Influenza as a Proportion of Pneumonia Mortality: United States, 1959–2007

Andrew Noymer, University of California, Irvine
Ann Nguyen, Palomar Pomerado Health

As causes of death, influenza and pneumonia are typically analyzed together, but sometimes influenza-only analysis is conducted. We quantify influenza's contribution to the combined pneumonia and influenza time series of mortality for the United States, 1959-2007. The analysis includes time series plots and calculation of distributions. A key statistic is the proportion of all pneumonia and influenza mortality that is accounted for by influenza, I/(P+I). Year-to-year, I/(P+I) is highly variable and shows long-term decline. Extreme values of I/(P+I) are associated with extreme P+I death rates and vice-versa, but the ratio I/(P+I) is a weak predictor of P+I mortality overall. Prominence of influenza in the medical news is associated with high I/(P+I). In population-level studies, influenza and pneumonia should be analyzed as a combined cause of death unless there is a specific, and especial, reason to separate the two causes.