Both molecular and serological assays have been used previously to determine the etiology of community-acquired pneumonia (CAP). However, the extent to which these methods are correlated and the added diagnostic value of serology for respiratory viruses other than influenza virus have not been fully evaluated. Using data from patients enrolled in the Centers for Disease Control and Prevention (CDC) Etiology of Pneumonia in the Community (EPIC) study, we compared real-time reverse transcription-PCR (RT-PCR) and serology for the diagnosis of respiratory syncytial virus (RSV), human metapneumovirus (HMPV), parainfluenza virus 1 to 3 (PIV1, PIV2, and PIV3), and adenovirus (AdV) infections. Of 5,126 patients enrolled, RT-PCR and serology test results were available for 2,023, including 1,087 children below the age of 18 years and 936 adults. For RSV, 287 (14.2%) patients were positive by RT-PCR and 234 (11.6%) were positive by serology; for HMPV, 172 (8.5%) tested positive by RT-PCR and 147 (7.3%) by serology; for the PIVs, 94 (4.6%) tested positive by RT-PCR and 92 (4.6%) by serology; and for AdV, 111 (5.5%) tested positive by RT-PCR and 62 (3.1%) by serology. RT-PCR provided the highest number of positive detections overall, but serology increased diagnostic yield for RSV (by 11.8%), HMPV (by 25.0%), AdV (by 32.4%), and PIV (by 48.9%). The method concordance estimated by Cohen's kappa coefficient (κ) ranged from good (for RSV; κ = 0.73) to fair (for AdV; κ = 0.27). Heterotypic seroresponses observed between PIVs and persistent low-level AdV shedding may account for the higher method discordance observed with each of these viruses. Serology can be a helpful adjunct to RT-PCR for research-based assessment of the etiologic contribution of respiratory viruses other than influenza virus to CAP.
Copyright © 2016 American Society for Microbiology.