Background: Multiple case definitions are used to identify hospitalized patients with community-acquired acute respiratory infections (ARI). We evaluated several commonly used hospitalized ARI case definitions to identify influenza cases.
Methods: The study included all patients from a population-based surveillance site in Damanhour, Egypt hospitalized for a broad set of criteria consistent with community acquired ARIs. Naso- and oropharyngeal (NP/OP) swabs were tested for influenza using RT-PCR. Sensitivity, specificity and PPV for influenza identification was compared between the 2014 WHO Severe Acute Respiratory Infection (SARI) definition (fever ≥38°C and cough with onset within 10 days), the 2011 WHO SARI definition (fever ≥38°C and cough with onset within 7 days), the 2006 PAHO SARI definition, the International Emerging Infections Program (IEIP) pneumonia case definition, and the International Management of Childhood Illness (IMCI) case definitions for moderate and severe pneumonia.
Results: From June 2009-December 2012, 5768 NP/OP swabs were obtained from 6113 hospitalized ARI patients; 799 (13.9%) were influenza positive. The 2014 WHO SARI case definition captured the greatest number of ARI patients, influenza positive patients and ARI deaths compared to the other case definitions examined. Sensitivity for influenza detection was highest for the 2014 WHO SARI definition with 88.6%, compared to the 2011 WHO SARI (78.2%) the 2006 PAHO SARI (15.8%) the IEIP pneumonia (61.0%) and the IMCI moderate and severe pneumonia (33.8% and 38.9%) case definitions (IMCI applies to <5 only).
Conclusions: Our results support use of the 2014 WHO SARI definition for identifying influenza positive hospitalized SARI cases as it captures the highest proportion of ARI deaths and influenza positive cases. Routine use of this case definition for hospital-based surveillance will provide a solid, globally comparable foundation on which to build needed response efforts for novel pandemic viruses.