The observed relationship between the degree of parasite aggregation and the prevalence of infection within human host populations for soil-transmitted helminth and schistosome infections

Measuring heterogeneities in soil-transmitted helminth transmission and control

The global effort over the past decade to control soil-transmitted helminths (STH) has resulted in communities with endemic infection reaching low prevalence levels suitable for the validation of elimination as a public health problem (EPHP), defined by the World Health Organisation (WHO) as <2% of infections classified as moderate or heavy intensity. The spatial scale in which this is validated is currently undefined. As the burden of STH infection decreases, the degree of aggregation of infection within individuals in a population increases. Identifying these remaining pockets of infection requires fine-scale monitoring and evaluation (M&E) programmes that are rarely implemented within current national neglected tropical disease (NTD) control. This review examines various heterogeneities that characterise the epidemiology of STH infections, and discusses their impact on control policy formulation.

Diagnosis of helminths depends on worm fecundity and the distribution of parasites within hosts

Helminth transmission and morbidity are dependent on the number of mature parasites within a host; however, observing adult worms is impossible for many natural infections. An outstanding challenge is therefore relating routine diagnostics, such as faecal egg counts, to the underlying worm burden. This relationship is complicated by density-dependent fecundity (egg output per worm reduces due to crowding at high burdens) and the skewed distribution of parasites (majority of helminths aggregated in a small fraction of hosts). We address these questions for the carcinogenic liver fluke Opisthorchis viverrini, which infects approximately 10 million people across Southeast Asia, by analysing five epidemiological surveys (n = 641) where adult flukes were recovered. Using a mechanistic model, we show that parasite fecundity varies between populations, with surveys from Thailand and Laos demonstrating distinct patterns of egg output and density-dependence. As the probability of observing faecal eggs increases with the number of mature parasites within a host, we quantify diagnostic sensitivity as a function of the worm burden and find that greater than 50% of cases are misdiagnosed as false negative in communities close to elimination. Finally, we demonstrate that the relationship between observed prevalence from routine diagnostics and true prevalence is nonlinear and strongly influenced by parasite aggregation.