A Novel Method for Surveillance of Novel H1N1 Influenza Using Automated Hospital Laboratory and Pharmacy Data

Background: Surveillance for influenza activity is mostly based on public health information, which is relatively broad, regional data and is usually delayed by several days before it is available to the public. 

Objective: We sought to identify “real time” surveillance methods based on hospital laboratory and pharmacy data for influenza. The H1N1 influenza pandemic in the spring 2009 presented a unique opportunity to analyze these methods.

Methods: The Detroit Medical Center (DMC) health system consist of 8 hospitals, has over 2,000 beds, and serves as a tertiary referral hospital for metropolitan Detroit and southeastern Michigan. During the initial outbreak of novel H1N1 influenza from April 2009–July 2009, Infection Prevention at DMC received weekly reports regarding utilization and results of the rapid EIA flu test and utilization of oseltamivir (Tamiflu").  We aimed to analyze the relationship between rapid EIA testing, Tamiflu dispensing and influenza-like illness (ILI) as reported by public health (Michigan Department of Community Health).

Results: During the study period, a total of 1879 EIA tests were performed and only178 were positive (9.5%) for Influenza A/B. There were 223 doses of Tamiflu dispensed for both treatment and prophylaxis. The peak period for EIA testing and Tamiflu dispensing occurred in late May and early June (Figure). The level of Tamiflu prescribing and EIA testing correlated closely with public health ILI surveillance.

Conclusions: Surveillance for novel H1N1 influenza can be easily performed using hospital data pertaining to influenza testing and Tamiflu prescribing.  These types of surveillance data are accurate and utilize readily available automated hospital data.