Search

School children are increasingly recognised as an important target for malaria control. However, limited evidence exists on the relative benefits of alternative school-based malaria control interventions or how impact may vary according to transmission intensity. Nested within the first evaluation of school-based intermittent screening and treatment (IST), in a region of low-to-moderate transmission, this thesis aimed to investigate the influence of heterogeneity in underlying transmission risk and variation in the fidelity of the intervention on the impact and possible operational applications of IST. The thesis utilized data from a cluster-randomised trial, over two years, evaluating the impact of school-based IST on health and education of children in 101 schools in south coast, Kenya. In the intervention children were screened for Plasmodium falciparum infection using rapid diagnostic tests (RDTs) once a term, with those found RDT-positive treated with artemether-lumefantrine. Cross-sectional analysis of baseline data, indicated marked heterogeneity in P. falciparum infection and anaemia at school-level. P. falciparum infection was strongly associated with anaemia, but health status was not associated with educational performance. Subsequent analysis of the impact of IST showed no benefit on the primary health or education outcomes even when stratified by baseline Plasmodium prevalence, or number of treatments received. Latent class analysis suggested reasonable diagnostic performance of RDTs, with seasonal variability observed. Bayesian geostatistical analysis of the spatial and temporal heterogeneity of school-level P. falciparum infection highlighted the relative stability in observed heterogeneity, despite periodic treatment of infection. Analysis at the individual-level suggested overdispersion of P. falciparum infection with prevalence-based models showing a greater proportion of individuals repeatedly infected than expected even after accounting for exposure. Collectively, these findings indicate that, in such a locally heterogeneous transmission setting, school-based IST provided no health or education benefits, with a number of factors identified for lack of impact. The persistence of high infection levels in certain schools despite periodic treatment highlights the importance of fine-scale targeting with a need for community-wide coverage in high-risk clusters. Moreover, the findings suggest a potential role for school-level screenings in both the identification of communities for targeted control, and periodic monitoring of impact.

In Ethiopia, malaria transmission is seasonal and epidemic-prone, with both Plasmodium falciparum and Plasmodium vivax being endemic. Such spatial and temporal clustering of malaria only serves to underscore the importance of regularly collecting up-to-date malaria surveillance data to inform decision-making in malaria control and improve responsiveness to potential epidemics. This thesis compares indicators and strategies used for the monitoring and surveillance of malaria in Ethiopia. Cross-sectional school-based surveys were conducted throughout Oromia Regional State, generating data on malaria prevalence by microscopy, risk factors for infection and intervention use. Filter paper blood samples collected during these school surveys were subsequently tested to determine exposure to malaria based on presence of anti-Plasmodium antibodies, and Bayesian geostatistical modelling was employed to predict P. falciparum and P. vivax seroprevalence across Oromia. In southern Ethiopia, a schoolbased syndromic surveillance system was piloted, exploring the utility of school absenteeism as a complementary indicator of malaria epidemics at community level. Finally, findings from the school surveys, measured and modelled seroprevalence, as well as data from the national Malaria Indicator Survey in 2011 were compared with spatially congruent estimates of malaria incidence collected from health facilities and to modelled parasite rate from the Malaria Atlas Project. Findings from this thesis demonstrate the limitations of microscopy as a primary indicator of malaria infection in cross-sectional surveys in areas of very low transmission. The work highlights the potential of serological indicators of Plasmodium exposure for inclusion in periodic large-scale malaria monitoring activities and develops a first ever geostatistical risk map based on serological indictors. This was supported by comparative analysis of a range of survey and modelling indicators against estimates of incidence from passive surveillance, indicating the inadequacy of cross-sectional surveys estimating population parasitaemia to reflect the spatial extent and temporal variability of transmission. The piloted syndromic surveillance system indicates that monitoring school absenteeism has potential as a complementary epidemic alert system, operating alongside the existing system at health posts, but is limited by low school enrolment in the piloted setting. The findings of this thesis indicate that existing periodic monitoring strategies and tools are insufficient to fully describe the extent of malaria in settings where Plasmodium transmission is spatially and temporally variable. Modifications to monitoring strategies are recommended, including incorporation of serological indicators and spatial modelling.

The last decade has seen significant progress towards the elimination of blinding trachoma as a health problem. However, gaps in our understanding of the epidemiology of trachoma at large scales are increasingly important in the context of programmatic scale up. This thesis therefore aimed to define the current distribution and burden of trachoma, in addition to investigating the spatial heterogeneity of trachoma and underlying risk factors at different scales. A systematic review of trachoma prevalence data was used to generate the Global Atlas of Trachoma, a unique spatially-referenced global database. In addition to highlighting important regional differences in the geographic distribution of trachoma, this database was used to quantify the disease burden in Africa; estimating nearly 153,000 disability-adjusted life years (DALYs) attributed to trachomatous vision loss and 155,500 additional DALYs to trichiasis. Detailed analyses of individual and cluster-level risk factors underlying the distribution of trichiasis in Nigeria and active trachoma in Kenya identified a number of key socio-demographic and environmental factors. Both analyses suggested that spatial dependency (generated by underlying associations with shared risk factors at larger scales) may vary in endemic areas. These findings emphasise the importance of local epidemiology and the need for robust and well-designed survey methodologies to identify areas of high risk. Computerised simulations were used to evaluate the performance of Integrated Threshold Mapping (ITM) in comparison to the accepted gold standard trachoma survey design. The results found that ITM tended to underestimate the prevalence of trachoma across a range of epidemiological contexts where attendance was low and/or the risk of disease was lower in school-going children. This thesis provides the first systematic investigation into the geography of trachoma; highlighting heterogeneities at different scales and their potential programmatic implications. In particular, the findings and methods from this thesis may help to inform future survey design.

School-aged children are the age group least protected by malaria interventions such as bed nets, despite being the age group most likely to be infected with malaria parasites. Effective targeting of malaria interventions among this age group is hampered by the lack of detailed data on the epidemiology of malaria. This thesis aims to describe the epidemiology of malaria among school children in Kenya and explores the usefulness of school-based approaches in malaria surveillance and control. A nationwide school malaria survey was carried across the different malaria ecologies in Kenya. These data allowed analysis in different settings of risk factors for Plasmodium infection and anaemia and the evaluation of alternate malaria diagnostic methods. A cluster randomised trial was conducted in coastal Kenya to evaluate the effectiveness of school-based distribution of mosquito nets. Finally, the congruence between reports of net use from school and household surveys was evaluated. Prevalence of Plasmodium infection was low overall, but varied markedly across the country. Risk factors for Plasmodium infection and anaemia varied by malaria transmission zone, with net use associated with reduced odds of infection in only coastal and western highland epidemic zones. The school-based distribution of mosquito nets was associated with an increase in reported net use but had no impact on Plasmodium infection or anaemia. In terms of identifying infection among individuals and populations, malaria rapid diagnostic tests represent a cheap diagnostic approach, especially in low and high prevalence settings. School surveys can also provide a reliable estimate of net use among both school children and households. Collectively, these results highlight the burden of malaria among Kenyan school children but show how this burden varies by transmission setting, emphasizing the need for a geographically targeted approach to tackling malaria. The results also demonstrate the role that schools can play in the surveillance and control of malaria.

Recent years have seen a shift towards integrated control of a number of tropical diseases. Such a strategy, however, relies on an understanding of the spatial distribution and overlap of different diseases. Using a combination of fieldwork, spatial and economic analyses and computerized simulations, optimal survey designs were explored for soil-transmitted helminths (STHs), Schistosoma mansoni and Plasmodium /alciparum in East Africa, and the potential of an integrated survey approach was evaluated. For STH, analysis indicated that hookworm clusters over larger scales than Ascaris lumbricoides and Trichuris trichiura, and that surveying small numbers of children, from four to five schools per district, provides a rapid and cost-effective approach to target treatment at district levels. For S. mansoni, Lot Quality Assurance Sampling (LQAS) was compared to a geostatistical survey design that allows spatial prediction at unsurveyed locations based on a subset of schools. Results showed that targeted treatment was more cost-effective than presumptive treatment and that, whilst LQAS correctly classified a higher proportion of schools requiring treatment, a geostatistical design proved more cost-effective. An investigation into the optimal spatial scale to conduct surveys for STH, S. mansoni and P. /alciparum in Kenya found that, over various cost scenarios, surveying fifty children from three randomly selected sites per sub-district provided a balance of performance and cost-effectiveness for all species. In sub-districts of low S. manson; and P.alciparum prevalence, LQAS should be used to target treatment. This thesis has shown that species-specific differences in spatial heterogeneity of infection and the costs of both mapping surveys and programme intervention have important implications for the optimal design of surveys. A two stage framework for integrated surveys is proposed allowing for a flexible approach to mapping. Similar studies in different settings are crucial and would help to assess whether changes in survey strategy are required as transmission drops due to control activities.