A two-pronged approach to address the neglect of female genital schistosomiasis: an analysis of the association with sexual and reproductive health conditions and evaluation of the cost-effectiveness of screening strategies.

Female genital schistosomiasis (FGS) is a chronic parasitic gynaecological disease that is highly prevalent in sub-Saharan Africa (SSA), yet it remains largely underreported, misdiagnosed, and untreated. Despite affecting an estimated 30-56 million women globally, only approximately 15,000 women have been screened to date. FGS disproportionally affects populations already burdened by HIV and cervical cancer. Previous studies have suggested an association between FGS and increased prevalence of HIV and high-risk (HR-) human papillomavirus (HPV), the main predictor of cervical precancer and cancer. Diagnosis of FGS is challenging and has traditionally required high levels of technical expertise based on visual inspection for FGS typical lesions of the genital tract, using a standard colposcope that is seldom available in endemic settings. More recent studies have piloted closer-to-the-user strategies for FGS screening and diagnosis, such as genital self-sampling for parasite detection by molecular methods. However, to ensure scalability, these strategies must be cost-effective; to date, no studies have evaluated their cost or cost-effectiveness.

An initial scoping review conducted as part of this PhD thesis identified a number of research gaps, including the need to measure the burden of FGS, to understand the interplay of FGS with other sexual and reproductive health (SRH) infections, and to accurately predict the cost-effectiveness of different screening and diagnostic strategies. To address these research gaps on the epidemiology of FGS and efficient resource allocation for disease control, this PhD thesis has three objectives:

1. Evaluate the burden of FGS and associated risk factors in two Schistosoma (S.) haematobium endemic areas in Southern Malawi using visual and molecular diagnostic methods. 2. Evaluate the association between FGS, diagnosed using visual and molecular methods, and HR-HPV among women of reproductive age in Zambia. 3. Evaluate the cost-effectiveness of home-based self-sampling and clinic-based sampling for FGS screening compared to a no screening approach (the current standard of care) using a novel health decision analytical model.

The first objective used data from the Morbidity Operational Research for Bilharziasis Implementation Decisions for Girls and Women (MORBID-FGS) study in Southern Malawi. This is the largest FGS burden of disease study to date, which has screened 994 women aged 15-65 years for FGS using both visual and molecular diagnostic methods. Results showed a moderate to high burden of FGS (between 8-27% depending on the diagnostic method used), with marked age differences in diagnostic performance. Further, some villages with low schistosomiasis prevalence (which would be excluded from control strategies per new WHO guidelines) had a significantly high burden of FGS, indicating the need for formal public health interventions.

For the second objective, baseline data from the ongoing Zipime-Weka-Schista longitudinal study was used. We found that, after adjusting for known confounders, women with FGS diagnosed using molecular methods had a higher risk of testing positive for HPV 16 and 18 and/or 45, the most oncogenic HR-HPV genotypes.

The last objective developed a novel health decision analytical model (the FGS-SCREEN model) to evaluate the cost-effectiveness of different FGS screening strategies. The model was then implemented as part of the Zipime-Weka-Schista study to compare the cost-effectiveness of home-based self-sampling and clinic-based sampling to a no screening approach (current standard of care). Home-based self-sampling had a lower unit cost per woman screened compared to clinic-based sampling. Neither strategy was found to be very cost-effective if implemented individually. However, a combined screening pathway incorporating home-based self-sampling in rural (80% coverage of screening-eligible population) and clinic-based screening in peri-urban areas (20% coverage of screening eligible population) was found to be very cost-effective under conservative thresholds.

Findings from this thesis offer significant methodological and empirical contributions to the FGS literature. They inform evidence-based policy, advocate for increased domestic and international investment in FGS programmes and provide a strong foundation to guide future research.