Abstract
School-aged-children are most vulnerable to schistosomiasis as exemplified by age-prevalence profiles although current understanding of these patterns needs improvement. Using epidemiological data from the southern shoreline of Lake Malawi, we investigated the dynamics of Schistosoma transmission and the main determinants of Schistosoma infection risk using a transmission dynamic model, considering urogenital and intestinal schistosomiasis respectively. Specifically, we assessed whether the proximity of primary schools to the immediate shoreline was a major geospatial and epidemiological determinant. Cross-sectional parasitology and malacological data previously collected and analysed was used, including age-infection profiles and interpolated predicted snail distributions for the southern part of Lake Malawi shoreline. A disease SEIRS ordinary differential equation model was created, and an observation prevalence model was formed using a binomial sampling distribution using the already published dataset. An optimisation using L-BFGS-B algorithm with upper/lower bounded box constraints was carried out to calibrate the model to find the best parameter values for each infection state transition given the disease model and dataset. The aim was to recapture the age-structure dynamics shown in the observation model representing the already published age-infection profiles. Concerning intestinal schistosomiasis, the best model for Biomphalaria sp. was the use of a single transmission rate for all the school's and no spatial effect. By contrast, for urogenital schistosomiasis, the best model for Bulinus spp. was found when using an independent transmission rate for each school and no spatial effect. There was some evidence that we were able to capture the age-structured dynamics of infection in SAC despite the expected outcome differing to statistical output due to sparse data. Within our study area, we found there was no significant effect on SAC exposure to Schistosoma infection risk based on school distance from the shoreline. Further, there was heterogeneity between schools in transmission rates estimated, although these did not have significantly different confidence intervals. However, schools considered in our study were all relatively close to cercaria infested shorelines. Further studies using a longitudinal cohort study could improve understanding of Schistosoma infection dynamics and allow for improved control method application.