Recent studies showed that SARS-CoV-2 RNA may be found in fecal specimens of COVID-19 patients, but the sample size is limited. This systematic review and meta-analysis examined the detection rate of SARS-CoV-2 RNA in fecal specimens of these patients according to their clinical characteristics.

MEDLINE, Embase, Scopus, and three Chinese biomedical databases were searched up to 25 March 2020 with no language restriction. We included original observational studies that reported the detection rate of SARS-CoV-2 RNA in fecal specimens of COVID-19 patients. Two separate reviewers conducted the review. Metaprop was adopted to conduct a meta-analysis of prevalence with variances stabilized by Freeman-Tukey Double Arcsine Transformation. A random-effects model was used. Heterogeneity across different studies was computed using Cochran's Q test and chi square statistics.

From 17 studies, the pooled detection rate of fecal SARS-CoV-2 RNA was 43.7% (95% CI 32.6%-55.0%) and 33.7% (95% C.I. 33.7%, 95% C.I. 20.1%-48.8%) by patient and number of specimens as a unit count, respectively. Female individuals (59.6% vs. 53.5%), those who presented with gastrointestinal symptoms (77.1% vs. 57.7%), and patients with more severe disease (68.3% vs. 34.6%) tended to have a higher detection rate.

A significant proportion of COVID-19 patients carry SARS-CoV-2 in their intestinal tract. Feces being a self-collected specimen bears a potential to improve case identification in community, especially for young children where proper respiratory sampling at home is difficult. Specific infection control strategies focusing on spread via fecal contamination and faulty toilet drainage are urgently needed.