Measuring, monitoring and improving mass dog vaccination programmes to control and eliminate rabies

Rabies is an acute viral infection which causes horrifying neurological symptoms that inevitably result in death. Every year at least 59,000 people are estimated to die from rabies and more than 10 million are treated with post-exposure prophylaxis (PEP). Over 90% of human rabies deaths occur in Asia and Africa following bites from domestic dogs. Although human rabies deaths are 100% preventable through the delivery of prompt PEP to bite victims following a bite, PEP is not accessible to many poor rural victims, most of whom subsist on less than US$1.25/day. Empirical and theoretical evidence shows that mass dog vaccination that reaches 70% of susceptible dog population can interrupt the transmission cycle. Rabies has been eliminated from industrialized countries through mass dog vaccination, and the continent-wide elimination of canine rabies from the Americas is now within reach. In contrast, no effective large-scale control of dog rabies has been achieved in Africa and information is still needed to optimise and sustain dog vaccination programmes. The aim of this thesis was to evaluate the rabies control programme in Tanzania. This thesis is presented as a series of three standalone chapters (Chapters 2-4) that are introduced and then summarised by a general introduction (Chapter 1) and a general discussion (Chapter 5) respectively. Achieving high coverage is the most important aim of any vaccination programme; however, assessing the vaccination coverage achieved is often neglected in rabies endemic countries. In Chapter 2, I compare three methods of measuring vaccination coverage (post-vaccination transects, school-based surveys, and household surveys) across 28 districts in different settings in southeast Tanzania and Pemba island in order to determine which is most precise method. These approaches were explored in detail in a single district in northwest Tanzania (Serengeti), where their performance in producing precise estimates of coverage was compared with