Comparing Active to Passive Surveillance In Philadelphia, 2005–2009: Implications for the Transition to Nationwide Passive Varicella Disease Surveillance

Background: Using active surveillance in West Philadelphia (WP), the Philadelphia Department of Public Health (PDPH) has documented substantial reductions in varicella-related morbidity since vaccine licensure. However, active surveillance is expensive and difficult to maintain.  The difference in case counts and reporting completeness between passive and active varicella surveillance remains unknown. 

Objectives: To compare detection of varicella cases from a passive surveillance system to an active system. 

Methods: For the ~300 WP active surveillance sites, PDPH mandates bi-weekly reports of varicella activity (even if none), and performs extensive investigations to collect clinical and demographic information.  In the remainder of Philadelphia, sites passively report varicella cases, and abbreviated investigations are conducted.  Varicella incidence, case-status determinations (confirmed, probable, or excluded), and completeness of investigation information were compared for 2005-2009 actively and passively received reports. Reporting completeness was assessed using capture-recapture analysis of 2-18 year old cases reported by two primary site types, schools/daycare centers and healthcare providers (HCP).

Results: In this period, PDPH received 4,068 varicella case reports (3,136 via passive surveillance and 932 through active surveillance). Most passive surveillance reports were classified as probable cases (23% confirmed, 52% probable, 25% excluded); whereas, nearly all the active surveillance reports were either confirmed or excluded (45% confirmed, 4% probable, 51% excluded). Passive surveillance had significantly higher varicella incidence for 5-18 year olds than active surveillance areas (6 versus 4 cases/1,000 persons, p<0.01). For reports from schools/daycares and HCPs, reporting completeness was higher in active than passive surveillance sites (66% versus 40%).

Conclusions: Active surveillance for varicella results in better classified cases and more accurate estimates of disease incidence.  However, additional considerations can be given to the follow up performed for passively reported cases to improve their classification.