Opportunities for funding

CLOSED: We recently ran a call for short-term mathematical and statistical modelling funding through the Coalition for Operational Research on NTDs (COR-NTD). The research questions addressed urgent practical problems for lymphatic filariasis (LF) control which could be informed by mathematical and statistical modelling.

The calls closed on Wendesday 19th March, but the details are included below for reference. we hope to announce the results by Friday 18th April.

Project details:

1. Loa loa intensity modelling

Issue:   To start safe treatment programs for LF and/or hypoendemic onchocerciasis in Loa-endemic areas

Research question: Is there a community Loa prevalence at which LF MDA can be safely administered?

An important obstacle to LF program scale-up is the lack of complete information about the geographic distribution of Loa loa and its overlap with LF (and hypoendemic onchocerciasis).   Better programmatic approaches to LF and oncho elimination are urgently needed in order to reduce the risk of serious adverse events (SAE) caused by ‘inadvertent’ treatment of loiasis with ivermectin during the onchocerciasis and LF elimination programs.  These severe adverse events are associated with high levels of loa microfilaremia (>30,000 microfilariae/ml). MDA is currently being implemented in hyper and mesoendemic onchocerciasis areas where the risk of blindness is considered a counterweight to the risk of SAE when persons with high levels of Loa microfilaremia (conservatively defined as > 8,000 mf/ml) are inadvertently treated.  For areas of overlap between Loa and LF or hypoendemic onchocerciasis, a more stringent safety standard is required.

The proportion of individuals at risk of SAE (with high mf counts) increases as community microfilarial prevalence increases.  What is needed is a tool for estimating the risk of SAE from a Loa loa survey, usually from microfilarial prevalence, but, where available, with a sample of individual microfilarial loads.

Aim: This research will assist program managers in determining where and how to implement MDA safely.

Work required: A statistical and/or modelling analysis to give an estimate of the proportion of the population with high Loa loa mf intensities given either the prevalence of mf in the community, or a sample of individual mf counts. The African Program of Onchocerciasis Control (APOC) has collected and published data on the distribution of mf counts in communities that differ by mf prevalence that will be used for these analyses under a suitable data-sharing agreement.  Other data that might be useful should be sought and utilized as well.

2. Estimating expected post-MDA prevalence

Issue:   To monitor LF programs for evidence of less than expected efficacy, due to reduced drug efficacy, coverage or other issues.

Research question: Can existing prevalence  data from sentinel sites be used to predict mf prevalence after a given number of rounds of MDA as a tool to identify program settings in which the response to MDA is less than predicted?

The scale-up of NTD programs is subjecting human parasites to unprecedented levels of drug pressure, leading to real concerns about the potential for emergence of drug resistance. The WHO NTD Working Groups have identified monitoring drug efficacy as a critical, but neglected, component of routine program monitoring and evaluation. The African Programme for Onchocerciasis Control (APOC) is currently using a model to determine whether or not declines in community onchocerciasis microfilarial (mf) prevalence following MDA are line with predictions (e.g. Habbema et al, Parasitology Today, 1992).  This approach is being used as an “alarm bell” to define program settings where the impact of MDA has been less than expected and should be further investigated to determine if, among other potential causes, drug efficacy is sub-optimal. 

Aim: This research will allow programs to identify areas where the MDA program is not working as expected.

Work required: Existing WHO (or other) sentinel site data will need to be aggregated and used to develop these models under a suitable data-sharing agreement. A transmission model will be fitted and validated against the sentinel site data and will be used to provide predictions for future surveillance data for these sentinel sites. Two treatment efficacies are needed because of the use of two different drug combinations (ivermectin/albendazole and diethylcarbamazine/albendazole).

3. Post-MDA surveillance

Issue:   To understand surveillance needs for LF programs after the end of mass drug administration (MDA) programs.

Research question: Can risk maps be used as a tool to define surveillance needs for the global LF program? 

Many LF programs are now scaling down MDA and beginning a period of post-MDA surveillance that will culminate with a request of WHO to verify the elimination of LF.  Although the specifics of the verification process are still being determined, countries will be requested to provide surveillance data, both for areas where MDA was administered and for areas judged to be non-endemic at the start of the program.  From the standpoint of practicality, the level of surveillance needed should be tailored to the level of transmission risk.  Risk might be influenced by the vector, local ecology, proximity to areas of high transmission and population density, among other factors.  Several teams have developed spatial models of the distribution of LF.  Could these maps be adapted to stratify districts by risk?  If so, this would establish a framework for an empirical comparison of surveillance strategies.

Aim: This research will provide a theoretical evidence base for designing detailed surveillance strategies programs based on risk of LF transmission.

Work required: Linking maps of LF risk and prevalence to a stratification of risk of recrudescence and reintroduction, which will subsequently be used to design surveillance strategies. It is expected that the applicants will use a combination of geostatistical covariates (including vector types, coverage with other interventions) and mathematical transmission models to identify areas of higher or lower risk. Ideally the model would provide a stratification of risk into 3-4 groups according to the estimated prevalence from mapping.