Clinical Features of Respiratory Viral Infections Among Inpatients at a Major US Tertiary Care Hospital.

Malhotra, Prashant; Luka, Arthur; McWilliams, Carla S; Poeth, Kaitlin G; Schwartz, Rebecca; Elfekey, Mohammed; Balwan, Sandy

OBJECTIVES Respiratory viral illnesses (RVI) are reliably diagnosed by respiratory viral panel using polymerase chain reaction (RVP-PCR); however, owing to the scant data, clinical presentation alone is unreliable in establishing viral etiology. The primary objective of this study was to characterize signs and symptoms of RVI among inpatients in a major tertiary care hospital. METHODS Between 2013 and 2015, adult inpatients with RVI undergoing RVP-PCR were prospectively enrolled in our study. Clinical data were collected by interviews and electronic medical record reviews. Data analysis was performed using χ(2) testing, analysis of variance for continuous variables, and logistic regression modeling. RESULTS Of 421 patients analyzed, 175 (41.7%) had a positive RVP-PCR. Patients were evenly matched at baseline except for renal disease. Multivariate logistic regression modeling demonstrated the following positive correlations: positive RVP-PCR with renal disease (odds ratio [OR] 2.08), cough (OR 2.28), and wheezing (OR 1.8); influenza with cough (OR 5.04), and renal disease (OR 2.17); metapneumovirus with age older than 65 (OR 3.24); respiratory syncytial viruses with wheezing (OR 3.42) and immunosuppression (OR 3.11); and parainfluenza with smoking (OR 3.16). Negative correlations included influenza with anosmia (OR 0.41); rhinovirus/enterovirus with feeling confined to bed (OR 0.3); metapneumovirus with smoking (OR 0.29); and parainfluenza with male sex (OR 0.22). CONCLUSIONS In this descriptive study, we noted specific viral associations with clinical signs and symptoms among 421 inpatients with RVIs. With increasing RVP-PCR use, studies similar to ours may be able to better define the clinical presentation of RVIs and lead to evidence-based, clinical presentation-guided diagnostic and management algorithms.