Co-infection of SARS-CoV-2 with other respiratory viruses and performance of lower respiratory tract samples for the diagnosis of COVID-19

Abstract

Objectives: This study was performed during the early outbreak period of coronavirus disease 2019 (COVID-19) and the seasonal epidemics of other respiratory viral infections, in order to describe the extent of co-infections of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) with other respiratory viruses. It also compared the diagnostic performances of upper respiratory tract (URT) and lower respiratory tract (LRT) samples for SARS-CoV-2 infection.

Methods: From 25 January to 29 March 2020, all URT and LRT samples collected from patients with suspected COVID-19 received in the virology laboratory of Pitié-Salpêtrière University Hospital (Paris, France) were simultaneously tested for SARS-CoV-2 and other respiratory viruses.

Results: A total of 1423 consecutive patients were tested: 677 (47.6%) males, 746 (52.4%) females, median age 50 (range, 1-103) years. Twenty-one (1.5%) patients were positive for both SARS-CoV-2 and other respiratory viruses. The detection rate of SARS-CoV-2 was significantly higher in LRT than in URT (53.6% vs. 13.4%; p<0.0001). The analysis of paired samples from 117 (8.2%) patients showed that SARS-CoV-2 load was lower in URT than in LRT samples in 65% of cases.

Conclusion: The detection of other respiratory viruses in patients during this epidemic period could not rule out SARS-CoV-2 co-infection. Furthermore, LRT samples increased the accuracy of diagnosis of COVID-19.

Keywords: Co-infection; Lower respiratory tract; Other respiratory viruses; SARS-CoV-2.

Figure 1

Comparison of upper respiratory tract (URT) and lower respiratory tract (LRT) for diagnosing infections with SARS-CoV-2 and other respiratory viruses, using paired samples from COVID-19 suspected patients. (A) Distribution of paired respiratory samples according to the concordance of detection of SARS-CoV-2 and other respiratory viruses in URT and LRT. (B) Distribution of paired respiratory samples according to the difference in SARS-CoV-2 load between URT and LRT. SARS-CoV-2 loads were estimated with cycle threshold (Ct) values and the difference of Ct values obtained between URT and LRT (ΔCt URT – LRT) was calculated. Paired samples were then distributed according to the ΔCt value of 3.3, given that this difference approximately corresponds with a SARS-CoV-2 load of 1 log.