Epidemiological assessment of eight rounds of mass drug administration for lymphatic filariasis in India: implications for monitoring and evaluation.

BACKGROUND: Monitoring and evaluation guidelines of the programme to eliminate lymphatic filariasis require impact assessments in at least one sentinel and one spot-check site in each implementation unit (IU). Transmission assessment surveys (TAS) that assess antigenaemia (Ag) in children in IUs that have completed at least five rounds of mass drug administration (MDA) each with >65% coverage and with microfilaria (Mf) levels <1% in the monitored sites form the basis for stopping the MDA. Despite its rigour, this multi-step process is likely to miss sites with transmission potential ('hotspots') and its statistical assumptions for sampling and threshold levels for decision-making have not been validated. We addressed these issues in a large-scale epidemiological study in two primary health centres in Thanjavur district, India, endemic for bancroftian filariasis that had undergone eight rounds of MDA. METHODOLOGY/PRINCIPAL FINDINGS: The prevalence and intensity of Mf (per 60 µl blood) were 0.2% and 0.004 respectively in the survey that covered >70% of 50,363 population. The corresponding values for Ag were 2.3% and 17.3 Ag-units respectively. Ag-prevalence ranged from 0.7 to 0.9%, in children (2-10 years) and 2.7 to 3.0% in adults. Although the Mf-levels in the survey and the sentinel/spot check sites were <1% and Ag-level was <2% in children, we identified 7 "residual" (Mf-prevalence ≥ 1%, irrespective of Ag-status in children) and 17 "transmission" (at least one Ag-positive child born during the MDA period) hotspots. Antigenaemic persons were clustered both at household and site levels. We identified an Ag-prevalence of ~1% in children (equivalent to 0.4% community Mf-prevalence) as a possible threshold value for stopping MDA. CONCLUSIONS/SIGNIFICANCE: Existence of 'hotspots' and spatial clustering of infections in the study area indicate the need for developing good surveillance strategies for detecting 'hotspots', adopting evidence-based sampling strategies and evaluation unit size for TAS.

Mathematical models for lymphatic filariasis transmission and control: Challenges and prospects.

BACKGROUND: Mathematical models developed for describing the dynamics of transmission, infection, disease and control of lymphatic filariasis (LF) gained momentum following the 1997 World Health Assembly resolution and the launching of the Global Programme to Eliminate Lymphatic Filariasis (GPELF) in 2000. Model applications could provide valuable inputs for making decisions while implementing large scale programmes. However these models need to be evaluated at different epidemiological settings for optimization and fine-tuning with new knowledge and understanding on infection/disease dynamics. DISCUSSION: EPIFIL and LYMFASIM are the two mathematical simulation models currently available for lymphatic filariasis transmission and control. Both models have been used for prediction and evaluation of control programmes under research settings. Their widespread application in evaluating large-scale elimination programmes warrants validation of assumptions governing the dynamics of infection and disease in different epidemiological settings. Furthermore, the predictive power of the models for decision support can be enhanced by generating knowledge on some important issues that pose challenges and incorporating such knowledge into the models. We highlight factors related to the efficacy of the drugs of choice, their mode of action, and the possibility that drug resistance may develop; the role of vector-parasite combinations; the magnitude of transmission thresholds; host-parasite interactions and their effects on the dynamics of infection and immunity; parasite biology, and progression to LF-associated disease. SUMMARY: The two mathematical models developed offer potential decision making tools for transmission and control of LF. In view of the goals of the GPELF, the predictive power of these models needs to be enhanced for their wide-spread application in large scale programmes. Assimilation and translation of new information into the models is a continuous process for which generation of new knowledge on a number of uncertainties is required. Particularly, a better understanding of the role of immune mechanisms in regulating infection and disease, the (direct or immune mediated) mode of action of current drugs, their effect on adult worms, their efficacy after repeated treatment, and the population genetics of drug resistance are important factors that could make the models more robust in their predictions of the impact of programmes to eliminate LF. However, if these models are to be user-friendly in the hands of programme managers (and not remain as research tools), it would be necessary to identify those factors which can be considered as the minimum necessary inputs/outputs in operational settings for easy evaluation and on-site decision making.