Performance of case definitions used for influenza surveillance among hospitalized patients in a rural area of India

OBJECTIVE: To assess case definitions for influenza in a rural community in India. METHODS: Residents of the study area who were hospitalized for any acute medical condition for at least one night between May 2009 and April 2011 were enrolled. Respiratory specimens were collected and tested for influenza viruses in a reverse-transcription polymerase chain reaction (PCR). The PCR results were taken as the "gold standard" in evaluating the performance of several case definitions. FINDINGS: Of the 3179 patients included in the final analysis, 21% (665) were PCR-positive for influenza virus, 96% reported fever and 4% reported shortness of breath. The World Health Organization (WHO) case definition for severe acute respiratory illness had a sensitivity of 11% among patients aged < 5 years and of 3% among older patients. When shortness of breath was excluded from the definition, sensitivities increased (to 69% and 70%, respectively) and corresponding specificities of 43% and 53% were recorded. Among patients aged ≥ 5 years, WHO's definition of a case of influenza-like illness had a sensitivity of 70% and a specificity of 53%. The addition of "cough and reported or measured fever" increased sensitivity to 80% but decreased specificity to 42%. CONCLUSION: The inclusion of shortness of breath in WHO's case definition for severe acute respiratory illness may grossly underestimate the burden posed by influenza in hospitals. The exclusion of shortness of breath from this definition or, alternatively, the inclusion of "cough and measured or reported fever" may improve estimates of the burden.