Fig 1.
Model-predicted association between annual biting rate, prevalence of skin microfilariae, and stability of transmission.
Bullets represent the average skin mf prevalence over 150 repeated simulations, with the shape of the bullet indicating the extinction probability (here defined as the proportion of repeated simulations in which transmission spontaneously faded out within 200 years). The red and purple lines (with k = 3.5 or 1.0 and one fly population) represent transmission scenarios with homogeneous mixing; the other coloured lines represent transmission scenarios with assortative mixing, assuming presence of two fly populations where some proportion fH of the human population with relative high exposure to flies has most of its contact with the fly population H. Parameter s represents the level of segregation of the two fly populations, e.g. s = 0 represents homogeneous mixing (presence of two populations but all humans have equal opportunity to be exposed to both) and s = 1 represents two completely segregated fly populations for which the biting affects two completely segregated human populations. See Methods section for details.
Fig 2.
The influence of mixing patterns on trends in prevalence of skin microfilariae during mass drug administration.
Lines represent results repeated simulations for a fixed annual biting that was tuned (given exposure heterogeneity k and assumed mixing pattern) to result in an average pre-control prevalence of about 50% in the population of age 5 and above. In each simulation, 7 mass drug administration (MDA) rounds are implemented at 65% coverage of the general population. Participation to MDA was assumed to be semi-systematic (some individuals are structurally more likely to participate that others). S4 Fig illustrates similar results for other pre-control endemicity levels and assumed patterns in MDA participation.
Table 1.
Impact of mixing patterns on probability of elimination.
Fig 3.
Pre-control arithmic mean density of mf in the skin among mf-positive individuals of age 5+ and above.
Lines are based on a generalised additive model with integrated smoothness estimation, fitted to predicted mf prevalences and intensities of repeated simulations for the same range and values of annual biting rate (ABR) used in Fig 1. For each value of ABR 150 repeated simulations were performed. Individual simulation results and the fit of the generalised additive model can be found in S5 Fig. Note that in all scenarios a mean microfilarial density below ~15 mf/ss in the mf-positive population is indicative of stochastic fade-out taking place (i.e. as incidence declines the worm population ages, resulting in lower mf production per female worm).
Fig 4.
Distribution of skin microfilarial density in mf-positive individuals in different transmission scenarios and endemicity levels.
Distributions are based on the average of 150 repeated simulation for each of three fixed values of the annual biting rate that result in an average pre-control mf prevalence of 40%, 50%, and 60% in the population of age 5 and above (three panels).
Fig 5.
Ratio of larval infection intensity in two spatially separate samples of blackflies around a single community as an indicator of assortative mixing.
Each bullets represents the result of a single simulation. Simulations were run using the same range and values of annual biting rate (ABR) as used in Fig 1, and for each value of ABR 150 repeated simulations were performed. For comparison, a ratio close to 1.0 (horizontal dashed black line) would indicate that flies from two spatially separate samples bite humans with a similar distribution of infection levels (i.e. under the assumption of homogeneous mixing). Lines are based on a generalised additive model with integrated smoothness estimation, fitted to individual bullets.