Keeping rooms ventilated and taking breaks are critical factors in stopping respiratory viruses like COVID-19 from building up inside during a gathering, new modelling research suggests.
Research published today in PNAS models the spread of respiratory disease during indoor meetings of varying lengths and looked at tradeoffs between masking, break times during meetings, testing, and ventilation.
In a case study of the COVID-19 outbreak following a US community choir practice in March 2020, the authors found that a single break period combined with medium mask compliance may have brought down the number of infections.
The SMC asked experts to comment on the research.
Dr Siouxsie Wiles, microbiologist and Associate Professor, University of Auckland, comments:
“In a new modelling study, Dixit and colleagues found that ventilation and break times were critical for preventing high viral loads from building up in indoor environments. In fact, they found that taking breaks of 10-25 minutes worked as well at reducing viral transmission as when half of the people in the room wore masks. This will be because taking breaks allows any contaminated air in the room to clear before participants return.
“With winter approaching and very few people wearing masks anymore, this study highlights how improving ventilation can prevent infections from airborne viruses and bacteria. For almost a year now, I’ve been carrying a CO2 meter around with me to measure how well or poorly a particular space is ventilated. It’s been fascinating and has highlighted to me just why we should all be wearing masks on public transport. I’d like to see all workplaces and public spaces providing real-time CO2 levels so people can assess just how safe or risky an environment is. I think this information is crucial, especially as the weather cools down and we spend more and more time inside with the windows and doors shut.”
No conflicts of interest.
Dr Amanda Kvalsvig, epidemiologist, University of Otago, Wellington, comments:
“Infection risk is a composite of human, pathogen, and environmental factors, but of the three, environments tend to be the easiest to modify, especially when it comes to airborne infections like Covid-19. The aim of the modelling framework presented in this paper is to explore options for reducing the spread of infections during indoor gatherings.
“The approach described by Dixit and colleagues is a practical one, recognising that there are tradeoffs between factors such as meeting time, air filtration quality, meeting breaks, mask adherence, and group size. This approach can enable decision-makers to ask and answer specific questions about the best ways to manage these tradeoffs.
“The approach appears to be adaptable for New Zealand and it could generate useful strategies to protect the public in the coming winter months. For example, a bus journey in NZ can be modelled as a type of ‘meeting’ where the meeting duration and group size can’t easily be changed. In the longer term, it’s likely that air filters will be needed for optimal protection. In the short term, it’s quicker to achieve an increase in the proportion of people wearing masks. So the policy question here could be to estimate the levels of mask adherence that would be needed for effective protection of drivers and passengers during winter 2023, and use that estimate as a benchmark to evaluate the impact of mask policies and public health messaging.
“All modelling findings are shaped by the baseline assumptions built into the model so before adopting any findings we would need to run a fine-tooth comb over the study assumptions to check that they’re appropriate for NZ and for the variants currently in circulation. But as an overall approach the paper findings suggest that this framework could provide useful policy insights to ensure safe access to a variety of health-, education-, social-, and work settings.
“A final note is that many other infections such as influenza and RSV are spread through the air in the same way as Covid-19, but these infections are generally much less transmissible. What this means is that strategies that effectively reduce Covid-19 transmission in public settings are likely to be even more effective in protecting people from a range of other infections. A systematic approach to safe indoor environments would be a win-win-win situation for population health.”
No conflict of interest declared.
Dr Mikael Boulic, Senior Lecturer, School of Built Environment, Massey University, comments:
“Despite limitations, this modelling study confirmed the right information that was given for NZ school environment. The Ministry of Education website says, “Taking refresh breaks to flush a space with fresh air by fully opening windows and doors for a short time – for example, 5-10 minutes.”
“We might want to increase the break time to maybe 15 minutes (as shown in this paper) to have a better impact on the viral load dynamic.
“Another common sense measure is decreasing the density of people in the classroom and decreasing lesson time. Following on from these results, maybe practical implications could be to extend time at school from 9am/3pm to 9am/4pm to allow an increase of the number of breaks and the break length between lessons (to allow proper refresh of classrooms).”
No conflict of interest declared.
Arindam Basu, Associate Professor of Epidemiology, School of Health Sciences, University of Canterbury, comments:
“It is an interesting paper based on the well-known Wells-Riley equations and updated modelling scenarios. My review is based just on the materials presented in the paper.
“I believe a few assumptions need to be reviewed. First, their assumption that if proportion ‘p’ wear masks, then the transmission probability is likely to be uniform is questionable, because even within the number of people, the transmission likelihood of the viral particles is likely to be different between an infected/infectious person wearing a mask as opposed to a susceptible person wearing a mask. Imagine a scenario where people who do NOT wear masks are largely recently infected individuals and hence infectious, and say 25% people have worn masks all of whom either have been infected in the past or have recovered or have been vaccinated. This scenario is likely to be very different from another situation where p = 25%. But here, people who are infected and near the end of their transmission period (not knowing whether they are infected but infectious anyway) are masked. The transmission likelihood will be different in the two situations, but there is no provision of how these variations are modelled in the modelling scenarios.
“In the paper, they have written, ‘Recovered individuals play no active role within a meeting; they neither exhale virus particles (unlike the infected types) nor do they get infected by inhaling the virus (unlike susceptible individuals) because they have immunity’. This assumption is questionable as with the new variants (XBB.1.5 e.g.), reinfection rates are not clear, which implies that they needed to consider other models than SIR (such as Susceptible Infected Susceptible or Susceptible Exposed Infected Recover models).
“Nonetheless, this finding from their model is significant: ‘Notice that, when there is low mask compliance (P ≈ 0), the infection level is highly sensitive to increments of the meeting length. In contrast, moderate and high mask compliance scenarios show smaller increments in the number of infected individuals, even for daily long meetings’
“This is justifiable and implies the importance of a high level of masking in indoor events, particularly the ones that are likely to span longer durations.
“This finding is also noteworthy along with Figure 6, where they write: ‘Our previous simulations show that the combination of moderate mask compliance levels and meeting length reduction is more effective on reducing the number of cumulative cases generated in recurrent meetings, compared to the combination of mask compliance and group reduction.’
“In summary, if you control for mask compliance, reducing the length of the meeting rather than controlling the number of people in the meeting will likely lead to a more effective drop in the total number of infected individuals.
“My take away from this paper would be that despite limitations, this modelling paper does point to the importance of ventilation, frequent breaks (longer breaks will likely to lead less viral loads), shorter meetings (shorter meetings keeping around 50 minutes at a stretch seems optimal), and wearing of masks (the higher the percentage the less likelihood of infection spread and viral load in the room).”
No conflicts of interest declared.