Applicability of influenza-like illness surveillance guidelines for H1N1 screening

Background: The Centers for Disease Control and Prevention (CDC) definition of influenza-like illness (ILI) as fever with cough and/or sore throat has major implications on public health surveillance and response. This study investigates the applicability of ILI surveillance guidelines as an efficient tool to screen for influenza A (H1N1) virus cases. Methods: This is a retrospective cross-sectional study design conducted September 5 to November 25, 2009 in the emergency department of an academic medical center. Six hundred thirty three patients received a nasopharyngeal swab followed by Polymerase Chain Reaction (PCR) testing for respiratory viruses. Geographic Utilization of Artificial Intelligence in Real-Time for Disease Identification and Notification (GUARDIAN), a syndrome surveillance program, was utilized to assign ILI status. Positive predictive value (PPV), negative predictive value (NPV), sensitivity and specificity as well as chi-square analysis were utilized to test the reliability of the ILI definition to detect H1N1 cases. Results: The PPV, NPV, sensitivity, and specificity of ILI compared to H1N1 PCR results were 34.17%, 86.01%, 59.85%, and 68.15%, respectively. The chi-square results were significant with 59.9% of H1N1 positive patients exhibiting ILI symptoms as compared to 32% of H1N1 negative patients. Conclusions: The ILI case definition casts a broad net and includes some of the positive H1N1 cases. The presence of a large number of cases with no ILI symptoms but positive H1N1 and cases with positive ILI symptoms but negative H1N1, needs further investigation for contributing factors such as the presence of other respiratory diseases, infants' inability to report symptoms, and physician practice.