Temporal Clusters of Salmonella Serovars in Humans in Alberta, 1990–2001

[[[ Background: ]]] Temporal clusters of salmonellosis are believed to occur but have not been quantitatively explored, thus, our objectives were to describe trends, seasonal patterns, and clusters of salmonellosis in humans in Alberta by examining isolates reported through passive surveillance systems. [[[ Methods: ]]] Cases of salmonellosis reported through Notifiable Disease Records between January 1990 and December 2001 were obtained from Alberta Health and Wellness. Least squares regression was used to characterize the distribution (long-term trends and seasonal patterns) of isolates. A cluster detection test was used to determine if and when isolates of specific serovars aggregated in time, over and above the background distribution. Comparisons were made to temporal patterns in Alberta livestock and to known outbreaks in humans. [[[ Results: ]]] S. Typhimurium, Enteritidis, Hadar, Heidelberg, and Thompson were the five most common serovars of the 9,188 isolates reported. The annual number of isolates was relatively stable over time, with a distinct summertime seasonal pattern. Clusters were observed in 23 of 32 serovars examined. More clusters occurred in September and October than in other months, and in 1998 through 2001 than in the early to mid-1990s. Also, more clusters were of short duration than long. [[[ Interpretation: ]]] Short-duration clusters likely indicate a point source of infection, while long-duration clusters may indicate exposure to a persistent common source or the occurrence of secondary infections. A sharp increasing trend and a large cluster of S. Heidelberg may have public health implications. Surveillance activities focussed on similarities between common serovars, trends, and temporal clusters in humans and animals, and studies on factors associated with autumn clusters may be useful in preventing outbreaks in humans.