Covid-19 lessons from a plagued flight

Written By: - Date published: 10:43 pm, January 11th, 2021 - 22 comments
Categories: covid-19 - Tags: , ,

In September there was a 18 hour flight Dubai that ended in Auckland. A relatively close group on that plane showed infection after arrival, and a genetically similar version. The evidence of in-flight infection is very strong and points to the on-going issues with shipping virus hosts around the world. This has some pretty strong implications for long-haul flights in particular. But also to the risks of having travel ‘bubbles’ without enforced quarantine.

On September 29, 2020, flight EK448, which originated in Dubai, United Arab Emirates, with a stop in Kuala Lumpur, Malaysia, landed in Auckland, New Zealand. During the required 14-day MIQ period, 7 passengers who had traveled on the flight received positive SARS-CoV-2 test results. The 7 passengers had begun their journeys from 5 different countries before a layover in Dubai; predeparture SARS-CoV-2 test results were negative for 5 (Figure 1). These 7 passengers had been seated within 4 rows of each other during the ≈18-hour flight from Dubai to Auckland. Because recent studies have reported conflicting findings of the risks associated with in-flight transmission (2–4), we undertook a comprehensive investigation to determine the potential source of infection of these travelers.

CDC – Emerging Infectious Diseases: “Genomic Evidence of In-Flight Transmission of SARS-CoV-2 Despite Predeparture Testing“

There is a lot of information in the early release article about the process of the travel of the people infected and of the flight. Reading through it, there are some obvious possible holes like mask and glove wearing on the flight. The 2 hour refuelling stop over in Kuala Lumpur had no passengers on or off. However it did have a auxilary power unit off for 30 minutes during that time, and therefore no environmental control during that period – a possible contributing factor.

The abstract states (my italics) in its conclusion..

Among 86 passengers on a flight from Dubai, United Arab Emirates, that arrived in New Zealand on September 29, test results were positive for 7 persons in MIQ. These passengers originated from 5 different countries before a layover in Dubai; 5 had negative predeparture SARS-CoV-2 test results. To assess possible points of infection, we analyzed information about their journeys, disease progression, and virus genomic data. All 7 SARS-CoV-2 genomes were genetically identical, except for a single mutation in 1 sample. Despite predeparture testing, multiple instances of in-flight SARS-CoV-2 transmission are likely.

CDC – Emerging Infectious Diseases: “Genomic Evidence of In-Flight Transmission of SARS-CoV-2 Despite Predeparture Testing“

For me the figures in the article are the most explanatory. So I’m going to put them up.

Figure 1. Countries of travel origins for 7 passengers who tested positive for severe acute respiratory syndrome coronavirus 2 infection after traveling on the same flight (EK448) from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020. Asterisks indicate where 6 other genetically identical genomes have been reported

Figure 1. Countries of travel origins for 7 passengers who tested positive for severe acute respiratory syndrome coronavirus 2 infection after traveling on the same flight (EK448) from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020. Asterisks indicate where 6 other genetically identical genomes have been reported

 

The most notable part of the image is the diversity of locations that the infected passengers came from. The probability of having the same covid-19 genome is very low. But the likely incubation times and times of the tests is even more interesting. 

Figure 2. Timeline of likely incubation and infectious periods, indicating testing dates, for 7 passengers who tested positive for severe acute respiratory syndrome coronavirus 2 infection after traveling on the same flight (EK448) from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020.

Figure 2. Timeline of likely incubation and infectious periods, indicating testing dates, for 7 passengers who tested positive for severe acute respiratory syndrome coronavirus 2 infection after traveling on the same flight (EK448) from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020.

 

As is the positioning of these seven passengers in a spaced out flight. Not to mention the locations of the passengers who didn’t get positive tests. Look at seats 24E, 24G and 28D and 26C – obviously either very lucky or had very nasty immune systems.

Figure 3. Seating arrangement (Boeing 777–300ER) for 7 passengers who tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection on flight EK448 from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020. Passengers F and G interchanged seats within row 24. Open circles represent nearby passengers who were negative for SARS-CoV-2 on days 3 and 12 while in managed isolation and quarantine. All other seats shown remained empty.

Figure 3. Seating arrangement (Boeing 777–300ER) for 7 passengers who tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection on flight EK448 from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020. Passengers F and G interchanged seats within row 24. Open circles represent nearby passengers who were negative for SARS-CoV-2 on days 3 and 12 while in managed isolation and quarantine. All other seats shown remained empty.

 

Figure 4 is the analysis genomic family. It is a extremely close match, especially from a group from such diverse locations.

The final figure is quite interesting as it uses the anticipated times and seating to try to look at possible transmission paths. Look at it in conjunction with figures 2 and 3. 

Figure 5. Network of likely severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission among 7 passengers who traveled on flight EK448 (Boeing 777–300ER) from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020. The gray shaded area illustrates likely in-flight virus transmission. Dashed circles represent likely virus transmission between travel companions.

Figure 5. Network of likely severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission among 7 passengers who traveled on flight EK448 (Boeing 777–300ER) from Dubai, United Arab Emirates, to Auckland, New Zealand, with a refueling stop in Kuala Lumpur, Malaysia, on September 29, 2020. The gray shaded area illustrates likely in-flight virus transmission. Dashed circles represent likely virus transmission between travel companions.

What is notable from this is that it shows the several notable characteristics of this disease.

It is infectious, but nothing like the same order as influenza. In an 18 hour flight even with precautions, a more virulent infection would have probably infected more people in the immediate vicinity. In particular the pattern in row 24 with the two people not getting a positive result is pretty abnormal.

The preflight checks were essentially useless and probably gave a false sense of security for both crew and passengers. That is my primary dislike for such tests – people can and will get tests both before getting infected and also return negative results if they have just been infected. On a long haul flight, they will infect others. The time taken before the disease is detectable by symptoms or test compared to the time to become infections is such that it is really hard to detect for flight onboarding.

That is the primary reason you won’t get my 61yo body in a long haul aircraft anytime soon. They are the ideal infection canisters for covid-19, if not on the way out, then on the way home again.

The same considerations apply for local no-quarantine travel bubbles where there are repeated unexplained community outbreaks – such as we have been seeing repeatably in Australia. Someone who is infected can be happily walking around after testing negative, and without a strong receiving quarantine infect others in-flight or after arrival.

Bearing in mind the financial and economic costs of suppression in the order of what we or the Australians have done to date, and the much smaller costs of strong border controls – this is a economic no-brainer for the country as a whole.

Incidentally, for an individual to have a vaccines don’t make as much of a difference as you’d expect. Ignoring the time to them becoming effective (a not inconsiderable consideration). Then even when and individual is fully vaccinated they are better at stopping the infection from spreading inside the individual than they are in stopping others being infected. The reason is that the bodies defences have noticeable delays at suppression after infection, and many if not most vaccinated people will be infectious.

Personally I don’t want to fly until I’ve had a chance to see what happens in the real world with vaccines in my age group. I sure as hell don’t want vaccinated Aussie tourists or returning kiwis here without quarantine until both countries have a sufficient induced population immunity to reduce the risks to an acceptable level. Certainly not to benefit airlines. tourism operators, and universities. After all they’re not paying me anything to negate the risks I would carry for them.

Just as importantly, this demonstrates the diagnostic utility of getting genomic analysis correlated with travel patterns.

There is more analysis in New York Times “One 18-Hour Flight, Four Coronavirus Infections“. However the title seems like a somewhat optimistic reading of the evidence.

22 comments on “Covid-19 lessons from a plagued flight ”

  • Comments are now closed

  1. lprent 1
    11 January 2021 at 11:05 pm

    On re-read, there were quite a few other lessons that I missed. I will leave them for comments to elucidate.

    But I will take the most obvious.

    Long haul flights obviously need wider separation between passengers. That is pretty good separation. It wasn't enough if people were as masked as they said.

    • Forget now 1.1
      11 January 2021 at 11:47 pm

      Eyeballing that separation, I'd definitely go for a window-seat if I couldn't get out of flying altogether. The aisle side looks a lot riskier, maybe due to bathroom walks by infected passengers? For extra safety you could have a depressurized plane filled with passengers all on independent individual oxygen supplies (scuba tanks is how I picture it, but probably ducted from central tanks). It'd be really cold and unpleasant though.

      To be of much benefit, the 3day test really should be alongside managed isolation in the country of origin. Which I seem to recall Baker including in his initial recommendations, though don't have a source to hand. The logistics and enforcement of that would be a bit tricky though.

      • lprent 1.1.1
        11 January 2021 at 11:57 pm

        Problem is that I don't want to rely on managed isolation in the Ukraine or most of the other places this set of passengers came from.

        I certainly wouldn't trust it from uncontrolled mayhem places like the US.

      • Treetop 1.1.2
        12 January 2021 at 8:33 am

        Probably the toilet would have the highest viral load on a long haul flight.

        I would like to know if swabs are taken from the toilet area and tested?

  2. Incognito 2
    11 January 2021 at 11:26 pm

    Very interesting and a few links to follow up on.

    Regarding the effects of vaccines on prevention or reduction of symptomatic disease (i.e. becoming sick) and prevention of spreading the infection, here’s a nice clarifying article on that:

    https://www.stuff.co.nz/national/health/coronavirus/123886555/covid19-few-vaccines-prevent-infection–heres-why-thats-not-a-problem

    In other words, we don’t know yet which kind of immunity we may enjoy. It could well be different with different vaccines.

    Maybe it is wise to wait with rolling out vaccines in NZ 😉

    • lprent 2.1
      11 January 2021 at 11:53 pm

      Saw a RNZ report on a Otago epidemiologist (not Baker!) saying we may need to do more blockage at the border from places with out of control epidemics. Worried about how thin the border defenses are.

      Problem is that is effectively everywhere in a few more months.

      Looking at border staff vaccinations as a extra will help. But it is just another few percent protection rather than a panacea.

      I would agree that inside the border has gotten really slack. Any case any outbreak inside will now need a wide set of lockdowns.

      • Incognito 2.1.1
        12 January 2021 at 12:11 am

        Yup, that was Prof. Nick Wilson.

        I posted a comment earlier today about the new variant becoming the dominant one globally: https://thestandard.org.nz/open-mike-11-01-2021/#comment-1774115.

        It does make sense to vaccinate frontline and border staff first.

        I agree that New Zealand has become complacent and is relying too much on the frontline defence, which will be breached again at some stage, unfortunately; it is simply a numbers game and therefore a matter of time, i.e. not “if” but “when”.

        When shit hits the fan, the blame games will start, again. Anyway, this is drifting away from your OP 😉

      • Treetop 2.1.2
        12 January 2021 at 2:24 pm

        Probably it was on one news last night the reason for not opening up more MIQ facilities is because there is not the staff to manage the isolation/quarantine.

        Not sure how popular to have a lock out before a lockdown for a limited period a month. This could be required as it is a not if but a when for a community out break.

    • gsays 2.2
      12 January 2021 at 7:17 am

      Thanks for the link, Incognito.

      Answered a few questions, but the answers largely fell into the 'We don't know yet' file. Not enough assurance for me to be pushing my way to the front of any queue.

      • Incognito 2.2.1
        12 January 2021 at 9:58 am

        Thank you for your response, which I find slightly puzzling.

        I assume you’re not keen to be the first to be vaccinated for some reason. Although our Government has ensured purchase & supply of vaccines, it is not rushing to roll out any vaccination program yet. In other words, NZ is not first in the queue.

        Why would you hesitate or delay? What “assurance” would change your mind, if I may ask?

        • gsays 2.2.1.1
          12 January 2021 at 1:05 pm

          Sure ask away.

          What would I be vaccinated for? Which strain of Covid? Do I run the risks of passing that on to Mum? Plus I have a healthy disrespect for Big Pharma.

          A bit like the annual flu jab, you get innoculated for last year's illness.

          The massive rush that this medication has arrived with makes the inner cynic in me wary.

          Another thread to the thinking is the false sense of security society has about the vaccine. Once we get the jab, everything will be alright, back to BAU.

          One of the positive sides to this pandemic is folk learning halfway decent hand hygiene. Coming out of a lifetime of kitchen work it's great that others have caught up.

          Edit, without telling Nana how to suck eggs, perhaps this is better on OM, we saw how quickly Mike Smith’s post on Julian Assange turned sour quick…

          • lprent 2.2.1.1.1
            12 January 2021 at 1:46 pm

            I’ll answer a few while I think about timers.

            Which strain of Covid?

            As far as I am aware most of the current released vaccines are targeting a particular attachment point (the bushy bits in a covid-19 virus) that it uses to attach to cells.

            This current strains have vary limited variations in that area because that particular bit was how they managed to species jump to humans. So current vaccines should trigger the immune system to recognise that. So they should all work with varying levels of effectiveness.

            A bit like the annual flu jab, you get innoculated for last year’s illness.

            It isn’t like the 1700 odd base pairs in a influenza variant that have virtually no error checking. This is a 32000 base pair virus with a lot of error checking. It mutates or recombines at a much lower rate. The only reason that we’re seeing any significant strains is because there were ancestral strains present at the outbreak in Wuhan, and then it had a very large number of hosts to play recombination games in.

            Different type of disease.

            With covid vaccines, the question is going to be different about effectiveness. It won’t be about coping with this years model like influenza. It will be how long the immune responses persist in humans. We won’t know that for years because we have to wait until the immune response stops either induced from vaccines or from ‘natural’ immunity after having it.

            But Covid-19’s close cousin SARS had immune responses that have persisted for up to 15 years. On the other hand Covid common cold immunities seem to last for only a few years. The difference appears to be to do with how much of a immune response was triggered in the first place. The people getting SARS tended to get very sick very fast, which was why it was so easy to stamp out. Getting covid common cold just makes people a bit miserable.

            So covid-19 vaccines need to trigger a strong response without making people sick. We won’t know how effective the first is for a while. The vaccine testing has mostly focused on how to make sure people didn’t get too sick.

            The massive rush that this medication has arrived with makes the inner cynic in me wary.

            It is one of the first diseases that came after we understood viruses, know how to produce vaccines for them (something that only really only started in the 1950s), and had a massive widespread economic impact that caused money to be thrown at it. It isn’t a stealthy retrovirus like HIV (which we still don’t have a vaccine for). It also helps when producing vaccines to have large numbers of test subjects – it abbreviates the testing process.

            But many of the vaccines have been going through the usual semi-public government testing. I don’t have a problem with that. There will be unforeseen effects as it gets rolled out to wider populations because people aren’t exactly the same. There will be a lot of data collecting – same as for every vaccine.

            Because we have the luxury of moderately effective border control – we can wait. But it is all about risk management. The first group that needs it are those at greatest risk of spreading – our border people and medical staff. After that the elderly who die from this too fast. etc etc…

            Edit, without telling Nana how to suck eggs, perhaps this is better on OM, we saw how quickly Mike Smith’s post on Julian Assange turned sour quick…

            I have a vaccine against people acting like non-listening arseholes. I just release my inner arrogant sarcastic arsehole, and then turn an amplifier up on it. This is one of my posts.

            • Treetop 2.2.1.1.1.1
              12 January 2021 at 2:39 pm

              I have never had an issue with vaccination. When it comes to vaccination for Covid I would want to know what the side affects are for rare auto immune conditions where several per million are diagnosed with the condition annually. There would be no data on this as it would take years to accumulate.

              • lprent …
                13 January 2021 at 8:04 am

                I would agree. We aren’t going to have as much information on these vaccines as we would normally have.

                If someone has conditions that could be an issue, then there has to be a tradeoff on risk.

                But that is always the case with vaccines or medicines. This one is just collapsed into a shorter time frame.

                The question is going to be if you take a vaccine with a possible unknown risk against any particular medical conditions. Or get a possibky fatal disease that may have long term consequences. Or constrain your life to minimize risk of infection. There are no absolute certainties.

                About the only certainty is that after the distribution of vaccines gets widespread, at some point the government and other citizens will start reducing the barriers to infection.

            • gsays 2.2.1.1.1.2
              12 January 2021 at 4:18 pm

              Thanks for yr considered response lprent.

              As you point out, here in NZ we have the luxury of time and distance.

          • Incognito 2.2.1.1.2
            12 January 2021 at 10:05 pm

            Ta

  3. Editractor 3
    12 January 2021 at 9:17 am

    An interesting article but it still leaves unanswered questions about the method of transmission, especially given the negative tests shown – from the last figure 26C is the luckiest I would say, assuming light blue is the original case (A).

    I am not sure " In particular the pattern in row 24 with the two people not getting a positive result is pretty abnormal." is necessarily correct though. If F and G were newly infected on the flight, they were themselves not likely to have been infectious during the flight. The last figure and the article text show that G caught coronavirus from their travel partner subsequent to the flight.

    It is possible that the people who were infected on the flight caught it by using the same toilet as the original case. I can well imagine that someone who is sick will relax their hygiene in an area where they can't be seen. If we discount cases B and G as having caught coronavirus on the flight, then 4 people did. From the seating plan, that's just under a 1 in 20 chance of catching it for those in the vicinity of case A, so how likely is it that there were a lot of infectious droplets in the air?

    • lprent 3.1
      12 January 2021 at 10:34 am

      I am not sure " In particular the pattern in row 24 with the two people not getting a positive result is pretty abnormal." is necessarily correct though.

      The main form of transmission for covid-19 appears to be from short distance airborne droplets, with a subsidiary and lesser transmission from some surfaces. This seems to be pretty clear with this flight – implies short distance droplets. The distance between infected and infectious is very low and entirely in aisles.

      If the passengers A and B were the infectious primary source, then they managed to infect C, D, E, and F in flight. Presumably F then infected G in quarantine.

      But as the seating diagram indicates, there are 9 people (including G) who were within similar droplet distances who didn't get directly infected during the flight according to subsequent tests.

      Since 4 did, that means a 4/13 probability of being infected at close range.

      The probability looks very high in aisles – 4/7.

      I presume that your 1/20 ratio is based on 4/87 – ie all passengers less A and B. But there has been little to no evidence anywhere of longer range infection past about 3 metres except for some rare cases of contact or enclosed spaces that don't clear the air – like lifts. If the environmental systems were a problem – then you'd expect a much wider spread issue.

      Presumably surface contact or residual droplets in the toilet would have had a similar probability for a lot more passengers than those we have a seating plan for. A lot of a passengers will use toilets in a plane – they always seem to have a queue on long haul.

      Pity that they obviously didn't have camera footage. I suspect that the pattern and frequency of contacts would have been enlightening. Like when and which toilets were accessed and who was seated and masked at the time. But my bet is that getting up and moving to toilets or stretching would be the theory to disprove. Droplets travel further when emitted at height.

      • barry 3.1.1
        12 January 2021 at 4:42 pm

        The main form of transmission may be short distance airborne droplets, but that doesn't rule out fomite transmission. I think that in this case fomite transmission (probably from the toilet) is the most likely cause. With the plane mostly empty there would have been a smaller intersection of possible infection candidates.

        • Incognito 3.1.1.1
          12 January 2021 at 9:55 pm

          How would the virus be internalised then? It doesn’t go through the skin so it could only happen by touching mouth, nose, or eyes with contaminated hands/fingers. I guess it is entirely possible that the ‘lucky’ ones had much better habits in the sense of not touching their faces.

  4. Patricia Bremner 4
    12 January 2021 at 9:21 am

    Yes travel to see our son… a long way off for all the reasons raised. There are still more questions than answers.

    I do think we should be praising and paying a premium to those doing the job of keeping us safe. I read many criticisms, yet our managed isolation is working.

    We need to train more staff, give real down time and rest for those charged with the task, so fatigue is not a cause of mismanagement, also a bonus could be paid if an isolation facility is rated safe and secure each calendar month. Carrot and stick.

    Perhaps in the end a certificate of vaccination/inoculation to travel will be needed, as some diseases have been managed previously.

    Let us hope we manage this infection as Queensland did, quickly reacting to an incursion of the new variant, until vaccination and herd immunity is achieved, which is some way off yet.
    That also assumes these vaccines will control the infection? That is the $64000 ? Only time will tell…so waiting…..

    • lprent 4.1
      12 January 2021 at 2:00 pm

      Only time will tell…so waiting…..

      Yep. Some of the botched distribution in the US and probably in the UK should be public enough to shed light on how not to do it. The US hopefully will soon lose theire major bullshit factor.

      Europe should provide a better controlled test for how to do it better.

      China has too few cases at present to get any good idea even if they didn't do the idiotic propaganda that they seem to be focused on.

      When India gets underway, that should be pretty clear. Some of the biggest vaccine manufacturing gets done there. The shortened distance between provider and recipients has proved effective in previous vaccine rollouts fro correcting problems.

      Russia will be interesting. While they have a wide public health system, it doesn't seem to be even and appears to have political aspects that aren't usually helpful at being effective. The early phases of covid there were perfect, all ok , and then a Brazil or US like high.