Physical Distancing, Face Masks, and Eye Protection to Prevent Person-Person COVID-19 Transmission: A Systematic Review and Meta-Analysis

Background: Coronavirus disease 2019 (COVID-19) is spread person to person by virus through close contact. The optimal person-person distance and use of facemasks and/or eye protection to prevent disease transmission in healthcare and non-healthcare settings is debated. Methods: We systematically reviewed the impact of distance, masks, and eye protection on transmission of COVID-19, SARS, or MERS from 21 standard, World Health Organization (WHO)-specific and COVID-19-specific data sources from inception to April 15, 2020 for studies of any design and language 1) comparing different distances between those infected and the people close to them, mask use, or eye protection, or 2) contextual factors of acceptability, feasibility, resource use, and equity of these interventions. We screened studies, extracted data, and assessed risk of bias in duplicate. Frequentist and Bayesian meta-analyses and meta-regression for the main outcome of viral transmission were by random effects. Secondary outcomes were contextual factors. We rated certainty of evidence rating per GRADE. PROSPERO: 177047. Findings: We identified 0 RCTs and 164 relevant observational studies in healthcare and non-healthcare (community) settings from 16 countries across 6 continents. A physical distance of one metre or more compared to less than one metre from those infected was associated with 1) a lower risk of viral transmission (n=7782, 5.3% vs 15.5%; RD -10.2% [95%CI -11.5% to -7.5%], pooled adjusted odds ratio [aOR] 0.18 [95%CI 0.09-0.38], moderate certainty) and 2) incremental benefits with increasing distance, change in relative risk (RR) per metre 1.57 (moderate certainty). Facemask use was associated with less infection (n=2647, 2.7% vs 17.4%; RD -14.% [95% credible interval [CrI] -15.9% to -10.7%]; aOR 0.15 [95%CrI 0.07-0.34], low certainty), with stronger associations with N95 or similar (including powered) respirators compared to disposable surgical or similar (e.g. reusable 12-16-layer cotton) masks, p interaction =0.090; posterior probabilities for RR 95% despite minimally informative priors; moderate certainty). Eye protection was associated with quantitatively similar lower risk of infection in 2 adjusted and 15 unadjusted studies (n=3751, 5.4% vs 16.0%; RD -10.6 [95%CI -12.5% to -7.7%]; RR 0.34 [95%CI 0.22-0.52]; aOR 0.22 [95%CI 0.12-0.39], low certainty). Interpretation: This meta-analysis supports physical distancing by more than one metre and provides quantitative estimates for models and contact tracing to inform policy. Although direct evidence is limited, the optimal use of masks, in particular N95 or similar respirators, may depend on risk assessment and contextual factors. Eye protection may provide significant additional benefits. Globally collaborative, well-conducted studies on preventative and therapeutic strategies are required but are challenging to achieve immediately and, thus, recommendations in the interim to curtail the COVID-19 pandemic should be informed by this systematic appraisal of current evidence. Funding Statement: Commissioned by the World Health Organization as a rapid review on March 25, 2020. The funders of the study helped with defining the scope of the question, but otherwise had no role in study design, data collection, data analysis, data interpretation, writing of the report, or the decision to submit it. Declaration of Interests: ML is an investigator of an ongoing clinical trial on medical masks versus N95 respirators for COVID-19 (NCT04296643). All other authors declare no competing interests. Ethics Approval Statement: The authors prospectively submitted the systematic review protocol for registration on PROSPERO (submission number 177047). This study followed PRISMA and MOOSE reporting guidelines.