Nosocomial infections among patients with COVID-19, SARS and MERS: a rapid review and meta-analysis.

BACKGROUND: COVID-19, a disease caused by SARS-CoV-2 coronavirus, has now spread to most countries and regions of the world. As patients potentially infected by SARS-CoV-2 need to visit hospitals, the incidence of nosocomial infection can be expected to be high. Therefore, a comprehensive and objective understanding of nosocomial infection is needed to guide the prevention and control of the epidemic.
METHODS: We searched major international and Chinese databases: Medicine, Web of Science, Embase, Cochrane, CBM (China Biology Medicine disc), CNKI (China National Knowledge Infrastructure) and Wanfang database for case series or case reports on nosocomial infections of COVID-19, SARS (severe acute respiratory syndromes) and MERS (Middle East respiratory syndrome) from their inception to March 31st, 2020. We conducted a meta-analysis of the proportion of nosocomial infection patients in the diagnosed patients, occupational distribution of nosocomial infection medical staff.
RESULTS: We included 40 studies. Among the confirmed patients, the proportions of nosocomial infections with early outbreaks of COVID-19, SARS, and MERS were 44.0%, 36.0%, and 56.0%, respectively. Of the confirmed patients, the medical staff and other hospital-acquired infections accounted for 33.0% and 2.0% of COVID-19 cases, 37.0% and 24.0% of SARS cases, and 19.0% and 36.0% of MERS cases, respectively. Nurses and doctors were the most affected among the infected medical staff. The mean numbers of secondary cases caused by one index patient were 29.3 and 6.3 for SARS and MERS, respectively.
CONCLUSIONS: The proportion of nosocomial infection in patients with COVID-19 was 44% in the early outbreak. Patients attending hospitals should take personal protection. Medical staff should be awareness of the disease to protect themselves and the patients. 2020 Annals of Translational Medicine. All rights reserved.

Introduction

COVID-19 is a respiratory infectious disease caused by a novel coronavirus, SARS-CoV-2. The first batch of COVID-19 patients were found in December 2019 (1). The disease is mainly transmitted through respiratory droplets and close contact, and all people are susceptible to it (2). SARS-CoV-2 is highly contagious (3), and has quickly spread to most countries and regions of the world. COVID-19 has become a global pandemic and has received great attention from all over the world (4,5). As of April 7, 2020, 1,214,466 confirmed cases of COVID-19 have been found in 211 countries and regions, causing 67,767 deaths (6).

The main clinical manifestations of COVID-19 are cough, fever and complications such as acute respiratory distress syndrome (1). Disease clusters and nosocomial infections have been reported (7,8). The proportion of nosocomial infections is high among diagnosed infections, and medical staff are at high risk of infection (8). One study on 44,672 patients showed that health workers accounted for 3.8% of the COVID-19 cases and five health workers died as a result of the infection (9). There is still no specific medicine for COVID-19, so preventing nosocomial infections is crucial.

This study compares the incidence of nosocomial infections during the COVID-19, SARS and MERS epidemics and analyzes the characteristics of the nosocomial infection, to enhance the understanding of nosocomial infection among medical and non-medical staff. We present the following article in accordance with the PRISMA reporting checklist (available at http://dx.doi.org/10.21037/atm-20-3324).

Methods

Search strategy

An experienced librarian searched the following databases from their inception to March 31, 2020 in the following electronic databases (10): the Cochrane Library, MEDLINE (via PubMed), EMBASE, Web of Science, CBM (China Biology Medicine disc), CNKI (China National Knowledge Infrastructure), and Wanfang database. We made no restrictions on language or publication status. We used the following search formula is as follow: (“Novel coronavirus” OR “2019-novel coronavirus” OR “Novel CoV” OR “2019-nCoV” OR “COVID-19” OR “SARS-CoV-2” OR “Middle East Respiratory Syndrome” OR “MERS” OR “MERS-CoV” OR “Severe Acute Respiratory Syndrome” OR “SARS” OR “SARS-CoV” OR “SARS-Related” OR “SARS-Associated”) AND (“Cross Infection” OR “Cross Infections” OR “Healthcare Associated Infections” OR “Healthcare Associated Infection” OR “Health Care Associated Infection “ OR “Health Care Associated Infections” OR “Hospital Infection” OR “Nosocomial Infection” OR “Nosocomial Infections” OR “Hospital Infections” OR “hospital-related infection” OR “hospital-acquired infection”). We also searched clinical trial registry platforms [the World Health Organization Clinical Trials Registry Platform (http://www.who.int/ictrp/en/), US National Institutes of Health Trials Register (https://clinicaltrials.gov/)], Google Scholar (https://scholar.google.nl/), preprint platform [medRxiv (https://www.medrxiv.org/), bioRxiv (https://www.biorxiv.org/) and SSRN (https://www.ssrn.com/index.cfm/en/)] and reference lists of the included reviews to find unpublished or further potential studies. Finally, we contacted experts in the field to identify relevant trials. The search strategy was also reviewed by another information specialist. The details of the search strategy can be found in the Supplement I.

Inclusion and exclusion criteria

We included case series studies and case reports about the proportion of cases of COVID-19, SARS and MERS who were infected in health facilities, about infections among medical staff and outbreaks in hospitals. Abstract, letter, new, guideline, articles for which we could not access all relevant data or full text were excluded.

Study selection

After eliminating duplicates, two reviewers (Y Gao and X Wang) independently selected the relevant studies in two steps with the help of the EndNote software. Discrepancies were settled by discussion or consulting a third reviewer (Q Zhou). In the first step, all titles and abstracts were screened using pre-defined criteria. In the second step, full-texts of the potentially eligible and unclear studies were reviewed to decide about final inclusion. All reasons for exclusion of ineligible studies were recorded. The process of study selection was documented using a PRISMA flow diagram (11).

Data extraction

Two reviewers (R Liu and X Wang) extracted the data independently using a standardized data collection table. Any differences were resolved by consensus, and a third auditor checked the consistency and accuracy of the data. The following data were extracted: (I) basic information: title, first author, country, year of publication, and type of study; (II) population baseline characteristics: age and sex distribution, and sample size; and (III) the proportion of nosocomial infections, the proportion of patients with occupation of medical staff, and for studies on hospital outbreaks, the number of index cases and total infections.

Risk of bias assessment

Two researchers (Z Wang and Q Shi) independently assessed the potential bias in each included study. The included studies were evaluated using appropriate assessment scales depending on the study type: for case control studies, the Newcastle-Ottawa Scale (NOS) (12), for cross-sectional studies and epidemiological surveys, the methodology evaluation tool recommended by the Agency for Healthcare Research and Quality (AHRQ) (13), and for case reports and case series, we used a methodology evaluation tool recommended by National Institute for Health and Care Excellence (NICE) (14).

Data synthesis

We performed a meta-analysis of proportions for dichotomous outcomes (nosocomial infection among the confirmed cases, and infections among the health care workers), reporting the effect size (ES) with 95% confidence intervals (CI) by using random-effects models. Two-sided P values <0.05 were considered statistically significant. Heterogeneity was defined as P<0.10 and I2>50%. All analyses were performed in STATA version 14. All results are limited to 0–100%.

Quality of the evidence assessment

Two reviewers (Z Wang and Q Shi) assessed the quality of evidence independently using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) (15,16). We produced a “Summary of Findings” table using the GRADEpro software. This table includes overall grading of evidence body for each prespecified outcome that is accounted in a meta-analysis. The overall quality can be downgraded for five considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) and upgraded for three considerations (large magnitude of effect, dose-response relation and plausible confounders or biases). The overall quality of evidence will be classified as high, moderate, low or very low, which reflecting to what extent that we can be confident the effect estimates are correct.

As COVID-19 is a public health emergency of international concern and the situation is evolving rapidly, our study was not registered in order to speed up the process (17).

Results

Characteristics and quality of included studies

Our initial search revealed 2,626 articles, of which 2,598 were left after deleting the duplicates (). After review the titles and abstracts, we screened the full texts of 66 articles, of which 40 were finally included (). Four studies were about COVID-19 (8,18-20), 25 studies about SARS (21-45), and 11 studies about MERS (46-56) (). Sixteen studies described the number of nosocomial infections in a selected patient population, 16 studies described the situation of nosocomial infections among the staff of medical institutions, and 13 studies reported the number of nosocomial infections caused by one or more than one patient. The quality of included studies was very poor: all cross-sectional studies scored less than 8 out of 11 in the evaluation by the AHRQ tool, half case series studies scored less than 5 out of 8 in the evaluation by the NICE tool, and only one case-control study scored 6 by the NOS tool. The details of the risk of bias of included studies can be found in the Supplement II ().

Figure 1

Flow diagram of the literature search.

Table 1

Characteristics of included studies

Study ID	Disease	Study type	Time	Location of the study	Sample size
Wang 2020 (8)	COVID-19	Case series	2020.01.01–2020.01.28	Wuhan	138
Wang 2020 (18)	COVID-19	Case series	2020.01.01–2020.01.28	Hubei	451
Jiang 2020 (19)	COVID-19	Case series	2019.12.15–2020.02.15	Wuhan	41
Shen 2020 (20)	COVID-19	Case control study	2020.01.15–2020.02.08	Wuhan	158
Bi 2003 (21)	SARS	Case series	2003.01.31–2003.02.17	Guangdong	25
Dai 2004 (22)	SARS	Cross-sectional study	203.01.18–2003.03.08	Guangdong	230
Zou 2004 (23)	SARS	Cross-sectional study	To 2003.05	Guangdong	2,635
Wang 2003 (24)	SARS	Cross-sectional study	2003.01.02–2003.04.17	Guangdong	966
Gao 2003 (25)	SARS	Cross-sectional study	2003.05.14–2003.05.17	Guangdong	86
Lin 2003 (26)	SARS	Cross-sectional study	To 2003.05	Guangdong	395
Xu 2003 (27)	SARS	Cross-sectional study	2003.01.13–2003.05.05	Guangdong	1,074
Gao 2003 (28)	SARS	Cross-sectional study	To 2003.07.07	–	669
Yuan 2003 (29)	SARS	Cross-sectional study	2003.01–2003.06.20	Shenzhen	53
Wang 2003 (30)	SARS	Cross-sectional study	2003.04.13–2003.05.08	Tianjin	175
Wang 2003 (31)	SARS	Cross-sectional study	2003.04.20–2003.05.18	Tianjin	2,300
Wu 2004 (32)	SARS	Cross-sectional study	2003.03.27–2003.06.24	Beijing	1,861
Huang 2003 (33)	SARS	Cross-sectional study	2003.02.02–2002.05	Guangdong	454
Li 2003 (34)	SARS	Cross-sectional study	2002.12.26–2003.01.19	Zhongshan	29
Fei 2003 (35)	SARS	Cross-sectional study	2003.03–2003.04	Beijing	33
Lu 2003 (36)	SARS	Case series	From 2003.04.05	Beijing	80
He 2003 (37)	SARS	Cross-sectional study	To 2003.05.20	Beijing	2,444
Ho 2003 (38)	SARS	Cross-sectional study	2003.03.25–2003.05.05	Hong Kong	1,312
Li 2003 (39)	SARS	Cross-sectional study	2003.03.15–2003.05.18	Beijing	740
Fowler 2003 (40)	SARS	Case series	To 2003.04.15	Toronto	38
					164
Varia 2003 (41)	SARS	Cross-sectional study	–	Toronto	128
Lau 2004 (42)	SARS	Cross-sectional study	–	Hong Kong	339
Zhou 2004 (43)	SARS	Cross-sectional study	2003.01.05–2003.05.09	Guangdong	1,645
Chen 2006 (44)	SARS	Cross-sectional study	To 2003.07	Singapore	105
Cooper 2009 (45)	SARS	Cross-sectional study	2003.02.21–2003.03.28	Beijng	41
		Cross-sectional study	2003.03.25–2003.04.12	Beijng	99
		Cross-sectional study	2003.04.16–2003.05.12	Tianjin	91
Oboho 2015 (46)	MERS	Cross-sectional study	2014.01.01–2014.05.01	Saudi Arabia	255
Xiang 2015 (47)	MERS	Cross-sectional study	2015.5.20–2015.7.13	South Korea	186
Assiri 2013 (48)	MERS	Case series	2013.04.01–2013.07.12	Saudi Arabia	447
Alenazi 2017 (49)	MERS	Cross-sectional study	2015.07.15–2015.09.15	Saudi Arabia	130
Memish 2015 (50)	MERS	Cross-sectional study	2013.08.24–2013.09.03	Saudi Arabia	306
Park 2016 (51)	MERS	Cross-sectional study	2015.05.20–2015.07.19	South Korea	76
					70
Al-Dorzi 2016 (52)	MERS	Case series	2015.08.25–2015.09.23	Saudi Arabia	276
Hunter 2016 (53)	MERS	Cross-sectional study	2013.01.01–2014.05.09	Saudi Arabia	65
Amer 2018 (54)	MERS	Cross-sectional study	2017.03.31–2017.07.15	Saudi Arabia	120
Cho 2016 (55)	MERS	Case series	2015.05.27–2015.05.29	South Korea	1,576
Hijawi 2013 (56)	MERS	Cross-sectional study	2012.04.01–2012.09.30	Jordan	13

Table S1

Cross-sectional studies

Study ID	Disease	Item 1	Item 2	Item 3	Item 4	Item 5	Item 6	Item 7	Item 8	Item 9	Item 10	Item 11	Scores†
Dai 2004 (22)	SARS	Yes	Yes	Yes	No	No	No	No	No	No	No	No	3
Zou 2004 (23)	SARS	Yes	Yes	Yes	Yes	No	No	No	No	No	Yes	No	5
Wang 2003 (24)	SARS	Yes	Yes	Yes	Yes	Yes	Yes	No	No	No	No	No	6
Gao 2003 (25)	SARS	Yes	No	Yes	No	No	No	No	No	No	No	No	2
Lin 2003 (26)	SARS	Yes	No	Yes	No	No	Yes	No	No	No	No	No	3
Xu 2003 (27)	SARS	Yes	Yes	Yes	Yes	No	No	No	No	No	Yes	Yes	6
Gao 2003 (28)	SARS	Yes	Yes	Yes	No	No	No	No	No	No	No	No	3
Yuan 2003 (29)	SARS	Yes	Yes	Yes	No	No	No	No	No	No	No	No	3
Wang 2003 (30)	SARS	Yes	Yes	Yes	Yes	No	Yes	No	No	No	No	No	5
Wang 2003 (31)	SARS	Yes	No	No	No	No	No	No	No	No	No	No	1
Wu 2004 (32)	SARS	Yes	No	Yes	No	No	No	No	No	No	No	No	2
Huang 2003 (33)	SARS	Yes	Yes	Yes	Yes	No	No	No	No	No	No	No	4
Li 2003 (34)	SARS	Yes	Yes	Yes	No	No	No	No	No	No	No	No	3
Fei 2003 (35)	SARS	Yes	No	Yes	Yes	No	No	No	No	No	No	No	3
He 2003 (37)	SARS	Yes	Yes	No	Yes	Yes	Yes	Yes	No	No	Yes	No	7
Ho 2003 (38)	SARS	Yes	Yes	Yes	Yes	Yes	Yes	No	No	No	No	No	6
Li 2003 (39)	SARS	Yes	Yes	Yes	No	No	No	No	No	No	No	No	3
Varia 2003 (41)	SARS	Yes	Yes	Yes	Yes	No	No	No	No	No	No	No	4
Lau 2004 (42)	SARS	Yes	Yes	Yes	Yes	No	No	No	No	No	No	No	4
Zhou 2004 (43)	SARS	Yes	Yes	Yes	Yes	Yes	No	No	Yes	No	No	No	6
Chen 2006 (44)	SARS	Yes	Yes	Yes	Yes	No	No	No	No	No	No	No	4
Cooper 2009 (45)	SARS	Yes	Yes	Yes	Yes	No	No	No	No	No	No	No	4
Oboho 2015 (46)	MERS	Yes	Yes	No	Yes	Yes	Yes	No	Yes	No	Yes	No	7
Xiang 2015 (47)	MERS	Yes	Yes	Yes	Yes	Yes	No	No	Yes	No	No	No	6
Alenazi 2017 (49)	MERS	Yes	Yes	Yes	Yes	No	No	No	No	No	No	No	4
Memish 2015 (50)	MERS	Yes	Yes	Yes	Yes	Yes	No	No	No	No	No	No	5
Park 2016 (51)	MERS	Yes	Yes	Yes	Yes	Yes	Yes	No	No	No	No	No	6
Hunter 2016 (53)	MERS	Yes	Yes	Yes	Yes	Yes	No	No	No	No	No	No	5
Amer 2018 (54)	MERS	Yes	Yes	Yes	Yes	Yes	Yes	No	No	No	No	No	6
Hijawi 2013 (56)	MERS	Yes	Yes	Yes	No	No	No	No	No	No	No	No	3

†, according to the methodology evaluation tool recommended by the Agency for Healthcare Research and Quality. This tool assesses the quality of bias according to 11 criteria. And each criterion is answered by “Yes”, “No” or “unsure”. The results were summarized by scoring method, for the “Yes” items, the score was 1, and for the “no” items, the score was 0. The maximum score is 11; the higher the score, the lower the risk of bias. The numbers 1 to 11 refer to the items of the tool: 1. defining the source of information (survey, record review); 2. listing the inclusion and exclusion criteria for exposed and unexposed subjects or referring to previous publications; 3. indicate time period used for identifying patients; 4. indicating whether the subjects were recruited consecutively (if not population-based); 5. indicating if evaluators of subjective components of the study were masked from the participants; 6. description of any assessments undertaken for quality assurance purposes (e.g., test/retest of primary outcome measurements); 7. explaining any exclusions of patients from the analysis; 8. description how confounding was assessed and/or controlled; 9. if applicable, explaining how missing data were handled in the analysis; 10. summarizing patient response rates and completeness of data collection; 11. clarification of the expected follow-up (if any), and the percentage of patients with incomplete data or follow-up.

Table S2

Case series

Study ID	Disease	Item 1	Item 2	Item 3	Item 4	Item 5	Item 6	Item 7	Item 8	Scores††
Wang 2020 (8)	COVID-19	Yes	Yes	Yes	No	No	Yes	Yes	Yes	6
Wang 2020 (18)	COVID-19	No	Yes	No	No	No	Yes	Yes	Yes	4
Jiang 2020 (19)	COVID-19	Yes	Yes	Yes	No	No	Yes	Yes	Yes	6
Bi 2003 (21)	SARS	No	Yes	No	No	No	No	Yes	Yes	3
Lu 2003 (36)	SARS	No	Yes	No	No	No	No	Yes	Yes	3
Fowler 2003 (40)	SARS	Yes	Yes	Yes	Yes	No	No	Yes	Yes	6
Assiri 2013 (48)	MERS	Yes	Yes	Yes	No	No	No	Yes	Yes	5
Al-Dorzi 2016 (52)	MERS	No	Yes	Yes	No	No	No	Yes	Yes	4
Cho 2016 (55)	MERS	Yes	Yes	Yes	Yes	No	No	Yes	Yes	6

††, according to the methodology evaluation tool recommended by National Institute for Health and Care Excellence. The risk of bias is evaluated according to eight criteria. The results were summarized by scoring method, for the “Yes” items, the score was 1, and for the “no” items, the score was 0. The maximum score is 8; the higher the score, the lower the risk of bias. The numbers 1 to 8 refer to the items of the tool: 1. case series collected in more than one centre, i.e., multi-centre study; 2. is the hypothesis/aim/objective of the study clearly described? 3. are the inclusion and exclusion criteria (case definition) clearly reported? 4. is there a clear definition of the outcomes reported? 5. were data collected prospectively? 6. is there an explicit statement that patients were recruited consecutively? 7. are the main findings of the study clearly described? 8. are outcomes stratified? (e.g., by disease stage, abnormal test results, patient characteristics).

Table S3

Case control study

Study ID	Disease	Selection					Comparability		Exposure			Scores†††
		Item 1	Item 2	Item 3	Item 4		Item 5		Item 6	Item 7	Item 8
Shen 2020 (20)	COVID-19	*	*	*	–		**		*			6

†††, according to the methodology evaluation tool of Newcastle-Ottawa Scale. It consists of eight domains, for each, we will grade with stars. The more stars, the lower the risk of bias. The maximum score is 9. A study can be awarded a maximum of one star for each numbered item within the Selection and Exposure categories. A maximum of two stars can be given for Comparability. The numbers 1 to 8 refer to the items of the tool: 1. representativeness of the exposed cohort; 2. selection of the non-exposed cohort; 3. ascertainment of exposure; 4. demonstration that outcome of interest was not present at start of study; 5. comparability of cohorts on the basis of the design or analysis; 6. assessment of outcome; 7. was follow-up long enough for outcomes to occur; 8. adequacy of follow up of cohorts.

Nosocomial infections among confirm cases

The proportion of nosocomial infections was 44.0% (95% CI: 0.36 to 0.51; I2=0.00%) among COVID-19 patients in the early outbreak, 36.0% (95% CI: 0.23 to 0.49; I2=97.8%) among SARS patients, and 56.0% (95% CI: 0.08 to 1.00; I2=99.9%) among MERS patients (). Thirty-three percent (95% CI: 0.27 to 0.40; I2=0.00%) of patients with COVID-19 were medical staff, and 2.0% (95% CI: 0.01 to 0.03; I2=0.00%), were nosocomial infections among people other than medical staff (such as inpatients or visitors). The corresponding proportions among SARS patients were 37.0% (95% CI: 0.25 to 0.49; I2=97.3%) and 24.0% (95% CI: 0.10 to 0.38; I2=86.6%), and 19.0% (95% CI: 0.04 to 0.35; I2=97.8%) and 36.0% (95% CI: 0.06 to 0.67; I2=99.3%) among MERS patients ().

Figure 2

The proportion of nosocomial infections among confirm cases of COVID-19, SARS and MERS.

Figure 3

Proportions of health care workers among confirmed cases of COVID-19, SARS and MERS.

Figure 4

Proportions of nosocomial infections excluding health care workers among confirm cases of COVID-19, SARS and MERS.

Infection among the health care workers

Twenty studies mentioned infection among the health workers, of which sixteen studies described the occupational composition of infected health care workers. Doctors accounted for 33.0% (95% CI: 0.24 to 0.44), nurses 56.0% (95% CI: 0.45 to 0.66), and other staff (such as carers, cleaners, hospital support staff) 11.0% (95% CI: 0.06 to 0.20) of COVID-19 cases among hospital staff in the early outbreak in Wuhan. For SARS, 30.0% (95% CI: 0.19 to 0.40; I2=91.1%) of the cases among hospital workers were doctors, 50.0% (95% CI: 0.45 to 0.55; I2=38.8%) nurses, and 21.0% (95% CI: 0.12 to 0.29; I2=85.2%) others. For MERS, for the corresponding proportions were 35.0% (95% CI: 0.14 to 0.56; I2=0.00%), 50.0% (95% CI: 0.29 to 0.71; I2=0.00%) and 16.0% (95% CI: 0.00 to 0.32; I2=0.00%). For all three conditions combined, the proportion of doctors among infected hospital staff was 30.0%, 51.0% for the proportion of nurses, and 19.0% for the proportion of others ().

Figure 5

Proportion of doctors among hospital staff with COVID-19, SARS and MERS.

Figure 6

Proportion of nurses among hospital staff with COVID-19, SARS and MERS.

Figure 7

Proportion of staff other than doctors or nurses among hospital staff with COVID-19, SARS and MERS.

Five studies described the protective measures of medical staff infected with SARS in hospital. Sixty-three percent (95% CI: 0.35 to 0.92; I2=96.1%) of the infected staff did not wear protective clothing), 58.0% (95% CI: 0.39 to 0.76; I2=0.00%) did not use gloves, 91.0% (95% CI: 0.80 to 1.00; I2=0.00%) did not wear goggles; 57.0% (95% CI: 0.00 to 1.00; I2=0.00%) did not take any hand disinfection measures), and 7.0% (95% CI: 0.00 to 0.16; I2=0.00%) did not wear masks (). One study described that among the 22 infected medical workers, 21 had no shoe cover. One study described that of 53 infected health workers, 47 wore cloth masks.

Figure 8

Proportion of health care staff with SARS who did not take protective measures.

Outbreaks in the hospitals

Six studies described SARS outbreaks, and five studies MERS outbreaks that happened in hospitals. The SARS studies reported on 23 patients, causing a total of 674 infections in hospitals, with an average of 29.3 infections per index patient. The MERS studies reported 24 patients causing 152 infections in hospitals, with an average of 6.3 infections per index patient ().

Table 2

Secondary infected by index patient in outbreaks in the hospitals

Disease	Study ID	Index patients	Number of secondary cases
SARS	Bi 2003 (21)	3	22
	Wang 2003 (30)	1	164
	Fei 2003 (35)	2	30
	Varia 2003 (41)	6	126
	Chen 2006 (44)	7	105
	Cooper 2009 (45)	4	227
	Total	23	674
MERS	Memish 2015 (50)	18	4
	Park 2016 (51)	1	23
	Hunter 2016 (53)	3	27
	Amer 2018 (54)	1	16
	Cho 2016 (55)	1	82
	Total	24	152

Quality of evidence

The results of GRADE on nosocomial infections showed that the quality of evidence were low or very low. The details can be found in the Supplement III ().

Table S4

Summary of findings

Outcomes	No. of studies	Sample size	Certainty assessment					Effect value (95% CI)	Certainty
			Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations
Nosocomial infections among confirm cases of COVID-19	2	179	Serious1	Not serious	Not serious	Serious3	None	44% (36%, 51%)	⊕⊕○○ low
Nosocomial infections among confirm cases of SARS	6	3,610	Serious1	Serious2	Not serious	Not serious	None	36% (23%, 49%)	⊕⊕○○ low
Nosocomial infections among confirm cases of MERS	6	1,049	Serious1	Serious2	Not serious	Serious3	None	56% (8%, 100%)	⊕○○○ very low
Health care workers among confirmed cases of COVID-19	2	179	Serious1	Not serious	Not serious	Serious4	None	33% (27%, 40%)	⊕⊕○○ low
Health care workers among confirmed cases of SARS	6	3,662	Serious1	Serious2	Not serious	Not serious	None	37% (25%, 49%)	⊕⊕○○ low
Health care workers among confirmed cases of MERS	6	1,049	Serious1	Serious2	Not serious	Not serious	None	19% (4%, 35%)	⊕⊕○○ low
Excluding health care workers among confirm cases of COVID-19, SARS and MERS	2	589	Serious1	Not serious	Not serious	Serious4	None	2% (1%, 3%)	⊕⊕○○ low
Excluding health care workers among confirm cases of SARS	4	267	Serious1	Serious2	Not serious	Serious4	None	24% (10%, 38%)	⊕○○○ very low
Excluding health care workers among confirm cases of MERS	6	1,049	Serious1	Serious2	Not serious	Serious3	None	36% (6%, 67%)	⊕○○○ very low
Doctors among hospital staff with COVID-19	1	79	Serious1	Not serious	Not serious	Serious4	None	33% (24%, 44%)	⊕⊕○○ low
Doctors among hospital staff with SARS	12	865	Serious1	Serious2	Not serious	Serious4	None	30% (19%,40%)	⊕○○○ very low
Doctors among hospital staff with MERS	3	20	Serious1	Not serious	Not serious	Serious3	None	35% (14%, 56%)	⊕⊕○○low
Nurses among hospital staff with COVID-19	1	79	Serious1	Not serious	Not serious	Serious4	None	56% (45%, 66%)	⊕⊕○○ low
Nurses among hospital staff with SARS	11	861	Serious1	Not serious	Not serious	Serious4	None	50% (45%, 55%)	⊕⊕○○ low
Nurses among hospital staff with MERS	3	20	Serious1	Not serious	Not serious	Serious3	None	50% (29%, 71%)	⊕⊕○○ low
Staff other than doctors or nurses among hospital staff with COVID-19	1	79	Serious1	Not serious	Not serious	Serious4	None	11% (6%, 20%)	⊕⊕○○ low
Staff other than doctors or nurses among hospital staff with SARS	11	846	Serious1	Serious2	Not serious	Serious4	None	21% (12%, 29%)	⊕○○○ very low
Staff other than doctors or nurses among hospital staff with MERS	2	17	Serious1	Not serious	Not serious	Serious4	None	16% (0%, 32%)	⊕⊕○○ low
Health care staff with SARS who did not wear protective clothing	5	222	Serious1	Serious2	Not serious	Serious4	None	63% (35%, 92%)	⊕○○○ very low
Health care staff with SARS who did not wear gloves	3	81	Serious1	Not serious	Not serious	Serious3	None	58% (39%, 76%)	⊕⊕○○ low
Health care staff with SARS who did not wear goggles	3	81	Serious1	Not serious	Not serious	Serious4	None	91% (80%, 102%)	⊕⊕○○ low
Health care staff with SARS who did not take hand disinfection measure	3	81	Serious1	Not serious	Not serious	Serious3	None	57% (0%, 100%)	⊕⊕○○ low
Health care staff with SARS who did not wear masks	3	81	Serious1	Not serious	Not serious	Serious4	None	7% (0%, 16%)	⊕⊕○○ low

1, downgrade one level: the risk of bias is high due to the limitations of study design. 2, downgrade one level: heterogeneity of data synthesis results, I2>50%. 3, downgrade one level: the confidence interval is too wide. 4, downgrade one level: the sample size is too small. CI, confidence interval; CS, cross-sectional study.

Discussion

Our rapid review identified a total of 40 studies. Low to very low-quality evidence indicated that the proportion of nosocomial infection among confirmed cases of COVID-19 was 44%, which is higher than for SARS but lower than for MERS. Most patients with COVID-19 and SARS infected in hospitals were medical staff, among whom nurses formed the largest group, followed by doctors. Both SARS and MERS outbreaks have been reported in hospitals, but we found no evidence of a COVID-19 outbreak.

SARS-CoV-2, the infectious agent causing COVID-19, is highly contagious, mainly spread by droplets and close contact. So far, a number of familial disease clusters have been reported, and some of the confirmed patients had been infected in healthcare facilities. As health care workers are in contact with a large number of suspected patients on a daily basis, strict precautions need to be taken to avoid outbreaks of infection in health care facilities. In the early stage of the epidemic, some hospitals, staff or publics did not have enough knowledge about the virus, leading to inadequate prevention and control measures, which may explain the reasons why the proportions of nosocomial infection are high in our study. The proportions may be higher than the real ones because the data of COVID-19 were from the early outbreak in Wuhan. When COVID-19 broke out in Wuhan at the beginning, medical resources were scarce, and various protective measures and management of hospitals were not in place, resulting in a high rate of nosocomial infections. Suspected patients did often not take any protection measures when they went to the hospital, which may have caused nosocomial infections and hospital outbreaks (19,20). A MERS study showed routine infection-prevention policies can greatly reduce nosocomial transmission of MERS (57). According to a report by the WHO, 20% of confirmed cases of SARS were among health care workers (58). Due to the rapidly evolving outbreak and spread of the disease, medical staff need to work in a state of high tension, but they should also protect themselves adequately and take the appropriate isolation measures to avoid cross infection in the hospital.

The high presence of the COVID-19 epidemic in the media is likely to improve the general public’s awareness. People with symptoms indicating a SARS-CoV-2 infection should take protective measures during the hospital or clinic visit, such as wearing a mask, minimizing the time of stay in the hospital, and if possible, making remote medical consultation in advance. Medical institutions should formulate sound infection prevention and control strategies, and strengthen the hospital's infection prevention and control efforts, such as the establishment of special departments for outpatients with fever, and a sound triage system: triage of early identification among suspected cases can avoid excessive gathering of patients in the hospital. Isolation wards should be established for suspected and confirmed patients needing treatment. In hospitals without single isolation wards or negative pressure isolation, indoor ventilation measures should be taken timely, and the management of patients should be standardized in these wards. Using adequate disinfection procedures can reduce the possibility of hospital transmission of the virus. During the epidemic, efforts should be made to publicize the knowledge of infection prevention and control, be alert to the possibility of the outbreak of nosocomial infection, and establish an early warning mechanism. Emergency plans or measures should be developed to deal with nosocomial infections.

Strengths and limitations

Our study included studies related to nosocomial infections among COVID-19, SARS and MERS patients. Our results can help the decision-making related to prevention, control and clinical management in hospitals. Some studies had missing data, and we used methods of meta-analyses of proportions to analyse those studies with available data, so the proportions estimated may not be accurate and similar to the actual data. Most of the results are based on low-quality research, so that the credibility of the results is low.

Conclusions

A large proportion of confirmed cases of COVID-19 were infected within healthcare facilities. Therefore, the patients who come to the hospital should do pay attention on personal protection. At the same time, medical institutions can reduce the spread of the virus through triage, and setting up separate fever clinic and isolation wards. Awareness of the disease needs to be improved among medical staff, so that they can protect themselves adequately and stop the spread of the virus within hospitals.

The article’s supplementary files as