Viral respiratory tract infections in adult patients attending outpatient and emergency departments, Taiwan, 2012-2013: A PCR/electrospray ionization mass spectrometry study

Viral etiologies of respiratory tract infections (RTIs) have been less studied in adult than in pediatric populations. Furthermore, the ability of PCR/electrospray ionization mass spectrometry (PCR/ESIMS) to detect enteroviruses and rhinoviruses in respiratory samples has not been well evaluated. We sought to use PCR/ESI-MS to comprehensively investigate the viral epidemiology of adult RTIs, including testing for rhinoviruses and enteroviruses. Nasopharyngeal or throat swabs from 267 adults with acute RTIs (212 upper RTIs and 55 lower RTIs) who visited a local clinic or the outpatient or emergency departments of a medical center in Taiwan between October 2012 and June 2013 were tested for respiratory viruses by both virus isolation and PCR/ESI-MS. Throat swabs from 15 patients with bacterial infections and 27 individuals without active infections were included as control samples. Respiratory viruses were found in 23.6%, 47.2%, and 47.9% of the 267 cases by virus isolation, PCR/ESI-MS, and both methods, respectively. When both methods were used, the influenza A virus (24.3%) and rhinoviruses (9.4%) were the most frequently identified viruses, whereas human coronaviruses, human metapneumovirus (hMPV), enteroviruses, adenoviruses, respiratory syncytial virus, and parainfluenza viruses were identified in small proportions of cases (< 5% of cases for each type of virus). Coinfection was observed in 4.1% of cases. In the control group, only 1 (2.4%) sample tested positive for a respiratory virus by PCR/ESI-MS. Patients who were undergoing steroid treatment, had an active malignancy, or suffered from chronic obstructive pulmonary disease (COPD) were at risk for rhinovirus, hMPV, or parainfluenza infections, respectively. Overall, immunocompromised patients, patients with COPD, and patients receiving dialysis were at risk for noninfluenza respiratory virus infection. Rhinoviruses (12.7%), influenza A virus (10.9%), and parainfluenza viruses (7.3%) were the most common viruses involved in the 55 cases of lower RTIs. The factors of parainfluenza infection, old age, and immunosuppression were independently associated with lower RTIs. In conclusion, PCR/ESI-MS improved the diagnostic yield for viral RTIs. Non-influenza respiratory virus infections were associated with patients with comorbidities and with lower RTIs. Additional studies that delineate the clinical need for including non-influenza respiratory viruses in the diagnostic work-up in these populations are warranted.