Table 1.
Parameters and the baseline values used in the rabies transmission model.
Fig 1.
Structure of the canine rabies transmission model.
(A) Example of the geographical distribution of buildings and roads in an area of 2x2 km2. Owned dogs reside only in G1 buildings, while unowned free-roaming dogs may occupy either G2 or G3 buildings. (B) Schematic of the transmission model. Based on the infection status, the model classifies each dog into susceptible (S), exposed (E), infectious (I), and vaccinated (V) classes. The dashed arrow represents transmission events, and the solid arrows indicate transitions between compartments. (C) Illustration of the probability of finding a dog at distance d from their home location, P (d), and the unnormalized encountering rate between dog i and dog j, Kij. (D) An illustrative example of the unnormalized encountering rate between a rabid dog residing at the centered black point and susceptible dogs living one kilometer apart. The blue and green dots on the map represent the home locations of susceptible dogs that are located near roads and far from roads, respectively. The unnormalized encountering rates are indicated by the colors of the dot circumferences.
Fig 2.
Study sites and geographical distribution of buildings and roads.
(A) The locations of the study sites, Hatyai and Tepha district, in Thailand. (B and C) Spatial distribution of buildings (green patches) and roads (black lines) across Hatyai and Tepha district. The base layer of the map was obtained from https://data.humdata.org/dataset/thailand-administrative-boundaries.
Fig 3.
Building density and distribution.
(A and B) Spatial distribution of building density in Hatyai and Tepha. The color bar indicates the building density in the unit of buildings per square kilometer. Note that the color bar in Hatyai represents 10 times higher building density than in Tepha. The black circles show the centroid points of the building distribution. (C) The median pairwise distance between two buildings with a different rank of closeness (from shortest to longest) in Hatyai (HY) and Tepha (TP) district. The inset illustrates an example of a building arrangement with different ranks of closeness relative to the stared building. (D) The distribution of the shortest building-to-road distance in Hatyai (HY) and Tepha (TP).
Fig 4.
Effect of the geographical distribution of buildings and roads on the rabies transmission dynamics.
(A) Comparison of the cumulative number of locally transmitted rabid dogs obtained from the simulations cases simulated in both Hatyai and Tepha and the five-year averaged reported data from Hatyai (years 2016–2020). (B) The number of cumulative locally transmitted cases, and (C) the number of cumulative imported cases in Hatyai (HY) and Tepha (TP).
Fig 5.
Geotemporal spreading pattern of canine rabies in Hatyai and Tepha.
The snapshots showing the spreading patterns of rabies, averaged from 100 simulations, in Hatyai (A) and Tepha (B) at 0, 30, 90, 120, 240, and 365 days after the introduction of an index rabid-dog. The greyscale indicates the density of buildings (buildings/km2), and the warm-color scale indicates the density of the cumulative number of rabid dogs (dogs/ km2) in each cell.
Fig 6.
Impact of intervention strategies on the likelihood of rabies transmission in Hatyai (A, B, and C) and Tepha (D, E, and F). The likelihood of rabies transmission caused by an imported infected dog under the intervention scenarios of (A and D) reducing the dog population size, (B and E) reducing the fraction of free-roaming dogs, and (C and F) increasing the mass vaccination coverage. A sensitivity analysis where the dog traveling distances are scaled by factors of 0.5 (x0.5) and 2 (x2) was also shown (dashed lines).