Model-Based Geostatistical Design and Analysis of Prevalence for Soil-Transmitted Helminths in Kenya: Results from Ten-Years of the Kenya National School-Based Deworming Programme

Authors:

Abstract
Background: Infection caused by soil-transmitted helminths (STH) comprise the most widespread neglected tropical diseases (NTDs) globally. Kenya is endemic for STH with 27 counties and over 6 million children being at risk. In 2012, Kenya launched a National School Based Deworming Programme (NSBDP) with a goal to eliminate parasitic worms as a public health problem. In this study, we used a novel model-based geostatistical (MBG) method to design and analyse the impact of NSBDP after 10 rounds of annual deworming with a view to inform treatment strategy changes. Methods: We utilized a cross-sectional design to survey 200 schools across 27 counties from parts of seven regions of Kenya. In each school, a minimum of 70 children were randomly sampled. The selection of the 200 schools followed the MBG approach that incorporated the available historical data on treatment frequency, morbidity, environmental and other geospatial covariates to efficiently predict the helminths prevalence in Kenya. The analysis was performed using the MBG approach with the primary objective of determining the predictive probabilities of various levels of implementation units lying within pre-defined prevalence ranges. Findings: The overall NSBDP geographic area prevalence for any STH infection was estimated to sit between 2% to 0.999. Species specific thresholds were between 2% to Ascaris lumbricoides, 0% to <2% for hookworm, and 0% to 0.999. County specific thresholds showed variation, with 10 counties estimated to sit within 0% to <2% prevalence, another 10 counties within 2% to <10%, and 7 counties within 10% to <20%. The analysis was extended to determine sub-county variation in prevalence thresholds. Interpretation: Using the MBG approach, the STH treatment requirements can now be confidently refined. Following the available World Health Organization (WHO) guidelines, 10 counties may warrant suspending treatment and establish a cost-effective surveillance system, another 10 counties will require preventive chemotherapy (PC) once every two years, and the remaining seven counties will require PC once every year. This change in treatment strategy will maintain the future cost-effectiveness of deworming going forward. Funding: The financial support for this study was provided by The END Fund, GiveWell and Epic Foundation via Evidence Action. Collins Okoyo is supported by Evidence Action through Kenya Medical Research Institute as a Principal Statistician, and further supported by GlaxoSmithKline (GSK) Africa Non-Communicable Disease Open Lab through the DELTAS Africa Sub-Saharan African Consortium for Advanced Biostatistics (SSACAB) training programme, grant number D1702270-01, as a PhD fellow. Declaration of Interest: All authors declare no competing interests. Ethical Approval: Ethical approval for the study protocol was obtained from the KEMRI Scientific and Ethics Review Unit (SERU Number 2206). Permission to access the schools for the survey was obtained from the national-level Ministry of Education. At county-level, approval was provided by the respective county health and education authorities. At school-level, parental consent from parents/guardians of the children was based on passive, opt-out consent rather than written opt-in consent due to the low-risk nature of the survey procedure. Additionally, individual assent was obtained from each child before participation in the study. All data used were anonymized.