Burden of Influenza Is Underestimated When Based on Pneumonia Cases
Influenza is a highly contagious viral disease that can be fatal in older adults. As the virus is constantly mutating, influenza outbreaks can reach epidemic and even pandemic proportions. Public health agencies are challenged each season with implementing the most effective prevention and control program well in advance of expected outbreaks. The use of disease surveillance data provides invaluable information that has proven helpful to this process. Influenza pneumonia surveillance is often performed on hospitalization data using statistical models to provide estimates of the total burden of the influenza disease. However, a recent study reported in PloS One found that influenza surveillance systems with case definitions requiring the presence of pneumonia might be underestimating the total influenza burden.1
Related Content:
Bacterial Pneumonia Acquired in Nursing Homes
What Geriatric Care Providers Should Know for the 2014-2015 Flu Season
A team of researchers from the University of Washington conducted a modeling study to examine if hospital influenza surveillance with a case definition restricted to pneumonia diagnoses can underestimate the total burden of influenza. The team used the State Inpatient Database from three states (Washington, California, and Arizona) and surveillance data from the same geographical region provided by the Centers for Disease Control and Prevention. Data covering a 6-year period (2003-2009) were collected. Two categories of influenza-associated cases were targeted in the study: (1) respiratory or circulatory influenza; and (2) critical illness influenza. Using negative binomial regression models, the total number of adult influenza-associated hospitalizations was estimated to be 80,834 (95% confidence interval [CI] 29,214-174,033) respiratory or circulatory cases, and 26,760 (95% CI 14,541-47,464) critical illness cases. The statistical analyses were repeated while excluding cases with a pneumonia diagnosis. There were 24,816 respiratory or circulatory cases (95% CI, 6,342 – 92,624), and 8,213 critical illness cases associated with influenza (95% CI, 3,764 – 20,799). These results suggest that approximately 30% of hospitalizations from each targeted category do not have a pneumonia diagnosis and would, therefore, likely be missed in a hospital-based pneumonia influenza surveillance.
According to the researchers, the findings of this study indicate “disease modeling approaches will estimate 50% fewer cases of influenza-associated hospitalizations if pneumonia ICD-9-CM codes are required as compared to more sensitive respiratory or circulatory ICD-9-CM codes.” They report that this has significant implications for influenza surveillance performed in hospitals and the estimates of total influenza burden. “A substantial proportion of influenza-associated hospitalizations may be missed,” and particularly for countries relying on local or regional burden of disease data, “such underestimates might affect the perceived value or cost-effectiveness of influenza vaccination.”
-Allison Musante
Reference
1. Ortiz JR, Neuzil KM, Cooke CR, et al. Influenza pneumonia surveillance among hospitalized adults may underestimate the burden of severe influenza disease. PLoS One. 2014;9(11):e113903.
