WHAT'S NEW - COVID-19 Risks on Commercial Flights

COVID-19 Risks on Commercial Flights

By Sean Derrig

Much has been written regarding the potential risks of contracting COVID-19 on commercial airline flights and given some of the graphics purporting to show how the virus could spread in an aircraft cabin one can be forgiven for thinking even contemplating a flight is a likely death sentence. 

The question goes beyond the basics how far droplets travel or the likelihood of fomite transmission or the other usual variables. All these play a part but it’s a complex problem in a complex system – there are multiple interacting components which make predictions challenging. 

Also people don’t just sit still during flights, particularly ones lasting longer than a few hours. While there may be social pressure or direct instructions from cabin crew not to stretch their legs passengers will likely visit the lavatory or take items from the overhead bins. This is just one of the layers of complexity; there are many more which is why the literature can be conflicting and confusing.

So, this is a complex question but it’s not one that defies analysis. And it seems – provided sensible hygiene rules are followed – the risk is likely lower than you might think. 

Proximity

The WHO defines ‘contact’ as being within two rows of an infected person for more than 15 minutes (1) but this is rather arbitrary; one often-cited NEJM paper (2) claims that during the 2003 SARS outbreak a passenger aboard a flight from Hong Kong to Beijing infected people well outside the two-row boundary and the WHO criteria “would have missed 45 percent of the patients with SARS”.

A more recent PNAS study came to a different conclusion: that transmission risk was minimal more than one row away from the contagious person (3). That study also suggested that those infected may have become so in the boarding area rather than on the aircraft. 

Yet another recent study suggests COVID-19 is primarily droplet-borne on aircraft but that many studies are biased by pre-flight contact - and notes that on a 15-hour flight from Guangzhou to Toronto an infectious passenger with an active cough did not pass COVID-19 on despite having 25 other passengers within two metres (4). 

So, in the face of conflicting data, how can we go about estimating the likely risk? 

Breaking Down the Question

To estimate the risk to an uninfected passenger from an infected one we need to consider 3 main questions:

• What is the probability that there is a passenger on board who is contagious?

• What is the reduction in risk provided by universal masking etc in preventing a contagious passenger from spreading the virus?

• What is the impact of the relative location on the aircraft of the contagious and uninfected passengers?A statistical model (5) has been developed to estimate the probability of this happening: P = Q x QM x QL where

• P is the probability that a particular uninfected passenger contracts COVID-19 during the flight;

• Q is the probability that a given passenger on the flight has COVID-19;

• QM = the probability that universal mask-wearing on aircraft fails to prevent transmission of COVID-19 and:

• QL = the conditional probability that a contagious passenger transmits COVID-19 to the uninfected one if the mask fails

Quantifying the Risk

The maths above rapidly becomes complicated and unwieldy because there are many variables and confounders: how full is the flight, is the airline blocking middle seats to effect social distancing etc. 

A recent US calculation (5) using this metric suggested a passenger in coach becoming infected on a full flight has a mortality risk of about 1 in 430,000. This means the risk of death from COVID caught on a flight would appear higher than that of dying in a US air crash unrelated to COVID-19 (1 in 34 million (6)) but is still low. In the USA you’re five times more likely to die from a lightning strike and over 4,000 times more likely to die in a motor vehicle accident (7). 

From this, with an assumed mortality risk of 1% we can infer that on a full, narrow-bodied jet the risk of contracting COVID-19 from a nearby passenger is about 1 in 4,300 – or about the same lifetime odds of winning the jackpot in a 49-ball lottery if you play every week. 

But there are other factors which reduce the risk still further. 

• The calculation above is based on US data from late June 2020 when approximately 1 in 120 Americans had COVID-19 on a given day. In the UK it is currently 1 in 2,300 (5).

• The numbers above are based on full flights. Current flight occupancy ex-UK is more like 20% but this will vary from route to route. 

• The assumptions in the model do not account for physical barriers to transmission such as seat backs.

• The estimates presented above do not consider the possibility of infection during boarding and leaving the plane - but because these risks involve short periods potentially near a contagious person they are likely lower than any risk associated with seat proximity.

Also while enclosed or confined spaces are generally considered higher risk, passenger aircraft ventilation systems split the cabin into zones of 5-7 rows. The air is changed every few minutes with 50/50 outside air and filtered, recirculated air: the HEPA filtration will capture 99.97% of all droplets larger than 0.3μm diameter – far smaller than respiratory droplets.

For these reasons the ‘avoid confined spaces’ advice would appear to be less applicable to commercial aircraft than other enclosed spaces. 

Conclusions

While the absolute risk is difficult to accurately quantify, the models seem to suggest the risk of catching COVID-19 on a commercial flight is low and can be diminished further with standard physical and hygiene measures. 

On the aircraft itself there needs to be another passenger or crew member with an active infection which is likely a low probability at current incidence of 1 in 2,300 as well as with current low flight occupancy rates. 

Consider as well that if another person on the flight is indeed infectious, all the other interventions intended to stop spread would need to break down too for you to catch it.  

But precautions need to encompass all aspects of travel, not just the flight. The journey to the airport, waiting in queues for security screening and pre-boarding can also mean potential exposure to the virus. 

While the risk in flying is not zero the models and the literature suggest it is low and can likely be well controlled with masks, assiduous hand and surface hygiene and other standard interventions. 

Take disinfectant wipes for the immediate area, sanitise hands after touching anything at all and do everything possible to stay a sensible distance from others. And remember the air coming from the nozzles above your seat is clean, HEPA-filtered air so leave them on!

References

1. World Health Organisation. Contact tracing in the context of COVID-19. apps.who.int. [Online] 20 May 2020. https://www.who.int/publications/i/item/contact-tracing-in-the-context-of-covid-19.

2. Transmission of the Severe Acute Respiratory Syndrome on Aircraft. Olsen, Sonia J and al, et. 18 December 2003, New England Journal of Medicine, Vol. 349, pp. 2416-2422.

3. Behaviors, Movements, and Transmission of Droplet-Mediated Respiratory Diseases During Transcontinental Airline Flights. Hertzberg, Vicki S and Weiss, Howard. [ed.] Burton H. Singer. 14, 3 April 2018, Proceedings of the National Academy of Sciences, Vol. 115, pp. 3623-3627.

4. Lack of COVID-19 transmission on an international flight. Schwartz, Kevin L, Murti, Michelle and Finkelstein, Michael. 15, 14 April 2020, CMAJ, Vol. 192, p. E410.

5. Barnett, Arnold. Covid-19 Risk Among Airline Passengers: Should the Middle Seat Stay Empty? s.l. : MIT Sloan School of Management, 8 July 2020. medRxiv.

6. Aviation Safety: A Whole New World? Barnett, Arnold. 1, 4 July 2019, Transportation Science, Vol. 54, pp. 1-297.

7. National Safety Council. Odds of Dying. Injury Facts. [Online] 2018. https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/.

8. Office for National Statistics. COVID-19 Infection Survey. [Online] ONS, QVIA, Oxford University and UK Biocentre, 20 July 2020. [Cited: 22 July 2020.] https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulletins/coronaviruscovid19infectionsurveypilot/latest.