Setting mental health science priorities for the COVID-19 pandemic

The OII's Prof. Andrew Przybylski is an author of a piece published today in The Lancet Psychiatry, on multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science. Based on survey data and with input from a panel of experts convened by the UK Academy of Medical Sciences and the mental health research charity MQ, it explores the psychological, social and neuroscience impacts of COVID-19, and sets out immediate priorities and longer-term strategies for mental health science research. 

We caught up with Andy, to discuss the mental health implications of the COVID-19 pandemic, and how it bears on his own research interests in digital technologies and wellbeing.  

David:

The paper has a long and impressive list of authors, including neuroscientists, psychologists, psychiatrists, and population health experts. What perspectives did you bring yourself, to the question of current research priorities? 

Andy:   

I did my best to bring my experience working on ways of improving the suitability of scientific evidence generated by psychologists, and other quantitative sciences, for policymaking to bear in the document. You’re right that there are a lot of moving parts when a diverse group of experts tries to set a broad national agenda. Many of the members of the advisory group are distinguished in areas I have only dabbled or occasionally published in. My experience thinking and applying models for making data more transparent, studies more incremental and informative and research efforts more scalable and less wasteful was where I could make a contribution.    

David:

We already know that those with chronic illness can suffer from anxiety, and that the elderly are often lonely -- could the fact that we are *all* now potentially affected (e.g. by anxiety and loneliness) actually help focus more attention on the importance of mental health, generally? Could the pandemic bump it up the public health agenda?  

Andy: 

I really hope so. You can think of there being at least three groups that receive different levels of attention and resourcing from policymakers and agenda setters. There are the relatively well off, those at risk, and those in crisis. All too often the latter two categories are missed from serious investment. If this crisis gets us to think more critically and systematically about our mental health infrastructure, then that’s a very good thing. The primary challenge for the latter two groups has consistently been resourcing. Ideas will arise amongst the relatively well off, such as ‘is social media making teenage girls depressed?’. We’ll debate things like this ad nauseum but the studies are poorly done and then the follow up, in terms of resourcing is entirely absent. If we took the mental health crisis among young people seriously, whether it was related to social media or not, you would see material movement in terms of making sure that young people had timely access to clinical support. This was lacking in ordinary times and I would hope that this crisis highlights this disconnect.   

David: 

I guess the pandemic provides an unexpected natural experiment for many areas (air quality, carbon emissions, the economy, etc.) -- including mental health. Presumably there are a lot of basic questions we could be asking about how we function (or don't..) as social animals, under these unprecedented conditions of social isolation. How might the Internet fit into this question -- in a way, we've never been better connected? 

Andy:  

There are a lot of opportunities out there for good basic and applied research, but I have found many of the projects now underway that have cast themselves as addressing COVID-19 are either wasteful, poorly considered, or peripherally related. It’s easiest for me to see this in my fellow psychologists.  For example, in the field of psychology, between February 10, 2020 and April 12, 2020, at least 74 new manuscripts were written about COVID-19 and shared to the field’s primary preprint server, PsyArXiv. It strains credulity that this work meets reliable standards for evidence. That said, I strongly suspect that there is a broader challenge to the credibility of the quantitative social sciences.  

I believe social, psychological, and data scientists we need consolidate our research efforts and run consolidated large-scale studies. This means open data, open measures, and pooled efforts that don’t waste the goodwill of participants or resources of funders. My dissatisfaction is part of the reason I joined the AMS/MQ effort here to set a proactive agenda for what internet-based research should look like, and how open science fits in to learning as much as we can from this crisis so that we can be better prepared to act as it and future crises unfold.  

David:

Of course, good research relies on quality data. In terms of Internet-related data, what could we be doing in this area? What's out there? And what still needs to be collected? 

Andy:  

I think there is a great deal that social media and online gaming platforms can do. It is somewhat ironic that we are living in an extremely data rich age but that most of this data is held by a relatively limited number of private companies. I think that well-coordinated collaborations between these platforms and teams of scientists with focused agendas are needed to make the most of these resources.  

David:

I actually found the "call to action" a bit overwhelming, given it's so amazingly comprehensive, and ambitious. There are probably all sorts of studies being launched all over the place at the moment: what's the importance of setting these broad research agendas? And how do they actually get (hopefully) put into practice?  

Andy:  

Understanding I where I am starting from, and these are the parts of the paper I wrote, I think that scientific transparency and empirical rigorour are the keystones of an effective response to this and future crises. Registered reports, a method of scientific review whereby peer reviewers evaluate and tweak study protocols before studies are conducted is a new and essential tool for ensuring we have high quality and actionable evidence. Chris Chambers has created a rapid workflow now adopted by a wide range of STEM fields that has streamlined the review process to take about a week. A scientist who thinks they have something to say about COVID-19 has no excuse to not go through this process. Hypotheses and materials are reviewed as are the statistical models which will be used before the data are collected. Further the data and the peer reviews are made publicly available so that those disagreeing with methodology, or looking to improve upon it, can do so easily and quickly.   

David:

The paper notes that (as of 1 April), there were 643 COVID-19 studies found on MEDLINE and Embase, only 10 of which were on the mental health impacts. Are you surprised? Given the mental health impacts might be evident for years, the number of studies will presumably rise -- how do we ensure they form a useful, usable body of work? What do we need, to usefully inform public policy? 

Andy:  

I wonder about this. My sense is that the immediate impacts will be felt by specific groups who are directly confronted with the crisis such as medical workers and key workers, so I am deeply concerned but I am not surprised. I think that mental health is difficult to ethically study and that many of the society-scale effects will take shape over time. It will be a slow burn. That said, we need to keep mental health at the centre of the national health agenda because it is both simultaneously difficult to understand, expensive to study, and critically important to our survival as a culture.  

David:

And finally, as a researcher, what are you most struck by in terms of the current situation? And how might it inform your own research agenda?  

Andy:  

I’m not sure, I’m more disappointed than I expected to be. I think a lot of researchers have good intentions to respond to this crisis, but I fear many are placing undue faith in existing science and methods which are known to be unreliable. There is a kind of optimism or exuberance for collecting data on COVID-19 that I find personally unseemly. I’ve lost count of the number of studies that have been launched in the last month on social media, mental health, and loneliness. People have asked me what’s the best measure of technology use a couple dozen times in the last few weeks and without exception I’ve been met with blank stares when I ask what their research questions are. This is not a good sign.  

In terms of my own research agenda, I don’t know. I’m focusing my efforts mainly on applied work, trying to connect people and groups to make a difference. This crisis has deepened my personal and professional dissatisfaction with the standards that many of my peers apply when they pursue research so I’m doing my best to address that directly by contributing to efforts the AMS/MQ initiative and writing a bit for broader audiences. I think it’s critically important that we link basic and applied research to real-world interventions that will generate true impact. 

If research on COVID-19 is as urgent as many scientists are claiming, it is urgent that we don’t cut corners and get the science right.   

Read the full paper: Holmes, E.A., et al. (2020) Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science. The Lancet Psychiatry. 

Dr Andrew Przybylski is an Associate Professor at the Oxford Internet Institute, University of Oxford, where he investigates the effects of digital technologies like video games and social media on wellbeing. Follow: @ShuhBillSkee. 

Andy was talking to David Sutcliffe, OII Science Writer.