Heterogeneity in the spread and control of infectious disease: consequences for the elimination of canine rabies.

Elaine A Ferguson,John Friar,Ramona Consunji,Marlon Pancipane,Joji Jimenez,Katie Hampson,Sunny E Townsend,Sarah Cleaveland,Daniel T Haydon,Raffy Deray

doi:10.1038/srep18232

Abstract

Understanding the factors influencing vaccination campaign effectiveness is vital in designing efficient disease elimination programmes. We investigated the importance of spatial heterogeneity in vaccination coverage and human-mediated dog movements for the elimination of endemic canine rabies by mass dog vaccination in Region VI of the Philippines (Western Visayas). Household survey data was used to parameterise a spatially-explicit rabies transmission model with realistic dog movement and vaccination coverage scenarios, assuming a basic reproduction number for rabies drawn from the literature. This showed that heterogeneous vaccination reduces elimination prospects relative to homogeneous vaccination at the same overall level. Had the three vaccination campaigns completed in Region VI in 2010–2012 been homogeneous, they would have eliminated rabies with high probability. However, given the observed heterogeneity, three further campaigns may be required to achieve elimination with probability 0.95. We recommend that heterogeneity be reduced in future campaigns through targeted efforts in low coverage areas, even at the expense of reduced coverage in previously high coverage areas. Reported human-mediated dog movements did not reduce elimination probability, so expending limited resources on restricting dog movements is unnecessary in this endemic setting. Enhanced surveillance will be necessary post-elimination, however, given the reintroduction risk from long-distance dog movements.

Highlights

The basic reproduction number of a disease, R0, describes the mean number of secondary cases produced by a single infected individual in a fully-susceptible population, and for rabies is consistently found to lie below two, typically around 1.28,9
This study showed that the relative successes of different vaccination strategies for eliminating rabies depend on the frequency of human-mediated dog movement, and that, when vaccination coverage is spatially heterogeneous, frequent human-mediated movements reduce the probability of elimination
Vaccination coverage in Region VI at the time of the household survey was high in general, averaging 57% (95%CIs: 55.9–58.8%), though there was some variation at the province/city level (Table 1)

Summary

Introduction

The basic reproduction number of a disease, R0, describes the mean number of secondary cases produced by a single infected individual in a fully-susceptible population, and for rabies is consistently found to lie below two, typically around 1.28,9. We used these data to determine the spatial pattern of vaccination coverage and estimate human-mediated dog movement rates in the region This information was used to parameterise a spatially explicit, individual-based transmission model, previously developed to examine rabies elimination prospects[9,20], to determine the likely consequences of the vaccination campaigns given alternative dog movement rates (including those estimated here for Region VI, and those previously inferred for Bali, Indonesia[9]). We used this model to explore the interplay between control efforts, both for endemic rabies as in Region VI, and for epidemic rabies as was the case for Bali. These analyses are a vital step towards ensuring programme targets are met through adaptive management, generating recommendations to improve prospects of rabies elimination within a reasonable timeframe

Methods

Results

Conclusion