behaviours (feeding and resting), and the distribution of malaria transmission intensity across Rwanda. Human Landing Collections (HLC) and Pyrethrum Spray Collections (PSC) were the methods used in order to collect comparable data in line with the sampling methods previously used in Rwanda. It was shown that Anopheles arabiensis became the dominant malaria vector and replaced An. gambiae s.s. and An. funestus, previously considered as primary malaria vectors. The peaks of Anopheles gambiae s.l. bites occur at the second half of the night and consequently effective indoor vector control interventions should provide adequate protection to the population against indoor transmission of malaria. However, we also observed trends towards earlier and outdoor transmission, although this was variable across study sites. Vector control strategies have to be adapted to the above challenges. However, the intensity of outdoor transmission has not been measured and future research should focus on this issue as well as on the evaluation of existing and new outdoor mosquito control methods. The patterns of exophilic and zoophilic behaviours of malaria vectors, and the time when the majority of the human population goes to bed and gets up should be prioritized in future research in order to better estimate the risks of residual malaria transmission in Rwanda. Results from Chapter 2 made clear that alternative mosquito sampling strategies are desired for objective monitoring, especially because HLC and PSC are labour-intensive and not favoured for ethical reasons. Chapter 3 therefore describes the results from a comparison of mosquito sampling methods, namely Centers for Disease Control (CDC) light traps and Suna traps equipped with a nylon, odour-baited strip, and set up with or without a light. The CDC light traps proved effective in sampling An. arabiensis, the dominant malaria vector species in the study site, both inside and outside houses. Interestingly, the Suna trap was more sensitive in sampling the high ratio of An. gambiae s.s. The experiment was limited to one study site and similar experiments should therefore be conducted in other malaria transmission settings or with different dominant malaria vector species. Furthermore, for improvement of its catches, the Suna trap should again be evaluated with an addition of carbon dioxide, as an activator of host-seeking behaviour, or with the CO2 mimic 2-butanone. Because current malaria control strongly depends on the proper use of LLINs by local communities, there is a need to assess their physical durability under real-life conditions. Chapter 4 therefore presents the outcomes on the physical durability and attrition rate of LLINs under usage at community level, and addresses the assumption of protection that they provide to the users against infective mosquito bites. It was found that the physical serviceable life of LLINs in Rwanda was closer to two years, which is lower than the distribution-replacement cycle of three to five years indicated by bed net suppliers. The attrition of LLINs at community level varied according to the net fabric and the level of malaria endemicity. The findings confirmed previous observations on the decreasing impact of LLINs on malaria burden, frequently found after two years following the countrywide mass distribution of LLINs. The results stress the importance of maintaining universal coverage of LLINs through the annual routine LLINs distribution, to shorten the LLIN replacement cycle to two years, and then highlight the importance of continuous LLIN durability monitoring by National Malaria Control programmes. Further assessment of the reasons of high attrition and the physical deterioration of LLINs under field conditions will help to identify potential solutions which could be implemented by community members with technical support from other malaria control stakeholders. Besides attrition of LLINs hampering malaria control, there are also increased reports of resistance of the mosquito vectors towards the insecticides used on LLINs or used within IRS campaigns. In Chapter 5, six insecticides were tested for measuring resistance levels of local vector populations throughout the country. For the first time, resistance to pyrethroids, used both for treatment of LLINs and for IRS application, was detected in 2011 in a few sites, and its onward spread was confirmed for many sites across the country in 2013. These results helped to explain the continued increase of malaria that occurred from 2013 in regions where communities benefit from universal protection with LLINs. Results also provided information for decision making on the prevailing resistance mechanisms, such as genetic-based kdr resistance and metabolic detoxification associated with esterase enzymes. Future research should emphasize the extent of each resistance mechanism, the resistance intensity and determine the presence of other potential resistance mechanisms involving mono-oxygenases and gluthathione S-transferase enzymes. This information will further guide decision making on when the use of a given insecticide should be stopped for vector control and to make an appropriate choice of an alternative insecticide. Moreover, the findings presented in this chapter provided basic information for the development and adoption of a national strategy for insecticide resistance management including the rotation of insecticide products every two to three years, and consequently sustain the efficacy of LLINs and IRS interventions. Because insecticide-based strategies pose more and more challenges (Chapters 4 and 5), there is a need to evaluate and implement alternative malaria control interventions. With the principles of IVM in mind, we therefore evaluated the impact of a new biological control tool for the larval stages of mosquito vectors, the bacterium Bacillus thuringiensis var. israelensis (Bti), and focused on the role that local communities could play themselves in implementing such programmes. Chapter 6 therefore explores the feasibility of empowerment of local communities for implementing mosquito larval control. First, by using the Open Space method as community engagement strategy, we established a preliminary list of community members, identified their expectations and co-designed the strategies adapted to their commitment. Mosquito larval source management was most often expressed by community members as an area for empowerment in knowledge. The most engaged community representatives consisted of rice farmers, community health workers and local leaders at village level. Then, two local community teams were set up and assigned separate responsibilities: the application of Bti and the entomological surveillance to monitor the impact of Bti, respectively. The first team was mainly made up of rice farmers, familiar with crop protection using sprayer pumps. Then, they were trained in the techniques for calibration of spraying equipment, preparing the required dosages and on how to apply Bti in different types of breeding sites. During the application of Bti, participants were split into two teams, one supervised by the project experts and another as community-based independent team. The findings showed that the community can be successfully empowered and participate in monitoring the impact of malaria vector control interventions such as larval source management. Moreover, the entomological outcomes resulting from the Bti application between the two separate teams, were similar in terms of reduction of larval and adult mosquitoes in the study area. This chapter demonstrates that communities can themselves take up the responsibility of larviciding and thus contribute to vector control in the framework of an integrated approach. The cooperatives of rice farmers, the community health workers and the community leaders represent the key stakeholders to supplement the existing vector control interventions using larval source management. The general discussion, Chapter 7, links the key findings to the primary research questions and to their implications for national vector control policies and strategies. The link between the results and the current implementation of the five pillars of IVM in Rwanda is discussed, as well as how to bring in the concept of leadership into IVM and the feasibility of malaria elimination in Rwanda. It is concluded that despite the ambitious goal to achieve malaria pre-elimination in Rwanda, the achievements previously obtained in malaria control appeared fragile. The major limitations and weaknesses of the core indoor vector control interventions related to the required quality of interventions and to new behaviours of malaria vectors. Furthermore, the changes in vector host-seeking, the shift in species composition, and the trends in earlier and outdoor biting indicate the potential existence of residual transmission. Thus, supplemental vector control tools are required and preferably should involve local communities and be implemented under the IVM approach. Larval source management using Bti proved to be effective and should be implemented through community networks and integrated into best practices of community-based organizations. The community showed the willingness to participate in larval source management if community members are empowered in knowledge and then products made available at an affordable cost. Malaria elimination is still feasible in Rwanda, but requires adaptive strategies based on the principles of IVM and integrating ‘leadership’ into its primary concept.