Table 1.
Description of state parameters where i is 1 or 2.
Table 2.
Rate parameter descriptions, values used and ranges for model versions based on heterogeneity, but no animal reservoir.
Fig 1.
Overview of the population structure and compartments of the model.
A): Human populations are divided in a stationary (Nh1) population that remains in low exposure habitats (e.g., a village), and a smaller population (Nh2) which commute and spend a proportion ξ of their time in a potentially high exposure setting (e.g., a plantation). Each of these habitats harbours tsetse (Nv1 and Nv2) and non-human vertebrate animal populations (Na1 and Na2) of varying sizes and characteristics. B): Compartmental diagram highlighting the transmissions between states of infection of the animal, tsetse, and human populations in the high exposure area 2. A similar diagram explains transmission in area 1, although there both human populations are exposed to tsetse bites. Solid lines depict transitions between compartments, while dashed lines represent transmission rates.
Fig 2.
Global sensitivity of sleeping sickness prevalence to model parameters without (left panel) and with the possibility of animal-tsetse transmission of T.b. gambiense (right panel).
Descriptions of parameters are provided in Table 2. The most important parameters without an animal reservoir were the ratio of commuters to non-commuters (N2/N1), the biting preference of tsetse for animals in the non-commuting area, σa1, and the proportion of infectious bites that lead to infection in hosts, b. With an animal reservoir, the most important parameters were b, the tsetse to human ratio in the focal (1st) area, V/H1, and the biting preference for humans, σh.
Fig 3.
Zero-growth isoclines (R0 = 1) of T.b. gambiense under perturbation of specific parameters.
The parameter values used were the median values obtained for the high transmission setting, except for those varied in the analysis. In the left plot, isoclines at different levels of vector mortality are shown, depending on the daily removal rate of infected humans (r) and the proportion of time commuters spend in the high exposure area (ξ). The areas above the isoclines represent values of R0 greater than 1, and below and to the right of the isoclines values smaller than 1. In the middle plot the impact of screening humans in the low risk setting (ra) in combination with screening commuting humans (rb) is shown for different levels of animal to human ratios (A/H1). In the right plot isoclines are depicted along removal rates (r) and tsetse density (V/H) in both areas when animals either do not contribute to transmission (ca = 0) or they can infect tsetse (ca = 0.003).
Fig 4.
Median (lines) and 95th percentiles (shaded areas) of simulations on the impact of interventions on prevalence over time in high transmission settings without (left) and with (right) animal-tsetse transmission, assuming a range of efficacies for screen & treat (solid line) and screen & treat with vector control (dashed line).
Fig 5.
The proportion of simulations where HAT was eliminated (prevalence < 1 x 10−6) and the mean time to elimination with standard deviation, depending on the percentage of the human population screened per year (expressed as absent (0%), low (20–40%), moderate (40–60%), or high (60–80%) coverage and indicated by the plot titles), with varying levels of vector control (expressed as additional vector mortality, indicated by the symbols in the legend), for a high transmission setting without an animal reservoir (A) and with an animal reservoir (B).
Fig 6.
The proportion of simulations where HAT was eliminated (i.e., prevalence < 1 x 10−6) and the mean time to elimination, depending on the percentage of the human population screened per year, with varying levels of vector control (expressed as additional vector mortality, indicated by the symbols in the legend), for a moderate transmission setting without an animal reservoir (A) and with an animal reservoir (B).
Fig 7.
The proportion of simulations where HAT was eliminated (i.e., prevalence < 1 x 10−6) and the mean time to elimination, depending on the percentage of the human population screened per year, with varying levels of vector control (expressed as additional vector mortality, indicated by the symbols in the legend), for a low transmission setting without an animal reservoir (A) and with an animal reservoir (B).