Effectiveness of community-based burden estimation to achieve elimination of lymphatic filariasis: a comparative cross-sectional investigation in Côte d’Ivoire

Abstract

For lymphatic filariasis (LF) elimination, endemic countries must document the burden of LF morbidity (LFM). Community-based screening (CBS) is used to collect morbidity data, but evidence demonstrating its reliability is limited. Recent pilots of CBS for LFM alongside mass drug administration (MDA) in Cote dIvoire suggested low LFM prevalence (2.1-2.2 per 10,000). We estimated LFM prevalence in Bongouanou District, Cote dIvoire, using a comparative cross-sectional design. We compared CBS implemented independently of MDA, adapted from existing Ministry of Health protocols, to a population-based prevalence survey led by formally trained nurses. We evaluated the reliability of case identification, coverage, equity, and cost of CBS. CBS identified 87.4 cases of LFM per 10,000; the survey identified 47.5 (39.4-56.3; prevalence ratio [PR] 1.84; 95% CI 1.64-2.07). CBS identified 39.7 cases of suspect lymphoedema per 10,000; the survey confirmed 35.1 (29.2-41.5) filarial lymphoedema cases per 10,000 (PR 1.13 [0.98-1.31]). CBS identified 100.3 scrotal swellings per 10,000; the survey found 61.5 (55.5-67.8; PR 1.63 [1.41-1.88]); including 26.6 (21.5-32.4) filarial hydrocoele per 10,000 (PR of suspect to confirmed hydrocele 3.77 [3.12-4.64]). Positive predictive values for case identification through CBS were 64.0% (54.5-72.8%) for filarial lymphoedema; 93.2% (88.5-96.4%) for scrotal swellings; and 33.3% (26.4-40.8%) for filarial hydrocoele. Households of lower socioeconomic status and certain minority languages were at risk of exclusion. Direct financial costs were $0.17 per individual targeted and $69.62 per case confirmed. We provide our CBS toolkit. Our community-based approach to LFM burden estimation appears scalable and provided reliable prevalence estimates for LFM, scrotal swellings and LF-lymphoedema. The results represent a step-change improvement on CBS integrated with MDA, whilst remaining at programmatically feasible costs. Filarial hydrocoele cases were overestimated, attributable to the use of case definitions suitable for mass-screening by informal staff. Our findings are broadly applicable to countries aiming for LF elimination using CBS.