Grappling with a novel virus that reared its ugly head barely six months ago, the world is facing many uncertainties. The SARS-CoV-2 virus is proving unpredictable and the pandemic is fast-moving. But one thing we do know is that older people bear the brunt of the impacts of COVID-19. The elderly are disproportionately affected, with those over 65 accounting for some 80% of hospitalisations due to the disease. And one in five over-80s with COVID-19 will need to go to hospital, compared with one in 100 individuals under 30.
So it’s worrying that new research based on global data IGHI has been gathering shows that elderly people aren’t behaving more cautiously than their younger counterparts.
Working with YouGov, IGHI has been tracking people’s behaviours and attitudes during COVID-19 across 29 countries, through bi-weekly surveys that are collecting nationally representative data. The motive behind this major piece of work has been to open the eyes of decision-makers so that they can understand how their citizens are responding to the crisis, and tailor their strategies based on evidence.
These data have revealed some “troubling” findings, according to Dr Jean-François Daoust, Lecturer at the University of Edinburgh. His published analysis has revealed that elderly people – those most vulnerable to COVID-19 – are no more likely to comply with public health measures than other age groups. In some cases, such as mask-wearing, older people are actually far less likely to engage in preventive behaviours.
This is surprising and concerning. How can governments take the necessary action to protect older people based on this knowledge? Lead researcher behind the project for IGHI, doctoral researcher Sarah P Jones, spoke with Dr Daoust to take a deep dive into the implications of these findings. Read below for more insights from Dr Daoust, who is calling for targeted public health policies in response.
What are the key global policy implications of your findings?
“The background of the pandemic is clear: several groups of people are at greater risk, hence we need to protect them to a greater extent. Among those at risk, elderly people or older people in general were a top priority. They are the group of people with the greatest – and by far – death rates. My findings show that elderly people are not substantially more willing to self-isolate, even though we know that this is the best preventive measure in the absence of a vaccine. Beyond self-isolation, they do not comply to a greater extent than their younger fellow citizens to public health preventives measures, which can range from washing hands to wearing a mask when outside the home.
“The implication is straightforward: our governments were not successful in conveying clear and convincing messages to that population and a lot of them died. There need to be better strategies implemented than the ones at the time of my study if we want to minimise the number of death from the pandemic.”
How can we ensure that policy-makers respond to these findings?
“We can evaluate case by case the public health policies that are targeting elderly people but, more broadly speaking, we should keep doing research to evaluate the effectiveness of governments’ strategies on the general population. The more effective for everyone, the less likely elderly people – like everyone in society – will put their and other peoples’ lives in danger. On top of this, we can evaluate particular measures for older people.”
Although the data did not cover the ‘why’, what are your thoughts on possible reasons behind the unexpected trends we are seeing in older people during COVID-19?
“These are just food for thoughts. It could be that they have far fewer social interactions when they are self-isolated at home, compared to other, younger people who would also still be self-isolated at home, but mastering technology to keep a minimum of social contacts with their friends and family. The lack of social contact could therefore drive older people to seek interactions outside the home because they do not master the technological tools to do so from their home.
“This certainly deserves a great deal of attention and while I have started to work on it, I hope that my research will spark interest on this to many researchers.”
We found ourselves willing to share data between researchers who might otherwise find themselves in competitive situations. Has this experience changed the way you think about academic collaboration especially during health emergencies?
“I have never felt that I was in competition. But when outcomes from findings can negatively affect people, I think that we should fully share the data and the codes, exactly the way that this IGHI/YouGov collaboration has for the datasets that I used.
“There can be reasons for not sharing a full dataset, but the minimum is to share what you are using in your publications and, ideally, everything you can if it does not prevent you from using the dataset as you expected. While this minimum was not reached in pre-pandemic times, it should be that way all the time.
“There are benefits for everyone in doing this: you are appreciated for sharing the data and codes by people who will use them, transparency, productivity, etc. The pandemic made us realise this because the stakes really increased in a very short amount of time. To allow greater research outputs, but also to ensure that they were replicable, more people began to adopt that view. Some people were already there in pre-pandemic times, and I think that others will realise that it is not a big cost to clean your replication files and make them publicly accessible, and that adding a dataset on a public dataverse does not prevent you from using it again for other research.”
We’ve chosen to follow behavioural responses. How important are behavioural insights to improving and evolving government response to this pandemic?
“This is crucial. We neglect it, but there is a fundamental descriptive job to do and to keep updated. At the end of the day, we can run fancy models on things such as compliance with preventive measures but we first and foremost need a good description and understanding of that variable. If we can’t make sense of a careful description of one variable or a simple bivariate relationship, and I tried to do the latter in my article at PloS ONE, how could fancy, complicated models deepen our understanding of our world?”
What are the benefits of having a cross country comparison for the impact of COVID-19 on older people?
“Case studies can be very useful, but we also need a comparative framework for at least two things I would say. First, you can test whether the relationships that you find can be generalised across countries. Second, if they are not generalisable, then you can benefit from more variance on the explanatory, contextual factors, and examine possible explanations on why some things work some countries, while others don’t.”
If you could add one question to the survey, what would that be?
“Actually, I would prefer to play and implement survey experiments to the actual questions. For example, we could randomly split samples where one group receives the same question as in the last survey, and another one receives a ‘face-saving question’. This allows respondents to admit non-compliance with a guilt-free answer choice, as I explained in another publication published in the Journal of Experimental Political Science recently (see my publication focusing on this).”
If you’d like to delve deeper into our global survey data on behaviours and attitudes during COVID-19, you can visit our publicly available dashboard here. Like Dr Daoust, researchers can also access the anonymised data via GitHub.