Categories
Covid-19

Some thoughts from earlier on in the pandemic

I am posting some of my earlier newsletters about the virus. Some of the questions have now been answered. On some issues my thoughts are different now., whereas on others our knowledge remains unchanged. It is interesting to look back

Categories
Covid-19 Mortality

Mortality from CoVid amongst Jews

[First circulated 29th June 2020]

Widely reported in both the Jewish and non-Jewish press was the high mortality from Covid amongst Jews in England and Wales.  This brief paper summarises the data, what is known and what is not known.

The study data : 

The Office for National Statistics in England and Wales linked the names on the death certificates that mentioned CoVid occurring between 2nd March and 15th May to those individuals’ self-reported religion on the 2011 census

The denominator was the total who self-reported one of the nine different religious groupings (including no religion) in that census

For each person who died the research also extracted from the census data:

  • Age
  • Gender
  • Self-reported ethnicity-ie white/non-white (not relevant for Jews but is important in comparing with data from Muslims 
  • Various indices of socio-economic deprivation
  • Self-reported health 

Comment: The number of Jews from the 2011 census amounts to 0.5% of the population.  Obviously like any estimate it is subject to errors. 

 Analysis

The research calculated the numbers of deaths per 100,000 in each religious group, separately by gender, overall and divided into those above and below 65.

They then calculated a full age-adjusted analysis to  compare these rates with the rate in Christians and other religions (including those who declared no religion).

Comment: The analysis was a standard epidemiological analysis.  One concern might be that using Christian as the comparator might be based on a very selective part of  the population.  In fact, 60% of the England and Wales population declared themselves as Christian in 2011 compared to just 7% who did not state their religion

The results 

There were a total of 453 deaths in people who were identified as Jewish on the 2011 census, 23 were aged 64 and under  and 430 above that age.  268 were males and 185 were female.  

The headline rates of deaths per 100,000, after adjusting for the age distribution of the population were  as below:

 Males: Females: 
Jews187.798.2
Muslims198.994.3
Hindus154.893.3
Christians92.654.6

* Note that the death rates for those with no religion or not prepared to say were similar  to the Christian population

  • Restricting the data to just the population aged over 64, this was the ranked order of deaths, taking account of the age structure, by religion
Ordered by highest death rateMales: Females: 
1stJewsJews
2ndMuslimMuslim
3rdHinduHindu
4thSikhSikh
  • Restricting the data to just the population aged 64 and under this was the ranked order, taking account of the age structure of deaths by religion
Ordered by highest death rateMales: Females: 
1stMuslimMuslim
2ndHinduHindu
3rdSikhSikh
4thJewishChristian

 * Note that the number of female deaths under the age of 65 was too small for some religions (including Jewish women and thus age-adjusted rates were not calculated  

  • After allowing for socio-economic and other indices of deprivation and self-reported ill health at the 2011 census this was the ranked order:
Ordered by highest death rate Males: Females: 
1stMuslimMuslim
2ndJewishHindu
3rdHinduJewish
4thSikhSikh

Comment: There is no doubt that Jews have a much higher death rate from CoVid than the background Christian population and indeed were similar to, or higher than, other religious groups This analysis takes into account that some of the differences in death rates may reflect differences in factors such as poverty and overcrowding.    Indeed, the death rate in Jewish males was still approximately twice that in Christians.  Although British Jews as a group may be better off economically, there are those in the growing Charedi community who are poorer and more overcrowded.  This analysis takes account of these differences 

The above analysis was repeated after allowing for self-reported ethnicity ie white/non-white.  This was the result:

Ordered by highest death rate Males: Females: 
1stJewishHindu
2ndMuslimJewish
3rdHinduMuslim
4thSikhChristian

Comment: This in some ways is the most interesting result.  Whilst Jews virtually always describe themselves as being of  ‘white ethnicity’ this is not true for Muslims.  (Detailed figures are available). After allowing for the ethnicity question, the effect of Muslim religion is much diminished.  The interpretation is that it is not the Muslim religion that is associated with the higher death rates but the ethnic aspect (ie white Muslims are not at risk of higher death rates)

How should the results be interpreted?

These are death rates and reflect the combination both  of how common CoVid is (the ‘incidence’ rate’)  in the different religions and how likely sufferers will die (the ‘case fatality’) in those who contract the virus. 

Is the higher death rate in Jews due to a higher incidence, a higher case fatality (or a mixture of both)?

Of these two possibilities the second is the less likely.  It is known that in the UK, Jewish males have death rates about 40% lower from all causes than the background population with females about 20-30% lower. Thus, British  Jews live longer with, for example, approximately twice as many being over 85 as the background population.  Also, though  not comparing like with like, Israel has one of the highest life expectancies in the world- being similar to Switzerland and Singapore

Thus, the more likely conclusion must be that there is (or was) a higher incidence of CoVid in Jews, rather than a higher case fatality 

It could be debated as to whether this was due to (i) being more likely to be exposed to the virus because of increased likelihood of social contact with an infected individual, and/or (ii) once exposed being more likely to get an actual infection with the same level of contacts.  Theoretically the latter is possible: ie  that Jews are less likely to remain asymptomatic for example are less likely to have pre-existing immunity. There are no data on this.

But one intriguing unanswered question!

In all the analyses on deaths, Jewish males have twice the age adjusted death rate as Jewish females.  That there is an excess death rate from CoVid in UK males  is well described and publicised but the data on Jews are different.  Indeed, Jewish males have around 2.2 times the death rate  of Christian males, whereas Jewish females have only about 1.5 times the death rate of Christian females.  Is it as simple as Jewish men spending more time mixing with each other??  

Categories
Covid-19

Can CoVid 19 spread through the air?

Last week 237 scientists from various disciplines signed a letter to a journal suggesting that infection with CoVid 19 could be from airborne transmission and not just by droplets. They called on the World Health Organisation to acknowledge this and advise accordingly.  In this blog I explain what this means and summarise the arguments for and against their suggestion.   I then consider what are the practical consequences if these scientists are right.  

What is the difference between droplet and airborne transmission?

In simple terms, we breathe out drops of different sizes.  When we cough, or sneeze, (even talk loudly or sing!) we breathe out large droplets.  These  can land on surfaces, or if we are close enough land on someone’s eyes, nose or mouths.  If these droplets carry the virus then this can lead to infection.  Because droplets can only travel for a short distance, by keeping a safe distance, thought to be around 2 metres, and especially washing hands in case they have touched surfaces on which there are active droplets, most carriage of infection could be stopped.  Indeed, that was the advice at the beginning of the pandemic.  In general, this has also been the advice for years to avoid catching infections such as flu.

The scientific question is  based on the fact that even when we breathe out normally, we breathe out smaller droplets that hang around in the air for longer times and longer distances.  These smaller droplets are referred to as aerosols (like the air freshener sprays we use at home).  The scientists who signed the letter suggested it is possible that the CoVid virus can also spread in these smaller drops and hence be able to lead to infection at larger distances than 2 metres.  This is illustrated below:  Source: (https://www.eurekalert.org/pub_releases/2020-04/quot-ipe041520.php

There are some key questions:

1. If someone is carrying the virus can they spread the virus via aerosols as well as by droplets?

There is only limited research on this and the data have not been subject to peer review (ie the quality of the research judged by experts to assess its quality). In one research study, Chinese scientists studied the breath of patients with CoVid 19 and looked for evidence of the virus in air samples as well as in their actually exhaled breath. What they found was that there were viral traces in 17% of samples of directly exhaled air (ie the ‘large drops’ in the picture above) but also in 4% of samples of ‘wider’ air (the ‘small drops’). This was a small study but the conclusion was there are traces of the virus in the air from the breaths of people with CoVid

2. Does this type of study prove that the exhaled air contains active virus?

This is a source of dispute between scientists. The studies such as the one above only show that there is viral genetic material (RNA) and cannot prove the actual presence of the ‘whole’ virus. RNA for example may be found in swabs of people who are recovering from infection but that does not mean they are infectious. The problem is that whilst it is easy to investigate for the presence of RNA, it is very difficult to prove the presence of the whole virus.

3. How far can the aerosols spread?

Forgetting CoVid for the moment, there is a general question which is how far aerosols can spread, and the short answer is we don’t know. and depends on so many factors including how they were produced ranging from normal breathing to speaking, sneezing and coughing. Droplet size is also important: larger droplets will fall to the ground closer to the person than smaller droplets. Droplets in aerosols can certainly travel more than 2 metres before dropping

4. How long would the virus last for in the aerosol?

The next question is in a person who has CoVid and is breathing out small droplets in an aerosol how long would a virus survive? This depends on where you are. In laboratory studies, 90% of viruses cannot survive more than 6 minutes outside in sunlight. Contrast that with being inside a poorly ventilated room with no UV light, and the virus within aerosol drops could survive for two hours

5. Would you receive  a big enough dose of the virus to actually get infected?

Thus far much of the above evidence is circumstantial. It is known that we produce aerosols when we breath out normally, and that virus traces can be found in exhaled breath. What is not known is how much virus is needed to trigger an infection. To give an idea of magnitude, a droplet of one ml can contain millions of virus particles and one minute of loud speech could generate 1000 virus-containing droplets, which can remain airborne for 8 minutes. There is also the concern that unlike even other coronaviruses eg SARS, in the laboratory aerosols with CoVid19 can remain infectious for 6 hours. It is important to remember that studying laboratory-based aerosols may not replicate what happens in real life

6. Is there any real life evidence that aerosols can cause CoVid at a greater safe distance than we were led to believe?

At the moment it is anecdotal and based on records of cases occurring in people who were in contact but not close contact. There was an interesting study of people who caught the infection in a restaurant. They proposed that those who did, and who did not, contract the infection could be explained by air conditioning flows rather than their proximity to each other. A now infamous choir practice in Seattle emerged with 52 members being infected. Although it is possible that one affected member spread the virus to all the others by direct droplet, an alternative view is that this person was exhaling aerosols throughout the evening and the high concentration of virus in the exhaled air lead to the high rate of infections

This all sounds quite scary how worried should we be?

Obviously more people would be exposed to the aerosol from an infected person than to their large droplets. Further, even before lockdown measures, an infected person typically spread the virus to less than 3 others (that is, the  R value was under 3).  Given how many people we are in contact with, if aerosol exposure was important then we might have expected a much higher transmission rate. 

In addition, the risk of infection, even by airborne route, depends on the prevalence  of the infection in the population.  As the number of new cases is decreasing, so does the likelihood of coming into contact with someone who is carrying the virus. 

What are the conclusions:

Like so much else with this pandemic despite our desire for hard facts there is much we don’t know.

  1. My conclusion is that whilst the large majority of infections occur from our contact with large droplets, the scientists are right to be cautious in not excluding the theoretical risk of aerosol transmission, even from normal activities in an asymptomatic person.
  2. Statistically the chance of becoming infected in this way is small, and especially small in areas where there have been no recent cases
  3. Outside and in well ventilated spaces and especially where there is UV light, social distancing should offer the necessary level of protection.
  4. SAGE’s advice that when indoors, especially in poorly ventilated places, people should wear face masks, avoid face to face interactions, and loud talking or singing seems sensible  

Interested in further reading:

“Mounting evidence suggests coronavirus is airborne-but health advice has not caught up”.  Nature, 8 July 2020

“How coronavirus spreads through the air: what we know so far” Scientific American, 12 May 2020

First10EM: http://first10em.com/aerosols-droplets-and-airborne-spread/

Categories
Covid-19

Trying to answer friends’ questions!

From the beginning of the Covid epidemic we have been inundated with data and there has never been so much air time or column inches given over to epidemiologists! Yet friends have been confused trying to make sense of it all both for individuals and understanding what’s going on.