Socio-spatial heterogeneity in participation in mass dog vaccination campaigns, Arequipa, Peru

Abstract

To control and prevent rabies in Latin America, mass dog vaccination campaigns (MDVC) are implemented mainly through fixed-location vaccination points: owners have to bring their dogs to the vaccination points where they receive the vaccination free of charge. Dog rabies is still endemic in some Latin-American countries and high overall dog vaccination coverage and coverage evenness are desired attributes of MDVC to halt rabies virus transmission. In Arequipa, Peru, we conducted a door-to-door post-campaign survey on >6,000 houses to assess the placement of vaccination points on these two attributes. We found that the odds of participating in the campaign decreased by 16% for every 100 m from the owner’s house to the nearest vaccination point (p=0.041) after controlling for potential covariates. We found social determinants associated with participating in the MDVC: for each child under 5 in the household the odds of participating in the MDVC decreased by 13% (p=0.032), and for every ten more years living in the area, the odds of participating in the MDVC increased by 9% (p<0.001), after controlling for distance and other covariates. We also found significant spatial clustering of unvaccinated dogs over 500m from the vaccination points, which created pockets of unvaccinated dogs that may sustain rabies virus transmission. Understanding the barriers to dog owners’ participation in community-based dog-vaccination programs will be crucial to implement effective zoonotic disease preventive activities. Spatial and social elements of urbanization play an important role in coverage of MDVCs and should be considered during their planning and evaluation.

Development, feasibility and potential effectiveness of community-based continuous mass dog vaccination delivery strategies: lessons for optimization and replication

“…Several local environmental, economic and sociocultural events affected feasibility of delivering CBC-MDV components. Structural community participation in initializing and implementing the intervention would clearly be valuable in identifying these events and issues, and replication of CBC-MDV across wider contexts would benefit from tailoring campaign schedules to local environmental and social events [36,37,39,40] or calendars.…”

“…However, because the vaccination activity of Strategy 1 occurs at a central point of the village, for many owners this Strategy likely posed a challenge of access as they will be required to travel further to reach the central point where the clinic was hosted. Living far from the point of the clinic is commonly cited as a reason for nonparticipation in other studies[34][35][36][37]. In comparison, Strategy 2, being hosted at the subvillage level, comes with a relatively lighter workload on each vaccination day for the…”

Spatial Association of Canine Rabies Outbreak and Ecological Urban Corridors, Arequipa, Peru

“…We used a Monte Carlo random-labeling simulation with 1000 trials to compare the average distance between observed rabid dogs and water channels to the distance from water channels of a randomly-generated spatial sample of households [ 20 ], matched on locality. We have found that the number of owned dogs per house can vary by locality [ 7 ], and we also wanted to take into account any unobserved difference in surveillance intensity at the locality level. Localities are informal district subdivisions which are widely used by the health inspectors to differentiate their catchment areas and organize their work.…”

“…Localities are informal district subdivisions which are widely used by the health inspectors to differentiate their catchment areas and organize their work. Under the null hypothesis, we assumed that all households are equally likely to have a rabid dog regardless of their distance to the water channels—on a door-to-door survey conducted in more than 4000 houses by our team, we did not find an association between the number of owned dogs per house and the water channels [ 7 ]. To determine whether the observed distance of rabid dogs to water channels was statistically different from the distance of the water channel to the random spatial sample, we compared the average observed distance to the distance distribution derived from the Monte Carlo trials.…”

“…In Arequipa, where 22% of owned dogs have free access to the street all the time, and 48% of owned dogs are unrestricted at some point during the day [ 7 ], it is unknown which proportion of dogs seen on the streets are actually stray dogs and which are owned. Upon detection of the first rabid dog, Ministry of Health (MOH) authorities attempted containment by ring vaccination of an approximately five-block radius, in accordance with the Peruvian technical guidelines to control and prevent human rabies [ 8 ].…”