Clustering of reported cases of leptospirosis among dogs in the United States and Canada.

A retrospective study was conducted to determine whether cases of leptospirosis diagnosed among dogs in the United States and Canada were clustered in time, space, or time-and-space. The study population was estimated--based on first examination of each dog--to consist of 1,035,366 dogs examined at 22 veterinary teaching hospitals between 1983 and 1998. The veterinary medicine database was searched for records of dogs in which a diagnosis of leptospirosis was made during the study period. The scan statistic (Poisson model) was used to examine for clustering of cases, using recorded date of diagnosis and the location of the veterinary teaching hospital at which the diagnosis was made. Three-hundred and forty dogs diagnosed with leptospirosis were identified. Cases were clustered in the period of 1992-1998. A cluster of dogs diagnosed at three veterinary teaching hospitals located in the midwest of the United States was identified. The composition of this spatial cluster was influenced by the age and gender distribution of the population at-risk. A cluster of cases involving 143 dogs diagnosed with leptospirosis between 9 April 1993 and 14 December 1998 at six veterinary teaching hospitals located in the midwest of the United States was identified. This time-space cluster was not influenced by the age, gender or breed distribution of the population of dogs presented at veterinary teaching hospitals during the study period. Leptospirosis was clustered during the 1990s in the midwest of the United States. Dog risk factors (age, gender, breed) do not explain the clustering that was observed. Leptospira serovar or environmental factors might be responsible for the clustering of canine leptospirosis in the United States and Canada.