Abstract
From a laboratory perspective, conventional respiratory viral tests (antigen detection by immunofluorescence and viral culture) are assumed to be less sensitive than nucleic acid amplification tests (NATs). We systematically reviewed the literature to estimate the diagnostic accuracy of conventional tests compared to NATs as the reference standard, when studied in the pediatric inpatient setting. We searched Medline, Embase, and bibliographies of relevant articles for studies that compared conventional respiratory viral tests to NATs. Descriptions of methodology were appraised using the QUADAS criteria. Thirty-three publications (10297 patients) were included. Viral diagnostic approaches varied considerably. The majority of studies reported less than 70% of QUADAS items. Important flaws comprised a lack of clear criteria for patient selection and the absence of blinding. Study results could not be pooled due to heterogeneity. The sensitivity of conventional tests varied between 0.00 and 1.00, whereas the specificity was mostly higher than 0.90. In conclusion, the sensitivity of conventional tests varied considerably between studies, compared to NATs as the reference standard. Because clinical studies used a large variety of viral test approaches, the anticipated inferiority of conventional viral tests could not be confirmed. International standards are needed to assure efficient viral diagnosis and patient management.
Keywords: Polymerase chain reaction, direct immunofluorescence, viral culture, respiratory tract infection, children, systematic review, nucleic acid amplification tests, NATs, QUADAS, respiratory syncytial virus, influenza viruses, parainfluenza viruses, immunofluorescence, PCR, rhinovirus, pneumonia, bronchiolitis, Nucleic Acid Sequence-Based Amplification, NASBA
Related Journals
4