Japan has been free from rabies since the 1950s. However, during the early 1900s several large-scale epidemics spread throughout the country. Here we investigate the dynamics of these epidemics between 1914 and 1933 in Osaka Prefecture, using archival data including newspapers. The association between dog rabies cases and human population density was investigated using Mixed-effects models and epidemiological parameters such as the basic reproduction number (R0), the incubation and infectious period and the serial interval were estimated. A total of 4,632 animal rabies cases were reported, mainly in dogs (99.0%, 4,584 cases) during two epidemics from 1914 to 1921, and 1922 to 1933 respectively. The second epidemic was larger (3,705 cases) than the first (879 cases), but had a lower R0 (1.50 versus 2.42). The first epidemic was controlled through capture of stray dogs and tethering of pet dogs. Dog mass vaccination began in 1923, with campaigns to capture stray dogs. Rabies in Osaka Prefecture was finally eliminated in 1933. A total of 3,805 rabid dog-bite injuries, and 75 human deaths were reported. The relatively low incidence of human rabies, high ratio of post-exposure vaccines (PEP) and bite injuries by rabid dogs (minimum 6.2 to maximum 73.6, between 1924 and 1928), and a decline in the proportion of bite victims that developed hydrophobia over time (slope = -0.29, se = 3, p < 0.001), indicated that increased awareness and use of PEP might have prevented disease. Although significantly more dog rabies cases were detected at higher human population densities (slope = 0.66, se = 0.03, p < 0.01), there were fewer dog rabies cases detected per capita (slope = -0.34, se = 0.03, p < 0.01). We suggest that the combination of mass vaccination and restriction of dog movement enabled by strong legislation was key to eliminate rabies. Moreover, the prominent role of the media in both reporting rabies cases and efforts to control the disease likely contributed to promoting the successful participation required to achieve rabies elimination.

Rabies was eliminated from Japan in 1957. In the 60 years since elimination, vaccination coverage has declined and dog ownership habits have changed. The purpose of this study was to assess the current risk of rabies spread in Japan. A spatially explicit transmission model was developed at the 1 km2 grid scale for Hokkaido and Ibaraki Prefectures. Parameters associated with dog movement and bite injuries were estimated using historical records from Japan, and were used with previously published epidemiological parameters. The final epidemic size, efficacy of rabies contingency plans and the influence of dog owner responses to incursions were assessed by the model. Average outbreak sizes for dog rabies were 3.1 and 4.7 dogs in Hokkaido and Ibaraki Prefectures, respectively. Average number of bite injury cases were 4.4 and 6.7 persons in Hokkaido and Ibaraki Prefectures, respectively. Discontinuation of mandatory vaccination increased outbreak sizes in these prefectures. Sensitivity analyses showed that higher chance of unintentional release of rabid dogs by their owners (from 0.5 to 0.9 probability) increased outbreak size twofolds. Our model outputs suggested that at present, incursions of rabies into Japan are very unlikely to cause large outbreaks. Critically, the reaction of dog owners to their dogs developing rabies considerably impacts the course of outbreaks. Contingency measures should therefore include sensitisation of dog owners.

No abstract

Rabies is a lethal zoonotic disease mainly transmitted to humans by dog bites. The purpose of this study was to assess the efficacy of rabies control policies in Japan, which resulted in the elimination of the disease from the country in 1957. Using historical records from the Kanto region (Chiba, Kanagawa, Saitama and Tokyo Prefectures) between 1947 and 1956 where the final canine cases were recorded, we undertook a descriptive epidemiological study, applying spatio‐temporal scan statistics using SaTScan and estimating the effective reproduction number (Rt) for the clusters and each prefecture using the growth rates. There were 1,567 dog rabies and 161 human rabies cases recorded during this period. Vaccination coverage in registered dogs was over 70% after 1951, with much lower coverage in free‐roaming and unregistered dogs. Eight clusters of dog rabies cases were identified: the first appeared in 1947 in Tokyo and was linked to three further clusters in peripheral prefectures between 1947 and 1951. Three more clusters occurred in Tokyo again between 1952 and 1954, and the last cluster was in Tokyo and Kanagawa between 1955 and 1956. Rt in the first cluster was 1.68, and Rt values in the others ranged between 1.18 and 1.86, with an exception of 4.05 in the smallest cluster in Tokyo in 1952 (10 cases). The moving average of Rt coincided with the clusters. As dog vaccination and dog management progressed, and the number of dog rabies cases declined, the moving average of Rt declined to below 1. Delays in the implementation of dog management policies in Kanagawa may have prolonged this last outbreak. These results demonstrate the effectiveness of coordinated control policy involving dog vaccination and management of free‐roaming dog populations for rabies elimination.