The second limitation of the study is the relatively small sample size. Wuhan, China reported that 8/108 (7.4%) patients were RP (Cao et al., 2020). Moreover, 14.5% of convalescent patients (= 38) were RP for SARS-CoV-2 RNA in RT-PCR tests of both anal and nasopharyngeal swabs (An et al., 2020). RP is usually observed in young patients who had moderate or moderate COVID-19 symptoms around the first admission, and several significant characteristic features including early RNA-negative conversion, fewer comorbidities, and more frequent upper respiratory symptoms (An et al., 2020). No obvious clinical symptoms were reported on the second admission of these patients (An et al., 2020; Lan et al., 2020; Cao et al., 2020). In recent reports, close contacts of RP patients were tested unfavorable for SARS-CoV-2 RNA (An et al., 2020; Lan et al., 2020). However, as the patients were usually in quarantine after discharge, the infectivity of PTGS2 the patients might be underestimated. Two patients who were RP continued to be positive for SARS-COV-2 RNA for more than 90 days (Cao et al., 2020). Careful Amlexanox consideration should therefore be given to the potential for patients who are RP to become chronic virus-carriers. The cause of RP remains controversial, and consequently it is difficult to set standards for discharge and follow-up of patients. False negatives in Amlexanox qRT-PCR assessments may partially explain the RP in some cases, because the lower limit of detection (LOD) of commercial RT-PCR kits is usually relatively high (An et al., 2020). The residual viremia could be another factor that leads to RP. Pathological examination of a ready-for-discharge patient who had COVID-19 identified viral particles in the pneumocytes. Several studies have reported Amlexanox the presence of SARS-CoV-2 RNA and viral particles in the gastrointestinal tracts of patients who had COVID-19, and that fecal samples remained viral RNA-positive after patients were respiratory-negative for SARS-CoV-2 and were discharged from hospital (Xu et al., 2020), thus highlighting the prolonged presence of SARS-CoV-2 in the gastrointestinal tract. Here, we investigated whether the intestine might be a reservoir of SARS-CoV-2 and one of the potential causes of RP. Between January 21 and March 8, 2020, a total of 173 patients who had COVID-19 were discharged from hospitals in Hefei, China, including the First Affiliated Hospital of University of Science and Technology of China (USTC). The patients were enrolled in this study for analysis of clinical parameters, and were followed for at least one month. During the monitoring, 12 out of 173 patients were found to be RP (Table S1). All 12 of these Amlexanox patients had gastrointestinal symptoms, including nausea, diarrhea, anorexia, abdominal pain and belching Amlexanox (Table S2), and three out of those patients who were RP developed a gastrointestinal symptom as the onset symptom (Table S2). In contrast to the reported 2%C39% of patients with COVID-19 who develop gastrointestinal symptoms (Chen et al., 2020; Zhang et al., 2020; Zheng et al., 2020), patients who were RP in Hefei had a much higher incidence of gastrointestinal symptoms (= 0.026, chi-squared contingency table tests), which supported a potential causal relationship between intestinal contamination and RP. Applying both untargeted and targeted metagenomics approaches, three near full-length SARS-CoV-2 sequences and one partial genome sequence were recovered from four fecal samples (Fig. S1), indicating active viral contamination in the gastrointestinal tract of these patients. Patient F experienced a fever and whole-body aches from February.