A rapid review on the sensitivity of SARS-CoV-2 RT-PCR done on different clinical specimens

ABSTRACT

Background@#RT-PCR using respiratory tract specimens, most commonly nasopharyngeal swab (NPS), has been used to confirm the diagnosis of COVID-19. NPS is a relatively invasive procedure that causes patient discomfort and risks viral transmission. Other specimens are therefore being investigated for the detection of SARS-CoV-2 RNA. @*Objective@#To determine the sensitivity of non-respiratory tract specimens in detecting SARS-CoV-2 RNA in patients with COVID-19. @*Methodology@#This review summarized the results of eight studies obtained from a literature search done in May 2020 in PubMed MEDLINE, Cochrane Library and MedRxiv. Two independent investigators reviewed and appraised the studies that were included, and pooled estimates of sensitivity for each specimen were determined using Stata’s Metaprop function@*Results@#The sensitivity in detecting SARS-CoV-2 RNA in non-respiratory tract specimens of diagnosed COVID-19 patients are as follows: Saliva 77% (95% CI 71-83%), stool/rectal swab/anal swab 22% (95% CI 22-37%), blood/serum/plasma 2% (95% CI 1-3%), and urine 22% (95% CI 18-25%). @*Conclusion@#SARS-CoV-2 RNA is detected in saliva, stool/rectal swab/anal swab, blood/serum/plasma and urine. Among these, saliva has the highest estimated sensitivity. However, more studies are needed to correct the heterogeneity brought about by factors such as timing of specimen collection, disease severity and treatment.