Face masks to prevent transmission of respiratory diseases: Systematic review and meta-analysis of randomized controlled trials *

  1. Hanna M. Ollila
  2. Markku Partinen
  3. Jukka Koskela
  4. Riikka Savolainen
  5. Anna Rotkirch
  6. Liisa T. Laine

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Objective

To examine the effect of face mask intervention in respiratory infections across different exposure settings and age groups.

Design

Systematic review and meta-analysis.

Data sources

PubMed, Cochrane Central Register of Controlled Trials, and Web of Science were searched for randomized controlled trials investigating the effect of face masks on respiratory infections published by November 18th, 2020. We followed PRISMA guidelines.

Eligibility criteria for selecting studies

Randomized controlled trials investigating face masks in respiratory infections across different exposure settings. Two reviewers performed the search, extracted data, and assessed the risk of bias. Random effects meta-analysis with risk ratio, adjusted odds ratios, and number needed to treat were performed. Findings by source control or wearer protection, age groups, exposure settings, and role of non-compliance were evaluated.

Results

Seventeen studies were included, (N=11,601 cases and N=10,286 controls, follow-up from 4 days to 19 months). Fourteen trials included adults and children and three trials included children only. Twelve studies showed non-compliance in treatment and eleven in control group. Four studies supported the use of face masks. Meta-analysis across all studies with risk ratios found no association with number of infections (RR=0.957 [0.810 − 1.131], p=0.608). Meta-analysis using odds ratios adjusted for age, sex, and vaccination (when available) showed protective effect of face masks (OR=0.850 [0.736 − 0.982], p=0.027). Subgroup meta-analysis with adjusted odds ratios found a decrease in respiratory infections among adults (14 studies, OR = 0.829 [0.709 − 0.969], p=0.019) in source control setting (OR = 0.845 [0.7375 − 0.969], p=0.0159) and when face masks were used together with hand hygiene OR = 0.690 [0.568 − 0.838], p=0.0002). Overall between-study heterogeneity was large also in the subgroup analyses.

Conclusion

Despite the large between study heterogeneity, compliance bias and differences by environmental settings, the findings support the use of face masks to prevent respiratory infections. PROSPERO registration number CRD42020205523 .

Article activity feed

  1. Version published to 10.1101/2020.07.31.20166116v3 on medRxiv
  2. ScreenIT

    SciScore for 10.1101/2020.07.31.20166116: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Institutional Review Board Statementnot detected.
    RandomizationWe limited the searches to randomized controlled trials on humans published by November 18th 2020.
    BlindingThe following characteristics were evaluated: Random sequence generation, allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and non-compliance in the treatment group and control group.
    Power Analysisnot detected.
    Sex as a biological variablenot detected.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    2.1 Search strategy: We performed the searches using the Cochrane Central Register of Controlled Trials (CENTRAL), PubMed, and Web of Science (science and social science citation index).
    Cochrane Central Register of Controlled Trials
    suggested: (Cochrane Central Register of Controlled Trials, RRID:SCR_006576)
    PubMed
    suggested: (PubMed, RRID:SCR_004846)
    We performed the PubMed search using Medical Subject Headings (MeSH) listed in Supplement A.
    MeSH
    suggested: (MeSH, RRID:SCR_004750)
    2.4 Risk-of-bias assessment: Two review authors (HMO and LTL) independently assessed the risk of bias using the Cochrane Risk of Bias tool.[
    Cochrane Risk of Bias
    suggested: None
    Following the Cochrane tool for risk assessment, we denoted the overall risk of bias as low if all the categories were at a low risk of bias, high if at least one domain was at a high risk of bias.
    Cochrane tool
    suggested: None

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    With these limitations in mind, we calculated that, for respiratory infections, the NNT might range from 23 to 455. To put this into context, let us presume that, in a metropolitan area with a population of one million, 30% of the residents use face masks. With NNT=455, this might prevent 600 respiratory infections. This effect size is comparable to the NNT of pharmaceuticals. For example, the NNT for statin, one of the most widely prescribed drugs, in primary prevention of major vascular events at low levels of a CVD risk (5-10% within 5 years) ranges from 67 to 170 and is of a similar scale to face masks [36]. We show that the studies where hand hygiene was assessed together with mask use, effects with multiplicative protective measures were seen. Our results support use of face masks in preventing respiratory infections and, hence, the WHO guidelines that recommend the use of face masks together with physical distancing and hand hygiene in controlling the spread of COVID-19. 4.3 Limitations: First of all, the population studied here had residual heterogeneity. Indeed, as respiratory infections are time- and exposure-dependent, it is possible that differences in follow-up times and in symptom assessments (ILI, respiratory illness or COVID-19) have affected the power to detect associations. Second, while all the studies reported the numbers in the treatment and control arms, we did not have access to raw data and could not adjust the analysis by within-study variables. As a ...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  3. Version published to 10.1101/2020.07.31.20166116v2 on medRxiv
  4. Version published to 10.1101/2020.07.31.20166116v1 on medRxiv