Supporting rabies elimination with evidence: Quantifying domestic dog populations, ecology and vaccination economics

Rabies remains a major public health threat in many low and middle-income countries, with domestic dogs as the principal reservoir. Effective control requires sound planning, implementation, and monitoring, based on reliable data on dog populations, their movements, and vaccination costs. This thesis had three objectives: to estimate the size of the domestic dog population in Machakos County, Kenya, to characterise movement ecology across areas with different human population densities, and to evaluate the implementation and costs of mass dog vaccination to inform county-level planning and long-term sustainability. I used household surveys, photographic sight–resight methods, GPS tracking, and economic modelling. The cross-sectional household survey estimated an owned dog population of 421,079, nearly double prior national projections, implying greater resource needs for control and a stronger case for education on responsible ownership. In a rapidly urbanising area, the free-roaming dog population was estimated at 3,807 dogs (95% CI: 1,780–5,833) when extrapolated to the total administrative area, and 3,150 dogs (95% CI: 1,473–4,826) when extrapolated to the inhabited area, with a density of 108 dogs per km² and a human-to-dog ratio ranging from approximately 1:24 to 1:30. Household data indicated that most free-roaming dogs were owned, highlighting the importance of improved confinement and responsible ownership. To assess spatial behaviour, 100 free-roaming dogs were GPS-tracked in low-density and high-density settings. Dogs in low-density areas had larger home ranges, while those in high-density areas exhibited more spatially constrained movement but slightly higher movement intensity. Movement patterns were primarily influenced by population density and age. These differences suggest that in rural settings, vaccination campaigns must cover larger geographic areas to reach widely roaming dogs, while in high-density settings, effective coverage can be achieved within smaller, more localised areas. The cost evaluation drew on a four-year county campaign that vaccinated 136,143 dogs. The mean cost per dog vaccinated was USD 2.41. Coordination, vaccines, and personnel were the main cost drivers. A Monte Carlo simulation estimated that achieving 70% vaccination coverage across the county would require a median annual cost of KES 47.2 million (approximately USD 0.36 million), exceeding the current Machakos County veterinary budget of KES 35.5 million. Across alternative operational scenarios, total campaign costs remained broadly similar (approximately KES 55 million), indicating that changes in campaign design had a limited impact on overall expenditure. However, substantial differences were observed in workforce requirements. A 360-day delivery schedule reduced the number of teams required by approximately 75% compared with a 60-day intensive campaign, offering a more feasible approach where workforce capacity is constrained. Vaccine donation reduced total programme costs by 24.5%, representing a substantial reduction in overall expenditure, with delivery-related components, particularly personnel and transport, continuing to account for a large share of total costs. Overall, the thesis demonstrates the value of integrating demographic, ecological, and financial evidence to plan rabies control efforts. I provide locally grounded estimates of owned dog population, movement patterns that guide spatial targeting, and realistic costs for campaign planning. These insights support scalable, context-responsive strategies for Machakos County and similar endemic settings.