This study was designed to investigate the epidemiology of rabies in Thailand and to
evaluate the rabies control programme. The study involved retrospective analysis of
seven years data (1993 to 1999), concerning rabies incidence and control activity. Five
datasets from four different organisations responsible for rabies control in Thailand
were collected between March and June, 2000.
Control activity data were found to be incomplete due mainly to problems associated
with the data collection form and with methods of data storage and retrieval in
provincial offices. Recommendations are made in this thesis to modify the data
collection form and data storage methods to improve data collection and the usefulness
of information collected in the future.
The annual cumulative incidence of rabies in both animals and humans has declined
considerably between 1993 to 1999 as a result of the control programme implemented
under the Rabies Act B.E. 2535 (1992). Rabies in humans is more common in males
than females and is more common in young children (less than 10 years of age) than
any other age group. Dogs were the most commonly submitted animal for rabies
diagnosis as part of the surveillance programme, accounting for 87% of the total of
40,649 submissions. Cows were more likely to return a positive rabies diagnosis than
any other animal (57% of a total of 409 animals submitted) while dogs returned more
cases than any other animal species (43% of a total of 35,338 animals submitted).
Detailed investigations were confined to dogs. Rabies was more common in non-owned
dogs, non-vaccinated dogs, dogs showing signs consistent with furious rabies, and dogs
with a history of having attacked or bitten either other animals or both animals and
people.
The spatio-temporal pattern of rabies in both dogs and people, showed changes in the
distribution of rabies cases during the seven year period of the study. Mixed effect
Poisson regression models incorporating spatial and non-spatial random effect terms
were used to investigate risk factors associated with rabies. Factors associated with
increasing risk of rabies in people at the province level included: increasing dog
density, increasing incidence of rabies in dogs, and reduction in human population
density. In the 1999 dataset, there was evidence of spatial clustering of rabies risk. In
addition a cluster of provinces showed an elevation in rabies risk that was unexplained
by the explanatory variables included in the model. The reason for this cluster of
elevated risk remains to be elucidated.
Factors associated with an increasing risk of rabies in dogs at the province level
included increasing dog population density and increasing proportion of dogs
vaccinated against rabies. Relative risk estimates were close to one suggesting the
strength of the association for both terms was weak. Increased dog vaccination against
rabies is likely to occur in provinces with a higher risk of both animal and human
rabies.
