Understanding how a complex intervention works: designing large-scale vaccination programs

Haydon, Cleaveland, Kao, Hampson, Townsend

Despite possessing a sophisticated understanding of the theory underlying many aspects of the dynamics and control of infectious diseases, this knowledge is only rarely applied directly to the design and management of large-scale public health interventions. Substantial resources are routinely invested in disease control programs, but without accompanying studies that analyse how these complex interventions are really working, important opportunities are missed to learn why these programs succeed in some circumstances but not in others. As a consequence, the effectiveness and impact of these interventions can be diminished. Rabies is a horrific zoonosis responsible for an estimated 55,000 human deaths in Asia and Africa each year, mainly in children.

The Bill and Melinda Gates Foundation (BMGF) have recently provided $10M to support extensive, long-term rabies vaccination programs of domestic dogs in Tanzania, the Philippines, and the Republic of South Africa, to demonstrate the feasibility and effectiveness of dog vaccination in preventing human rabies. This intervention provides a rare and important opportunity to learn about the large-scale control of rabies, the design of efficient and effective vaccination programs, and the adaptive management of the roll-out of a substantial public health intervention. The BMGF vaccination program will generate a wealth of spatial data on dog population size and distribution, vaccination coverage, and rabies incidence, but resources do not exist within this program to undertake sophisticated in-depth analyses of these data. Here, we have assembled an experienced, multi-disciplinary team of epidemiologists, modellers, veterinarians, virologists, and public-health workers to chaperone these field vaccination programs.

The team will conduct real-time modelling and analysis of the implementation and performance of this program to learn how the intervention works, and through a process of feedback, improve its efficiency, effectiveness, and likely sustainability. While most of the primary data will be generated through the BMGF program we are requesting limited resources for the acquisition of additional data where the resulting added-value is significant - namely trace-back contact studies, and genome sequencing of selected rabies isolates to enhance our understanding of the origins and circulation of rabies virus in the study areas. Our proposed project will deliver enhanced understanding of the relationship between population size and required vaccination coverage, algorithms that prioritise the order and importance of vaccinating different parts of the study areas, the interval between pulse vaccination events, the cause of local control failures and the design of appropriate response actions.