Effectiveness of infection prevention and control interventions, excluding personal protective equipment, to prevent nosocomial transmission of SARS-CoV-2: a systematic review and call for action

Summary Many infection prevention and control (IPC) interventions have been adopted by hospitals to limit nosocomial transmission of SARS-CoV-2. The aim of this systematic review is to identify evidence on the effectiveness of these interventions. We conducted a literature search of five databases (OVID MEDLINE, Embase, CENTRAL, COVID-19 Portfolio (pre-print), Web of Science). SWIFT ActiveScreener software was used to screen English titles and abstracts published between 1st January 2020 and 6th April 2021. Intervention studies, defined by Cochrane Effective Practice and Organisation of Care, that evaluated IPC interventions with an outcome of SARS-CoV-2 infection in either patients or healthcare workers were included. Personal protective equipment (PPE) was excluded as this intervention had been previously reviewed. Risks of bias were assessed using the Cochrane tool for randomised trials (RoB2) and non-randomized studies of interventions (ROBINS-I). From 23,156 screened articles, we identified seven articles that met the inclusion criteria, all of which evaluated interventions to prevent infections in healthcare workers and the majority of which were focused on effectiveness of prophylaxes. Due to heterogeneity in interventions, we did not conduct a meta-analysis. All agents used for prophylaxes have little to no evidence of effectiveness against SARS-CoV-2 infections. We did not find any studies evaluating the effectiveness of interventions including but not limited to screening, isolation and improved ventilation. There is limited evidence from interventional studies, excluding PPE, evaluating IPC measures for SARS-CoV-2. This review calls for urgent action to implement such studies to inform policies to protect our most vulnerable populations and healthcare workers.


S U M M A R Y
Many infection prevention and control (IPC) interventions have been adopted by hospitals to limit nosocomial transmission of SARS-CoV-2. The aim of this systematic review is to identify evidence on the effectiveness of these interventions. We conducted a literature search of five databases (OVID MEDLINE, Embase, CENTRAL, COVID-19 Portfolio (preprint), Web of Science). SWIFT ActiveScreener software was used to screen English titles and abstracts published between 1st January 2020 and 6th April 2021. Intervention studies, defined by Cochrane Effective Practice and Organisation of Care, that evaluated IPC interventions with an outcome of SARS-CoV-2 infection in either patients or healthcare workers were included. Personal protective equipment (PPE) was excluded as this intervention had been previously reviewed. Risks of bias were assessed using the Cochrane tool for randomised trials (RoB2) and non-randomized studies of interventions (ROBINS-I). From 23,156 screened articles, we identified seven articles that met the inclusion criteria, all of which evaluated interventions to prevent infections in healthcare workers and the Introduction Since the beginning of the COVID-19 pandemic, hospitalacquired SARS-CoV-2 infections have been reported across the world [1e5]. Healthcare workers are at higher risk of SARS-CoV-2 infection than the general population [6,7], in turn increasing the risk of transmission to their patients, coworkers and household members [6,8]. Hospitalised patients are often older and have more comorbidities than the general population [3] and hence are at a higher risk of becoming seriously ill [9].
SARS-CoV-2 infection is transmitted through close contact with an infected individual via droplet transmission, as well as via airborne and fomite transmission [10]. In healthcare settings, interventions such as mask wearing by patients and healthcare workers [11e13], screening of patients [14e16] and healthcare workers [17e19], and triaging of patients [20,21] have been implemented to reduce transmission of SARS-CoV-2. These interventions remain relevant despite emergence of new variants and uncertainties in effects of vaccines on transmissibility.
To determine the evidence for these interventions, we conducted a scoping review up to 28th January 2021 (Appendix A) for reviews of the effectiveness of IPC strategies to reduce SARS-CoV-2 transmission in hospital-based populations. We identified reviews of physical distancing [22], and masks [22e26] though none performed a meta-analysis due to a lack of intervention studies. Most notably, we identified a living rapid review by Chou et al. [23] reviewing the effectiveness of masks in health care and community settings in preventing transmission of respiratory viruses including SARS-CoV-2. Given the presence of such effort, we decided to exclude personal protective equipment (PPE) as part of our interventions under investigation. Instead, we focused on wider IPC measures, such as screening protocols, triaging, cohorting, ventilation, or physical barriers to transmission, for which we found no reviews.
Hence, the aim of this systematic review is to collate and assess evidence on the effectiveness of IPC interventions, excluding PPE, intended to reduce transmission of SARS-CoV-2 between patients, between patients and healthcare workers, and between healthcare workers within the hospital setting.

Methods
We report the results of this systematic review following the guidelines of the Preferred Reporting Items for Systematic Reviews and Network Meta-Analyses (PRISMA) checklist [27].

Protocol and registration
This systematic review is registered with the International Prospective Register of Systematic Reviews (PROSPERO) with the registration number CRD42021246617.

Eligibility criteria
We included intervention studies as defined by the Cochrane Effective Practice and Organisation of Care (EPOC) group [28]. These studies include randomised trials, non-randomised trials, cluster randomised trials, repeated measures studies, interrupted time series studies, and controlled before-after studies. Observational studies were excluded.
We included any IPC strategy except personal protective equipment (PPE). PPE, as defined by the World Health Organisation in the context of COVID-19 includes medical masks, gloves, face shields, gowns, respirators (such as N95 or FFP2 standard or equivalent) and aprons [29]. Only studies that evaluated outcomes indicative of SARS-CoV-2 infections including clinical or laboratory confirmed diagnoses were included.
The population of interest included hospitalised patients and healthcare workers working in hospitals and thus were at risk of hospital-acquired SARS-CoV-2 infections. We included only English language studies.
We excluded studies that were published as letters, editorials, opinions, or brief communications.

Information sources
Public Health England Knowledge and Learning Services conducted a search of Ovid Medline, Embase, CENTRAL, COVID-19 Portfolio (preprints), and Web of Science databases including studies between 1 January 2020 and 6 April 2021. The search strategy included terms to identify studies on SARS-CoV-2 infections, IPC, and the hospital setting. Each group of these terms was combined with an 'AND'. A comprehensive search strategy is available in Appendix B.

Study selection
We used the software SWIFT ActiveScreener [30] to screen titles and abstracts. Each record was reviewed by two reviewers split between four reviewers (CL, GK, MY, YJ). SWIFT ActiveScreener uses active learning to incorporate user feedback during the screening process and prioritize articles. A negative binomial model then identifies the number of relevant articles remaining [30]. Using this software, we reviewed a subset of articles until we reached a 95% estimated recall which is the probability of having included relevant articles. Conflicts were resolved by consensus between the two reviewers of each record.
Each included full text was reviewed by two reviewers split between four reviewers (CL, GK, MY, YJ) and conflicts were resolved by discussion.

Data collection and data items
We collected information on study characteristics (publication status, study design), intervention characteristics, population characteristic (healthcare workers, patient, age, sex, co-morbidities), outcomes (clinical, laboratory based), and results (infections in study arms).

Risk of bias in individual studies
Risk of bias in RCTs were assessed using the RoB2 tool [31]. The domains included risk of bias arising from the randomisation process, risk of bias due to deviations from the intended interventions, risk of bias due to missing outcome data, risk of bias in measurement of the outcome, risk of bias in selection of the reported result, followed by an overall risk of bias. A value judgement of low, high, or some risk of bias concern was assigned to each domain.
The risk of bias in non-RCTs were assessed using the ROBINS-I tool [32]. The ROBINS-I tool was used to assess bias due to confounding, bias in selection of participants into the study, bias in classifications of interventions, bias due to deviations from intended interventions, bias due to missing data, bias in measurements of outcomes, and bias in selection of the reported results, and an overall risk. A value judgement of low, moderate, serious, or critical risk of bias or no information was assigned to each domain.
Data from each article was extracted by one reviewer (YJ) and reviewed by a second reviewer (MY). Conflicts were resolved by discussion.

Summary measures
We performed a descriptive analysis as the interventions were not comparable and were too few to conduct a meta-analysis.

Results
Our initial search identified 35,158 records ( Figure 1). After de-duplication, 23,711 records were identified and uploaded to identification screening included Figure 1. Study selection. SWIFTActive Screener. We (CL, MY, GK, YJ) double screened 8,807 (37%) records, and reached 97% recall. Through this process, 238 records which matched our search criteria were identified, as well as 531 more duplicates identified during the review process resulting in a final of 228 full records to be reviewed. Of these, seven records met the inclusion criteria ( Figure 1). Six of these studies were randomised control trials while one was a non-randomised trial. Studies were conducted in Argentina [33], Canada [34], India [35], Pakistan [36], Spain [37], and the USA [34,38]. All studies evaluated efficacy of interventions in healthcare workers only, with no studies conducted in patient populations.
We found no studies which evaluated the effectiveness of other widely used strategies including isolation, cohorting of patients and staff, improved ventilation strategies, use of air filters, or enhanced environmental cleaning.

Summary of studies
Studies did not have comparable interventions, and hence a meta-analysis was not conducted. Summary details of studies are presented in Table I, with a more detailed extraction in Appendix C.
A non-randomised trial [36] evaluated the effectiveness of an audio-visual triage system separating the patient and staff by a glass barrier, at a distance of six feet, and connected by an audio system. Sixty staff were enrolled, with a PCR test used to identify SARS-CoV-2 infections. The authors found a statistically significant lower rate of infection in the intervention group (intervention: 3/30 (10%), control: 9/30 (30%), P¼0.001).
Two studies [34,38] noted that authors received income from pharmaceutical companies, not funding the work related to the studies. In one study [37], a pharmaceutical company partly funded the study while in another study [35] it was unclear whether a drug under investigation was donated. One study [36] did not include a statement on conflict of interest while another [33] did not identify any conflicts of interest.

Risk of bias
Using RoB2 for the RCTs, two studies [33,39]   of bias due to deviation from intended interventions [28,32,33] ( Figure 2). All studies had a low risk of bias due to missing outcome data. One study [39] had a high risk of bias in the measurement of the outcome, while all studies had a low risk of bias of the selection of the reported results. Three studies [33,35,39] had a high overall risk of bias. Results of individual studies assessed using the RoB2 are presented in Table II.
Using the ROBINS-I tool to assess the risk of bias in one included non-RCT [36], the risk of bias was considered moderate for confounding and selection bias, and low risk for classification of interventions, reporting bias, deviations from interventions, missing data, and measuring outcomes. Overall risk of bias was moderate. More detail can be found in Table III.

Discussion
We found seven studies matching our inclusion criteria, of which six focused on evaluating prophylactics and one study evaluated an audio-visual system during triage. None of the six proposed agents have good evidence for efficacy in preventing infection with SARS-CoV-2 [40]. Notably, we did not find any studies with appropriate study design for assessing interventions proposed by international [10,41] and national [42e45] guidelines and adopted by some hospitals, including cohorting, screening, isolating, use of single rooms, and use of environmental and engineering strategies such as improved ventilation and use of filters. Similarly to a review on the effectiveness of ultraviolet-C (UVeC) in hospitals [39], we did not find any studies using UV-C or any other sterilisation strategies to prevent transmission of SARS-CoV-2. Consequently, generating high quality evidence for IPC against nosocomial transmission of SARS-CoV-2 continues to be a priority, particularly for costly interventions or interventions that have the potential to cause harm such as mass administration of prophylaxis to healthcare workers [46].
All identified studies only included healthcare workers, and the majority evaluated pharmaceutical interventions. Three studies evaluating hydroxychloroquine did not find a prophylactic effect and were stopped early. This was similar to the findings of Lewis et al. [18,19] and Bartoszko et al. [22] which reviewed effectiveness of hydroxychloroquine as prophylaxis though not specifically in a hospital setting. Bartoszko et al. [18], through their systematic review of effectiveness of prophylaxes for prevention of SARS-CoV-2 infections in all settings, included the same trial on the evaluation of ivermectin and iota-carrageenan as found here. Similar to their findings, we found that despite the demonstrated impact of ivermectin and iota-carrageenan as a prophylaxis to reduce SARS-CoV-2 transmission, the quality of evidence was low due to high risk of bias, mainly due to the lack of blinding. Studies evaluating bromhexine hydrochloride [39] and chyawanprash [35] also suffered from high bias due to lack of blinding of interventions.
In the non-RCT, non-pharmaceutical intervention trial that evaluated an audio-visual system [36], the authors found a significant decrease in infections in the intervention arm. There was, however, a moderate risk of confounding with analysis only including a small number of covariates and selection bias with a poor description of recruitment strategy. Hence, this evidence does not support implementation of the evaluated system.
Chou et al. [23] last updated their live systematic review on effectiveness of respirators, face masks, and cloth masks to prevent SARS-CoV-2 infections in healthcare settings in July 2021 [47]. Their inclusion criteria included RCT and observational studies. They did not identify any RCTs through their search but identified ten observational studies. However, due to methodological limitations of the studies and heterogeneity of comparisons and results, meta-analysis was not conducted. Therefore, evidence on effectiveness of masks to prevent transmission of SARS-CoV-2 in healthcare settings remains insufficient.
Since there is limited evidence to guide hospital practice for the prevention of nosocomial transmission of SARS-CoV-2, learnings can be taken from studies collating evidence on the effectiveness of interventions in preventing hospital-acquired infections of other respiratory viruses. A review on infection and prevention strategies to prevent seasonal influenza [42] found limited evidence to support screening HCWs, patients, and visitors and isolation of infectious individuals. Evidence of intervention strategies for other types of hospital-acquired infections can also be informative. A review [48] of the effectiveness of cohorting in reducing transmission of C. difficile and multi-drug organisms found limited evidence in its effectiveness despite its widespread use.
This review highlights the lack of available evidence on the effectiveness of IPC strategies for SARS-CoV-2 infection in hospitals indicating further research is needed. This could be a result of increased workload of hospital and public health staff, inadequate understanding of transmission dynamics of the virus in the hospital setting, and challenges in implementing IPC interventional studies (e.g., many interventions involve behaviour change which is difficult to attain and assess, institutional commitment for systemic change, interventions are usually executed in bundles which are resource-intensive). These may have prevented timely and appropriate implementation of important clinical research to assess interventions to inform practice.
With an increased understanding of SARS-CoV-2 transmission and disease and more managed workload in many settings, there should be a renewed focus on evaluating IPC practices. For example, interventions to be investigated include type and frequency of patient and HCW screening, isolation practices, cohorting strategies of both patients and HCWs, improved ventilation and methods for air circulation, and environmental cleaning. Many of these are already suggested as part of the infection prevention and control recommendations by Public Health England [49]. The importance of SARS-CoV-2 vaccination status for both patients and HCWs could also be investigated and would be an important confounder in any future intervention study.
Furthermore, enhancing national and local guidelines on implementing interventional studies to evaluate IPC practices would greatly improve the feasibility, ease, and timeliness of their implementations and ensure appropriateness of studies particularly when such research may not be deemed a priority. Researchers should follow the Standards for Reporting Implementation Studies (StaRI) Statement and Checklist [50], which applies to a range of study designs, to help clarify planning and implementation and to ensure accurate and transparent reporting of studies. The StaRI statement consists of a checklist of 27 items, describing both the implementation strategy and the implemented intervention.
There were some limitations with our review. Firstly, we only included English language studies. The COVID-19 pandemic has occurred in every country in the world, each with their unique healthcare delivery system, practices, and challenges. By excluding non-English studies, we may have missed studies with transferrable evidence between settings. Secondly, we did not include studies evaluating the effectiveness of PPE interventions. Some of these studies, however, may have evaluated other interventions along with PPE but may not have explicitly identified them in keywords or abstract. Thirdly, due to the large number of identified records, we used a software, SWIFT-Active Screener, to screen titles and abstracts until we reached a recall probability of 97%. We may have missed some studies by not reaching 100% recall or reviewing all references. However, using this software greatly reduced the effort required to conduct this systematic review and we are confident that we would not have found any additional relevant articles [30].

Conclusions
There is currently very little evidence available on the effectiveness of interventions, excluding PPE, to prevent the spread of SARS-CoV-2 in hospital settings. Our systematic review revealed no appropriate studies specifically targeting prevention of transmission between hospitalised patients. While this likely reflects the pressures health systems have been, and are, under, and that many logical, likely effective IPC measures are in place, without unbiased, intervention studies we cannot say which are optimal and cannot maximise the protection given to our most vulnerable populations and those that care for them. This review underscores the need to generate evidence for IPC interventions in hospitals to prevent transmission of SARS-CoV-2, which could also be applicable to other pathogens and calls for high quality intervention studies to systematically determine which IPC measures to implement in healthcare settings.

Funding sources
This work was supported by a UK Medical Research Council Skills Development Fellowship (grant number MR/P014658/1 to GMK) and the Society for Laboratory Automation and Screening (grant number: SLAS_VS2020 to TMP). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect those of the Society for Laboratory Automation and Screening. This work was also supported by a joint grant from UKRI and NIHR (grant number: COV0357/MR/V028456/1 to GMK, supporting YJ, JMR, BC and JVR, grant number MR/V038613/1 for JMR). Support was also received from the National Institute for Health Research Health Protection Research Unit (NIHR HPRU) in Healthcare Associated Infections and Antimicrobial Resistance at Oxford University in partnership with Public Health