Novel surveillance system based on school-nurse visits used to detect trends in influenza-like illness — Washington DC, 2004–05

Background: Student absenteeism has been used in communities to monitor influenza activity trends, but absenteeism data are not routinely reported to the District of Columbia Department of Health (DCDOH). In 2004, DCDOH expanded surveillance by using reports of school-nurse visits, including influenza-like illness (ILI), to improve early outbreak detection. We assessed whether these data can be used as an indicator of community influenza activity. Methods: We compared peaks in school-nurse ILI visits during the 2004–05 influenza season with student absenteeism, emergency department (ED) and clinic ILI visits, and positive clinical specimens for influenza. ILI was defined as fever, sore throat, headache, or respiratory illness identified by school nurses, and fever with sore throat or cough by clinicians in EDs and clinics. We used a Statistical Process Control method to identify activity peaks. Results: Thirty-four/151 (23%) schools reported nurse visits. A major ILI peak occurred in mid-February and a minor peak in mid-December. The mid-February peak was concurrent with peaks in laboratory-confirmed influenza A and clinic visits for ILI. Two peaks in ED ILI visits were identified; both occurred 2 weeks after peaks in nurse visits. No peaks in student illness absenteeism were identified. Conclusions: The major mid-February peak in school-nurse ILI visits was probably caused by influenza A; the cause of the mid-December peak is uncertain. Used together, increases in school-nurse ILI visits and laboratory-confirmed influenza might help DCDOH better monitor community influenza activity. Respiratory virus testing of symptomatic students at nurse visits might improve system specificity and usefulness.