Estimating Influenza-Associated Deaths and Hospitalizations for the Same U.S. Catchment Areas by Using Negative Binomial Regression Models

• Activity Number: 139
• Type: Contributed
• Date/Time: Monday, July 30, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Po-Yung Cheng*+ and Hong Zhou and Lynnette Brammer and David Shay
• Companies: Battelle - Contractor for CDC and Columbus Technologies and Services, Inc. and CDC and CDC
• Keywords: negative binomial regression ; influenza ; pneumonia ; death ; hospitalization

Abstract

The disease burden of influenza is commonly assessed by estimating deaths and hospitalizations associated with virus circulation. We estimated influenza-associated deaths and hospitalizations using complete mortality data and hospitalization data for 13 states from 1993-1994 through the 2005-2006 respiratory seasons. We made estimates for 5 age groups (< 5, 5-19, 20-64, 65-74, and 75+ years) using negative binomial regression models that incorporated weekly influenza surveillance data as covariates. Two categories of deaths and hospitalizations were used to make these estimates, pneumonia and influenza (P&I) and respiratory and circulatory (R&C). For P&I, the annual averages of the rate ratios of influenza-associated deaths to hospitalizations were in a range of 0.005 (95% CI, 0.003-0.011), 0.010 (0.007-0.012), 0.023 (0.005-0.040), 0.034 (0.009-0.070), and 0.117 (0.026-0.179) for < 5, 5-19, 20-64, 65-74, and 75+ years, respectively. For R&C, the rate ratios were in a range of 0.003 (0.003-0.008), 0.018 (0.013-0.022), 0.049 (0.014-0.144), 0.101 (0.034-0.282), and 0.253 (0.095-0.601), respectively. These estimates of the rate ratios are important for disease prevention efforts.