Changing patterns of malaria epidemiology between 2002 and 2010 in Western Kenya: the fall and rise of malaria

Abstract

Background: The impact of insecticide treated nets (ITNs) on reducing malaria incidence is shown mainly through data collection from health facilities. Routine evaluation of long-term epidemiological and entomological dynamics is currently unavailable. In Kenya, new policies supporting the provision of free ITNs were implemented nationwide in June 2006. To evaluate the impacts of ITNs on malaria transmission, we conducted monthly surveys in three sentinel sites with different transmission intensities in western Kenya from 2002 to 2010.

Methods and findings: Longitudinal samplings of malaria parasite prevalence in asymptomatic school children and vector abundance in randomly selected houses were undertaken monthly from February 2002. ITN ownership and usage surveys were conducted annually from 2004 to 2010. Asymptomatic malaria parasite prevalence and vector abundances gradually decreased in all three sites from 2002 to 2006, and parasite prevalence reached its lowest level from late 2006 to early 2007. The abundance of the major malaria vectors, Anopheles funestus and An. gambiae, increased about 5-10 folds in all study sites after 2007. However, the resurgence of vectors was highly variable between sites and species. By 2010, asymptomatic parasite prevalence in Kombewa had resurged to levels recorded in 2004/2005, but the resurgence was smaller in magnitude in the other sites. Household ITN ownership was at 50-70% in 2009, but the functional and effective bed net coverage in the population was estimated at 40.3%, 49.4% and 28.2% in 2010 in Iguhu, Kombewa, and Marani, respectively.

Conclusion: The resurgence in parasite prevalence and malaria vectors has been observed in two out of three sentinel sites in western Kenya despite a high ownership of ITNs. The likely factors contributing to malaria resurgence include reduced efficacy of ITNs, insecticide resistance in mosquitoes and lack of proper use of ITNs. These factors should be targeted to avoid further resurgence of malaria transmission.

Figure 1. Dynamics of Plasmodium falciparum parasite prevalence (%) from 2002 to 2010.

Study site from top to bottom: Kombewa, Iguhu, and Marani. The timings of major changes in ITN/LLIN distribution policy were marked on top.

Figure 2. Dynamics of Plasmodium falciparum infection intensity and gametocyte prevalence (%) from 2002 to 2010.

Study site from top to bottom: Kombewa, Iguhu, and Marani.

Figure 3. Temporal dynamics of Anopheles gambiae s.l. (left panel) and An. funestus (right panel) density (number of females/house/night) from 2002 to 2010.

Figure 4. Species composition of Anopheles gambiae s.s. and An. arabiensis in Kombewa (top) and Iguhu (bottom) from 2003 to 2010.

Figure 5. A) Increase in bed net ownership, presented as the percentage of households owning at least one net, from 2004 to 2010 (top); B) percentage of bed net coverage among the population in 2010 (bottom).

Effective bed net coverage is defined as the proportion of the population who slept under LLINs in good condition, or slept under regular ITNs in good condition and that were routinely insecticide-retreated. Ineffective population coverage by bed net is defined as the proportion of the population who slept under torn nets, or slept under regular ITNs but not routinely insecticide-retreated.

Figure 6. Variations in monthly maximum temperature, minimum temperature, mean temperature and monthly rainfalls in Kombewa (top), Iguhu (middle) and Marani (bottom) from 2002 to 2010.