Diagnostic Accuracy of Novel and Traditional Rapid Tests for Influenza Infection Compared With Reverse Transcriptase Polymerase Chain Reaction: A Systematic Review and Meta-analysis

Abstract

BACKGROUND: Rapid and accurate influenza diagnostics can improve patient care. PURPOSE: To summarize and compare accuracy of traditional rapid influenza diagnostic tests (RIDTs), digital immunoassays (DIAs), and rapid nucleic acid amplification tests (NAATs) in children and adults with suspected influenza. DATA SOURCES: 6 databases from their inception through May 2017. STUDY SELECTION: Studies in English, French, or Spanish comparing commercialized rapid tests (that is, providing results in <30 minutes) with reverse transcriptase polymerase chain reaction reference standard for influenza diagnosis. DATA EXTRACTION: Data were extracted using a standardized form; quality was assessed using QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2) criteria. DATA SYNTHESIS: 162 studies were included (130 of RIDTs, 19 of DIAs, and 13 of NAATs). Pooled sensitivities for detecting influenza A from Bayesian bivariate random-effects models were 54.4% (95% credible interval [CrI], 48.9% to 59.8%) for RIDTs, 80.0% (CrI, 73.4% to 85.6%) for DIAs, and 91.6% (CrI, 84.9% to 95.9%) for NAATs. Those for detecting influenza B were 53.2% (CrI, 41.7% to 64.4%) for RIDTs, 76.8% (CrI, 65.4% to 85.4%) for DIAs, and 95.4% (CrI, 87.3% to 98.7%) for NAATs. Pooled specificities were uniformly high (>98%). Forty-six influenza A and 24 influenza B studies presented pediatric-specific data; 35 influenza A and 16 influenza B studies presented adult-specific data. Pooled sensitivities were higher in children by 12.1 to 31.8 percentage points, except for influenza A by rapid NAATs (2.7 percentage points). Pooled sensitivities favored industry-sponsored studies by 6.2 to 34.0 percentage points. Incomplete reporting frequently led to unclear risk of bias. LIMITATIONS: Underreporting of clinical variables limited exploration of heterogeneity. Few NAAT studies reported adult-specific data, and none evaluated point-of-care testing. Many studies had unclear risk of bias. CONCLUSION: Novel DIAs and rapid NAATs had markedly higher sensitivities for influenza A and B in both children and adults than did traditional RIDTs, with equally high specificities. PRIMARY FUNDING SOURCE: Québec Health Research Fund and BD Diagnostic Systems.