An epidemic of respiratory disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) began in China and has spread to other countries. Real-time reverse transcriptase-polymerase chain reaction (RRT-PCR ) from nasopharyngeal swabs has generally been used to confirm the clinical diagnosis. However, it is unknown whether the virus can be detected in samples from other sites and thus may be transmitted in ways other than respiratory droplets.


We investigated the biodistribution of SARS-CoV-2 among different tissues of hospitalized patients with coronavirus disease 2019 (COVID-19) diagnosed based on symptoms and radiology and confirmed by SARS-CoV-2 detection. This study was approved by the ethics committees of the participating hospitals, waiving informed consent.

Patients were included with samples collected according to clinical indications from 3 hospitals in Hubei and Shandong provinces and Beijing, China, from January 1 to February 17, 2020. Pharyngeal swabs were collected from most patients at 1 to 3 days after hospital admission.

Blood, sputum, stool, urine, and nasal samples were collected throughout the illness. Bronchoalveolar lavage fluid and fiberoptic brush biopsy samples were taken from patients with severe disease or undergoing mechanical ventilation. RNA was extracted from clinical samples and determined by RRT-PCR targeting the SARS-CoV-2 open reading frame gene 1ab as described previously.2 RRT-PCR cycle threshold values ​​were used as indicators of the SARS-CoV copy number.

RNA in samples with lower cycle threshold values ​​corresponding to higher viral copy numbers. A cycle threshold value less than 40 is interpreted as positive for SARS-CoV-2 RNA. Four high copy number SARS-CoV-2 positive faecal samples were cultured and then electron microscopy was performed to detect the live virus. Patterns were explored in a subgroup of patients with multiple samples collected during hospitalization.


A total of 1,070 samples were collected from 205 COVID-19 patients who had a mean age of 44 years (range, 5-67 years) and 68% were men. Most of the patients presented fever, dry cough and fatigue; 19% of patients had severe disease. Bronchoalveolar lavage fluid samples showed the highest positive rates (14 of 15; 93%), followed by sputum (72 of 104; 72%), nasal swabs (5 of 8; 63%), fiberoptic brush biopsy (6 of 13; 46%), throat swabs (126 of 398; 32%), stool (44 of 153; 29%), and blood (3 of 307; 1%). None of the 72 urine samples was positive.

Mean cycle threshold values ​​for all sample types were greater than 30 (<2.6 × 104 copies/mL), except for nasal swabs with a mean cycle threshold value of 24.3 (1.4 × 106 copies/mL), indicating higher viral loads.

Twenty patients had 2 to 6 specimens collected simultaneously. Viral RNA was detected in individual samples from 6 patients (respiratory samples, faeces or blood), while 7 patients excreted virus in respiratory tract samples and faeces (n = 5) or blood (n = 2). Live SARS-CoV-2 was observed in the stool sample of 2 patients who did not have diarrhoea.


In this study, SARS-CoV-2 was detected in multi-site samples from 205 COVID-19 patients, with lower respiratory tract samples testing positive for the virus more frequently. Importantly, the live virus was detected in faeces, implying that SARS-CoV-2 can be transmitted via the faeces. A small percentage of blood samples had positive results on the PCR test, suggesting that the infection can sometimes be systemic.

Respiratory and extra-respiratory transmission of the virus may help explain the rapid spread of the disease. In addition, multi-site sample testing can improve sensitivity and reduce false-negative results. Two smaller studies reported the presence of SARS-CoV-2 in anal or oral swabs and blood from 16 patients in Hubei province, and the viral load in throat swabs and sputum from 17 confirmed cases.

Limitations of this study include that some patients did not have detailed clinical information available, so the data could not be correlated with symptoms or disease course, and that the number of some sample types was small. Further investigation of patients with detailed temporal and symptom data and consecutively collected samples from different sites is warranted.