Table 1. The characteristics of all included studies.

Study	Population and characteristics	Control	Samples or Microbiota community	Timepoint at sample collection	Methods of microbiome characterization	Change in microbiota	Treatment/Outcome	Other key findings
Gu et al. (2020)	30 COVID-19 patients, China	(1) 24 H1N1 flu patients	Fecal samples	Samples were collected at admission.	V3-4 16S rRNA gene sequencing	Patients with COVID-19 or influenza had reduced bacterial diversity compared with controls. Stool samples of COVID-19 patients had an abundance of opportunistic pathogens, such as Streptococcus, Rothia, Veillonella, and Actinomyces. In contrast, Ruminococcaceae family and several genera from the family Lachnospiraceae were reduced in COVID-19 patients.	No outcome data	Richness, diversity, and structure of the gut microbiota were not significantly different between general COVID-19 and severe COVID-19.
	13 were women	(2) 30 healthy controls (matched for age, sex, body mass index)
	The median age was 55 (Range 29–70) years
	10 patients had comorbidities, including hypertension (30%)
Zuo et al. (2020a)	15 COVID-19 patients, China	15 healthy controls	Stool samples	2–3 times per week during hospitalization	Whole-genome sequencing	COVID-19 patients who were antibiotic naïve had increased opportunistic pathogens compared to controls.	Among 15 COVID-19 patients, 7 were antibiotic-naive, 8 received empiric antibiotics.	Clostridium ramosum and C. hathewayi were positively associated with severe COVID-19. In contrast, Alistipes onderdonkii and Faecalibacterium prausnitzii were negatively correlated with severe COVID-19.
	7 were males,	6 community-acquired pneumonia
	The median age was 55 (22–71) years
	11 patients had moderate-severe COVID-19
	Comorbidities: Hypertension (4), Hyperlipidemia (4), Diabetes (2), etc.
Zuo et al. (2020b)	30 COVID-19 patients, China	30 healthy controls	Stool samples (Fungal microbiome)	2–3 times per week during hospitalization	Whole-genome sequencing	The mycobiome in most patients (22 of 30 COVID-19 patients) was similar to that in healthy controls. In contrast, gut microbiome in COVID-19 patients (8 of 30) had alterations, with enrichment of Candida albicans and heterogeneous composition.	No treatment data	COVID-19 patients had increased proportions of opportunistic pathogens (Candida albicans, C. auris, Aspergillus flavus) during hospitalization.
	16 were males,	9 community-acquired pneumonia
	The median age was 46 (15–71) years
	11 patients had comorbidities
Zuo et al. (2020c)	Same population as Zuo et al. (2020a)	None	Stool samples	2–3 times per week during hospitalization	Whole-genome sequencing	Stool with high SARS-CoV2 infectivity had a higher abundance of bacterial species, including Collinsella aerofaciens, Collinsella tanakaei, Streptococcus infantis, and Morganella morganii.	Same treatment as Zuo et al. (2020a)	Stool with low to no SARS-CoV-2 infectivity had higher abundances of Parabacteroides merdae, Bacteroides stercoris, Alistipes onderdonkii, and Lachnospiraceae bacterium.
Zhang et al. (2020)	24 COVID-19 (nasopharyngeal swab) patients	Pneumonia cases	Nasopharyngeal swab, sputum samples	Samples were collected during inclusion in the cohort.	Whole-genome sequencing	COVID-19 patients had reduced alpha diversity in the airway microbiome. Opportunistic pathogens, including Candida albicans and human alphaherpesvirus 1, were frequently detected.	No treatment data	None
		36 non-COVID-19 (nasopharyngeal swab)
	14 COVID-19 (sputum) patients, China
		39 non-COVID-19 (sputum)
	37% were females
	The median age was 40.5 (25–82) years
De Maio et al. (2020)	18 COVID-19 patients, Italy	22 healthy controls	Nasopharyngeal swab samples	Samples were collected because of suspected COVID-19.	Sequencing of V5-V6 hypervariable region of bacterial 16S rRNA	The microbiota of the nasopharynx was not different in COVID-19 patients compared with that in controls.	No treatment data	None
	All patients had non-critical COVID-19
Shen et al. (2020)	8 COVID-19 pneumonia patients, China	25 community-acquired pneumonia	Bronchoalveolar lavage fluid samples	Samples were collected at bronchoscopic examination in clinical management.	Whole-genome sequencing	The BALF microbiota in COVID-19 patients was similar to that in patients with community-acquired pneumonia. The microbiome had an abundance of oral and upper respiratory commensal bacteria.	No treatment data	None
		20 healthy controls
Chen et al. (2020)	2 patients with COVID-19, China	None	Bronchoalveolar lavage fluid samples	Samples were collected at bronchoscopic examination for a diagnostic test.	Microbial next-generation sequencing analysis	Authors found Capnocytophaga species and Veillonella species in BALF samples.	Antiviral and anti-infectious treatment	None
	A 39-year-old male and a 21-year-old female
Ren et al. (2020)	5 patients with COVID-19, China	None	Bronchoalveolar lavage fluid samples	Samples were collected at bronchoscopic examination during hospitalization.	Microbial next-generation sequencing analysis	Most genome reads were viral (betacoronavirus), with bacterial pathogens such as Acinetobacter, Pseudomonas, Chryseobacterium, Escherichia, Streptococcus, Enterococcus, Rothia, and Lactobacillus.	5 patients received antibiotic therapy. Outcomes are recovered (1), hospitalized (3), and died (1).	None
	The median age was 52 (41–65) years
	2 patients were female
	All had moderate to severe COVID-19
	Comorbidities: hypertension, chronic liver disease
Fan et al. (2020)	20 deceased COVID-19 patients, China	None	Lung tissue samples	Samples were collected from deceased patients.	Sequencing of V3-V4 regions of bacterial 16S rRNA, ITS gene	Acinetobacter was the most common bacterial genus, followed by Chryseobacterium, Burkholderia, Brevundimonas, Sphingobium, and Enterobacterales. Cryptococcus was the most prevalent fungus, along with Issatchenkia, Wallemia, Cladosporium, and Alternaria.	All patients received antibacterial and antiviral therapy.	None
	14 were males
	The median age was 66 years
	Comorbidities: cardiovascular disease (10), hypertension (9), malignancy (7), diabetes (2), chronic kidney disease (2), chronic lung diseases (1)
Tao et al. (2020)	62 patients with COVID-19, China	33 seasonal influenza patients	Fecal samples	Samples were collected at first time of visit to the hospital.	Next-generation sequencing of V4 region of the 16S rRNA	COVID-19 patients had an abundance of Streptococcus, Clostridium, Lactobacillus, and Bifidobacterium in gut microbiota. In contrast, lower levels of Bacteroides, Roseburia, Faecalibacterium, Coprococcus, Parabacteroides were found.	No data	Alpha-diversity of gut microbiome decreased in COVID-19 compared with that in healthy control, flu patients.
		40 healthy controls
Yu et al. (unpublished)	2 male patients with COVID-19, China	22 healthy cohort (Data imported from Arumugam M, Raes J, Pelletier E, et al. Enterotypes of the human gut microbiome. Nature. 2011;473(7346):174–80.)	Anal swab samples	No data	No data	The proportion of gut microbiota, including Corynebacterium and Ruthenibacterium, was increased.	Both patients received antiviral and antibacterial agents. and both patients died.	None
	2 patients with COVID-19, China
	65 years old male and 78 years old male
	Comorbidities: prostatic hyperplasia, chronic bronchitis
Ai et al. (unpublished)	20 patients with COVID-19, China	33 pneumonia without COVID-19	Nasopharyngeal swab samples	Admission day	Multiplex RT-PCR assays	More than half of patients had co-infection with COVID-19 and another virus, such as influenza A/B, rhino- or enteroviruses, or respiratory syncytial virus.	No data	None
	10 were female				next-generation sequencing
	The median age was 37 years
Budding et al. (unpublished)	46 patients with COVID-19, Netherlands	89 SARS-CoV2 (-) patients	Throat swab samples	Samples were collected for routine diagnostic tests.	16S rDNA sequencing	Haemophilus parainfluenzae, Neisseria cinerea,	No data	None
						S. mitis, S. bovis, Leptotrichia buccalis, and Rothia mucilaginosa were the main composition of throat microbiome.
Xu et al. (a) (abstract only)	No data, China	No data	Intestinal microbiome	No data	No data	COVID-19 patients had intestinal dysbiosis with decreased Lactobacillus and Bifidobacterium.	No data	None
Xu et al. (b) (abstract only)	No data, China	No data	Gut microbiome	No data	No data	Decreased proportions of Lactobacillus and Bifidobacterium were observed in COVID-19 patients.	No data	None