select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

expert reaction to study looking at face masks in the home

A study, published in BMJ Global Health, looks at face masks use in the home.


Dr Antonio Lazzarino, Department of Epidemiology and Public Health, UCL, said:

“This study is not robust science, as it has several limitations in the conception and in the statistical analysis.

“The main limitation is that it was designed at the family level, rather than at the individual level.  This kind of flaw is called ecological fallacy.  In brief, the authors have not considered the individual risk of being infected with or without masks; they have instead assessed if families with the habit of wearing masks had more/less cases of infection, assuming that family members could be infected only by their family case.  Families who have that habit may have lower chances of infection for reasons that have nothing to do with masks.  For example, they may be based in areas where the spread of the virus is lower, or they may tend to isolate more.

“Furthermore, this is a questionnaire-based retrospective study, therefore potentially affected by recall bias.  Moreover, there was no distinction between a N95 mask, disposable surgical mask or just a common mask including cloth mask.

“In conclusion, this study alone is not enough to make recommendations of any kind.”


Prof Sally Bloomfield, Honorary Professor, London School of Hygiene and Tropical Medicine, and International Scientific Forum on Home Hygiene (, said:

“As someone who is asked to give advice to the public about social distancing and hygiene behaviour in their homes, this is an important paper because it comes at a time when – as lockdown is eased – the risk of a person entering the home who has become infected (e.g. whilst on public transport or in the workplace) but is unaware that this is so, is increasing.  Since “household transmission is a major driver of epidemic growth” the behaviour of family groups is likely to play a significant part in keeping the R value below 1.

“In giving advice to the public, one of the most difficult things has been to get people to accept is that behaving, at all times, in a manner whereby they do not spread infection, even though they may feel perfectly well, is as important as protecting themselves from becoming infected.

“It is becoming clear that the number of people who show clinical symptoms and test positive is the tip of an iceberg, but until we start track and trace testing we have no idea what the relative proportions might be.  The recent study1 of an outbreak on a “closed household” of 217 people on board a cruise liner where all passengers were tested, only 24 became showed symptoms out of 128 who tested positive.  This suggests that up to 81% of people in the community who are infected and thus potentially infectious may be asymptomatic i.e. are unaware they are infectious.

“The study will be very useful for persuading people that social distancing and strict hygiene in their homes can really work to reduce the risk of infection transmission to other household members but I think it could be very difficult to persuade families to wear masks in their own homes.  The most difficult thing will be to get them to adopt mask wear and social distancing where there is no evidence that any of the members of the household might be infectious, least of all the idea that “it could be you”.  This is despite the fact that the research suggests that this is the critical time i.e. the researchers found that “This practice was 79% effective at curbing transmission before symptoms emerged in the first person infected, but it wasn’t protective once symptoms had developed”.

“The weakness of the study lies in the fact that behaviours were self reported, but, the advantage is they were being asked to self report about a specific recent time period.”



Prof Babak Javid, Principal Investigator, Tsinghua University School of Medicine, Beijing, and Consultant in Infectious Diseases at Cambridge University Hospitals, said:

“This is an interesting and important study.  We still don’t know for sure the transmission routes for Covid.  The best way to investigate is by careful contact tracing in circumstances where the transmission chain is fairly clear.  Beijing had a relatively small number of Covid cases, and due to the response of the population to the Covid situation, most transmission within Beijing occurred in households, making study of household transmission in this setting extremely useful, since it was very unlikely that transmission occurred outside of the household in January and February 2020.  In this study, the authors investigated all cases of Covid-19 in Beijing: there were a total of 399 cases involving 181 families.  They were able to look in detail at transmission within 124 of these families in which at least one member (the “primary case”) had Covid.  In all cases, as was the practice in Beijing, the primary case was hospitalised, regardless of severity of illness, after the diagnosis was made.  In only 41/124 families were secondary cases identified: i.e. the primary case transmitted Covid to other family members. resulting in 77 additional cases.  This may be an under-estimate, since totally asymptomatic transmission was not considered at that time in China as meeting the case definition of Covid.

“The 77 cases were out of a total of 335 family members, i.e. the primary case transmitted Covid to approximately a quarter of family members, more commonly to adults than children, and 90% of secondary cases were mild.  The study authors contacted all families several weeks after the outbreak in Beijing and conducted a telephone interview and questionnaire about which behaviours they engaged in, and performed a rigorous statistical analysis for factors associated with either increased or decreased risk of secondary transmission.  Only 4 factors were robustly associated with altered risk.  Two factors increased risk a) whether the primary case had diarrhoea or not: this was associated with increased risk, probably due to sharing of toilets by family members, b) lack of isolation of the primary case from family members, such as close daily contact to have meals, watch TV etc.  Two behaviours were associated with significantly decreased risk: a) disinfecting the home every day with alcohol and bleach and b) wearing a mask by the primary case and/or by the family members, but only if before the primary case showed symptoms of illness.

“The single biggest risk factor was prolonged close contact: which was associated with 18-fold increased risk of transmission.  Both mask use and disinfecting the home was associated with 80% decreased risk of transmission.  Importantly, most households practiced good hand hygiene, but this was not associated with protection against transmission.

“This study highlights a number of factors: first, that most transmission occurs before patients show symptoms.  Waiting for people to develop symptoms will not help in breaking the transmission chain.  Secondly, there is a current debate as to whether Covid transmission is by aerosol (v. small droplets).  These data strongly suggest it is not, since isolation of the primary case within the home, but by simple physical distancing was associated with substantially decreased risk of transmission.   Thirdly, mask use, even within homes, can reduce transmission, but only if worn before the primary case shows symptoms.  This has obvious pragmatic implications, since it will be hard to enforce, but suggests that masks are most useful in preventing transmission from presymptomatic individuals.  Finally, the use of detergents/ cleaning surfaces was also protective: this suggests some transmission of Covid is by contamination of surfaces, and again, has wider-ranging public health interventions.”


Prof Trish Greenhalgh FMedSci, Professor of Primary Health Care Services, University of Oxford, said:

“This is a study from China which looked at various measures in the home to reduce transmission of COVID-19.  Of all the measures tested, mask-wearing seemed the most effective.  Disinfecting, opening windows and keeping 1 metre apart also seemed to help.  The mask-wearing is perhaps the most interesting as it’s something that few people currently do in their own homes, especially when not symptomatic.  Worryingly, whereas people who wore a mask before they became symptomatic with COVID-19 were significantly less likely to pass it on to others in the household, those who only started wearing the mask after they became symptomatic weren’t able to protect their family members.

“This study is based on a telephone survey undertaken retrospectively, so is highly susceptible to recall bias.  It offers relatively weak evidence to support the claim that people would be less likely to become infected if they wear masks all the time while at home.

“The findings must also be weighed against the practicalities and the absolute risks involved.  In localities where the incidence of COVID-19 is high, people might feel this measure is worth the inconvenience, especially if there is a vulnerable family member.  Where the incidence of the disease is low, the balance between benefits and hassles may lean more towards the latter.”



Reduction of secondary transmission of SARS-CoV-2 in households by face mask use, disinfection and social distancing: a cohort study in Beijing, China’ by Yu Wang et al. was published in BMJ Global Health at 23:30 UK time on Thursday 28 May 2020.

DOI: 10.1136/bmjgh-2020-002794


All our previous output on this subject can be seen at this weblink:


Declared interests

None received.

in this section

filter RoundUps by year

search by tag