Israel Medical Association Journal

Background: Syndromic surveillance systems have been developed for early detection of bioterrorist attacks, but few validation studies exist for these systems and their efficacy has been questioned.

Objectives: To assess the capabilities of a syndromic surveillance system based on community clinics in conjunction with the WSARE[1] algorithm in identifying early signals of a localized unusual influenza outbreak.

Methods: This retrospective study used data on a documented influenza B outbreak in an elementary school in central Israel. The WSARE algorithm for anomalous pattern detection was applied to individual records of daily patient visits to clinics of one of the four health management organizations in the country.

Results: Two successive significant anomalies were detected in the HMO’s[2] data set that could signal the influenza outbreak. If data were available for analysis in real time, the first anomaly could be detected on day 3 of the outbreak, 1 day after the school principal reported the outbreak to the public health authorities.

Conclusions: Early detection is difficult in this type of fast-developing institutionalized outbreak. However, the information derived from WSARE could help define the outbreak in terms of time, place and the population at risk.

Background: Each winter influenza activity is a major cause of morbidity and mortality both in Israel and worldwide.

Objectives: To identify the influenza viruses active in Israel during the winter season and to assess the extent of influenza morbidity.

Methods: Information was collected on a population of 18,684 individuals enrolled in two community clinics in central Israel. It included the total number of visits for acute respiratory infection – including influenza and influenza-like illness (ARI/flu-like) – during a 20 week surveillance period (23 November 1997 to 27 March 1998) and the percent of influenza virus isolates in nasopharyngeal specimens from a sample of patients with ARI/flu-like collected on a weekly basis during the same period.

Results: A total of 5,947 visits for ARI/flu-like were recorded among 18,684 enrolled patients in two community clinics (18.1%). The progressive increase in the number of visits for ARI/flu-like reached a peak on week 2/98 with 597 visits and a rate of 31.95 visits per 1,000 population. After this, a decrease to the initial values was evident by week 12/98. Most affected patients were in the age groups 5–14 and 65 years and over, with a rate of 733.5 and 605.3 visits per 1,000 population, respectively. Influenza virus was isolated from 92 of the 426 nasopharyngeal specimens (21.6%). The most commonly detected strain was A/Sydney/5/97(H₃N₂) like (77.2%). The peak rate of isolates was recorded at the beginning of January (01/98).

Conclusions: A/Sydney/5/97(H₃N₂) like-strain was the dominant influenza virus. Its presence did not prevent the simultaneous activity of influenza A/H₁N₁ virus. The dynamic of the clinical disease as expressed by the weekly visit rate for ARI/flu-like was similar to the temporal pattern of the virological findings. The extent of morbidity suggests moderate epidemic activity.