LSHTM Coronavirus & Health Survey

A total of 9741 people completed the survey between Tuesday 7th April 2020 and Thursday 7th May 2020.

Age

The chart below shows the age distribution of participants in our study. People aged 13 to over 80 years old have taken part in the study. We are especially keen to increase the number of participants in the under 40 and people aged 75 and above.

Ethnicity

The ethnic backgrounds of the people responding to our survey are shown in the table below, along with the national average in the UK. At present, we have an over-representation of White people, and an under-representation of Asian and Black participants. We are very keen to get more people from non-white backgrounds to participate in our study

Gender

Females significantly outnumber all other gender types, with male respondents making up just 20.89% of the respondents. Our study is not directly investigating the risk factors of biological sex on COVID-19 and as a largely anthropological study has more focus on gender. Where relevant we have performed some analysis on the assumption that genders of ‘male’ and ‘female’ are synonymous with biological sex in most cases. Where people have indicated a non-male, non-female gender, we assume that their biological sex is unknown and have excluded them from relevant analysis.

Employment and Work

The majority of our participants either have a job or are retired, although we have some representation of people who either do not work, cannot work or are jobseekers.

If someone told us that they had either a full-time or part-time job, or that they were self-employed, we asked them what sector they worked in. Our participants came from a wide range of different sectors of the economy, but there appears to be a bias towards people working in education and healthcare.

Household Income

We expect that there might be some big differences in the way that different households manage during the COVID-19 pandemic. This may have something to do with how wealthy different households are. So far, our survey seems to have received a relatively even balance of participation from across the full range of household income levels.

Educational Qualifications

People who have a degree or higher qualification have so far been more likely to respond to the survey. Over time and once more respondents have taken part, we hope that there will be a more even spread of people with different educational qualification

Where people live

Participants have responded from from across the UK country but around 15.1% live in London. We have under-representation from Northern Ireland

Most participants of the survey live in houses with gardens, but a large proportion of people live in flats with no gardens. Given that lockdown puts limits on how often people can leave their properties, we are keen to investigate the importance of private gardens in people’s lives during the COVID-19 pandemic. Access to private outdoors spaces is likely to be a key feature of our planned analysis.

What people are doing at the moment

Almost everyone is following the social distancing recomendations at the present time with many saying they are isolating. Around three quarters of people are regularly exercising and most people are keeping in touch with neighbours and their community and spending time with loved ones.

A total of 1564 (16.1%) of participants believe that they have had COVID-19 but have now recovered.

A total of 157 (1.6%) of participants believe that they have currently got COVID-19

These are not confirmed cases, but reflect self-reported disease. It is unlikely that all of these possible cases were true cases of COVID-19, but it will be interesting to see whether the percentages increase as the outbreak progresses.

The percentages are fairly similar for males and females, although it looks like around two times as many women think that they have a current case of COVID-19. It also seems that more women think that they have had a previous case of COVID-19.

Participants who thought that they had recovered from COVID-19

From what data are currently available, it looks like older people are less likely to have said that they believe that they have had a case of COVID-19. There is currently no way to know if this reflects the true rate of disease in different groups or if it might be something to do with the differences between the groups. Our numbers in the 13-19 year age group are very small and all data in this group are probably unreliable as a result.

Age Distribution of possible COVID-19 Cases

Duration of Symptoms in possible COVID-19 Cases

Within the group of people who thought that they had previously had COVID-19, we asked participants to tell us the number of days during which they had had symptoms of the disease. A large percentage of people told us that their symptoms had lasted ten or more days.

Because COVID-19 carries higher risk in older age groups, we were interested in finding out if people with possible (not confirmed) cases of COVID-19 had symptoms that lasted longer if they were in particular age groups.

So far we have limited amounts of data, particularly in the youngest and oldest age groups, so little can be learned from this chart at this stage. If anything, we might be starting to see a gradual increase in average duration of symptoms as people get older.

When split by gender, it seems that both men and women reported symptoms lasting longer than 10 days in most cases, with around a third of men and nearly half of women doing so. More men (as a percentage) reported symptoms lasting between 3 and 5 days than women and a smaller percentage of men overall reported symptoms lasting 10 days or more. A robust statistical analysis would be required to make any conclusions about this issue and this will follow in a later update to this report.

Number of cases in affected households

In households where COVID-19 cases were reported, we were interested in how many people in the household thought that they had been infected. The chart shows the number of people who thought they had had COVID-19, split by household size. The grey boxes show the size of the household and the blue bars show what percentage of households that size reported having n number of cases in them (where n is a number smaller than the household size).

There appears to be some misreporting as a number of people who said that they believed that they had suffered COVID-19 reported zero cases in their household. Nonetheless it is clear from the data that COVID-19 symptoms do not develop in all members of affected households.

It seems unlikely that many people living in a household with an affected case could avoid any exposure to virus particles. As such we think that this may relate to the fact that lots of people with COVID-19 have an infection but no symptoms. It is also possible that some people are resistant to infection.

Smoking

Smoking is a risk factor for many diseases. In our data we had a similar percentage of both current and past-smokers in both the group of people who had not had COVID-19 and in the groups who thought they had current or previous COVID-19. Whilst this does not provide obvious evidence that smoking is a risk factor for getting symptoms of COVID-19, the chart also does not necessarily indicate that smoking is not a risk factor. The number of smokers in the data is rather small and we have not yet performed tests to show whether there is a statistically significant difference between the groups. Smoking may also impact on severity of symptoms and on outcome of severe disease, which we have not yet investigated.

If you were in charge, what would you do to control the outbreak?

We asked people what they would do if they were in charge of controlling the outbreak. TPeople could provide their own text and could say anything they liked. With so many answers it can be hard to rapidly understand patterns in the answers and we will be doing some in depth analysis of these answers.

To provide a rapid view of the kinds of things that people said, we generated word-clouds. These show the words that were most frequently mentioned, with more frequent words being shown in larger text.

How you choose to interpret the meaning of these individual words in wordclouds is entirely up to you, but some such as ‘testing’, ‘PPE’, ‘stricter’, ‘vulnerable’ and ‘earlier’ might give a good idea of the kinds of topics people are most concerned about at the present time

Have you heard anything about coronavirus or COVID-19 that you think is being kept secret, covered up or hidden from the general public?

We are very interested in learning about the misinformation and rumours that are circulating about COVID-19. In outbreaks of other diseases we have seen that rumours can have a direct negative impact on efforts to control disease, so this is an incredibly important topic.

Around 31.4% of participants had heard rumours about secrecy and cover-ups.

There was a statistically significant association between whether participants used Facebook as a source of information and whether they had heard rumours about things being kept secret, covered up or hidden from the general public. Around three quarters of people who reported hearing such rumours were people who used Facebook as one of their information sources.

## 
## |                 |                 |              |              |               |
## |----------------:|----------------:|-------------:|-------------:|--------------:|
## |                 | `Facebook User` |           no |          yes |         Total |
## | `Heard Rumours` |                 |              |              |               |
## |              no |                 | 2226 (39.9%) | 3351 (60.1%) | 5577 (100.0%) |
## |             yes |                 |  781 (29.1%) | 1899 (70.9%) | 2680 (100.0%) |
## |           Total |                 | 3007 (36.4%) | 5250 (63.6%) | 8257 (100.0%) |

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  facebook_rumours$`Heard Rumours` and facebook_rumours$`Facebook User`
## X-squared = 90.248, df = 1, p-value < 2.2e-16

As with the previous question, we allowed people to describe in detail what rumours they had heard and we will be investigating this in greater detail in due course. For now we provide another word-cloud. Words like ‘Testing’, ‘PPE, ’Deaths’ and ‘numbers’ feature in this wordcloud, along with ‘China’ and ‘Government’. The word ‘Conspiracy’ appears on this wordcloud, though perhaps surprisingly is not one of the larger, more frequent words.

Have you heard of people doing interesting, unusual or different things to fight coronavirus?

During our work on previous infectious disease outbreaks, we have heard of many interesting and unusual approaches to treatment and cure. Many or most of these do not have any proven effect and some can be very harmful. Only a small percentage (11.7%) of participants reported hearing of unusual or interesting control methods. We will wait until we have much more data before we make any attempts to analyse this data. A wordcloud is provided for interest.

Stay Home, Save Lives