Objective: To identify delayed prophylaxis from a pre-existing database and strategies to improve performance.
Setting: A skilled nursing facility with 14 floors (4 buildings). The "outbreak unit" was a 50- to 60-bed floor.
Methods: We performed surveillance during six seasons using one protocol. Prophylaxis was started when influenza was cultured in the building and 10% of residents on the floor had a new respiratory illness within 7 days. We defined delayed prophylaxis as four or more residents on a floor with positive cultures whose specimens had been collected within 5 days before the application of prophylaxis.
Results: We identified 14 examples of delayed prophylaxis. In three, delayed prophylaxis was related to the 3.9-day delay between culture collection and culture report There was a high degree of commonality among building attack rates within a season. During six seasons, the first case in the last building occurred 27 to 64 days after the first case in the facility. The two seasons with the greatest activity (68 and 154 cases, respectively) began with explosive, multi-floor outbreaks in a single building. The match between the circulating strain and the vaccine was good, except in 1997-1998 when there were seven examples of delayed prophylaxis.
Conclusions: Influenza may involve buildings sequentially with a commonality of building attack rates. Explosive, multi-floor outbreaks early in the season could lead to a lower threshold for prophylaxis within a larger area when initial cases are encountered later in the season. This strategy could have prevented five examples of delayed prophylaxis. Rapid testing of multiple specimens while waiting for culture confirmation could have prevented three examples of delayed prophylaxis.