Introduction: The immediate impact of COVID-19 on morbidity and mortality has raised the need for accurate and real time data monitoring and communication. The aim of this study is to document initial observations from multiple digital services providers during the COVID-19 crisis, especially those related to mental health and wellbeing.
Methods: We used email and social media to announce an urgent call for support. Digital mental health service providers (N=45), financial services providers (N=4) and other relevant digital data source providers (N=3) responded with quantitative and/or qualitative data insights. People with lived experience of distress, as service users/consumers, and carers are included as co-authors.
Results: This study provides proof-of-concept of the viability for researchers and private companies to work collaboratively towards a common good. Digital services providers reported a diverse range of mental health concerns. A recurring observation is that demand for digital mental health support has risen, and that the nature of this demand has also changed since COVID-19, with an apparent increased presentation of anxiety and loneliness.
Conclusion: Following this study, we will continue to work with providers in more in-depth ways to capture follow-up insights at regular time points. We will also on-board new providers to address data representativeness. Looking ahead, we anticipate the need for a rigorous process to interpret insights from an even wider variety of sources in order to monitor and respond to mental health needs.
Early warning signs of a mental health tsunami: A coordinated response to gather initial data insights from digital services providers
During the COVID-19 pandemic, traditional mental health services and related activities have declined, in part, due to outpatient clinics being closed to adhere to social distancing requirements, mental health staff redeployment, and inpatient beds being converted into COVID-19 units. As governments attempt to contain the virus, we must mitigate the mental health impact of the pandemic and economic crisis, especially given that pre-COVID-19 predictions already indicated that by 2030 depression will be the leading cause of disease burden globally .
During the SARS (2002-2004) epidemic, social disengagement, mental stress, and anxiety were associated with increased suicide rates in the elderly population . Another study found that 30% of children, and 25% of parents, who were quarantined or isolated during pandemic diseases met the clinical criteria for post-traumatic stress disorder . Data from previous economic depressions and recessions suggest profound increases in substance use disorder, depression, and suicide [4,5].
In the current pandemic, frontline health care workers face the possibility of anxiety and burnout [6,7], alongside fears of becoming ill, especially among ethnic minority groups . For others living in highly conflicted households, social distancing has meant prolonged social contact, and abuse. For example, in the UK the number of suspected domestic homicide victims more than doubled during the first three weeks of the lockdown . In France, calls to the national violence against children helpline have increased by 89% . From an economic perspective, a survey of UK households three weeks into ‘lockdown’ found that 49% of households feel anxious about their finances, rising to 95% amongst the households experiencing serious financial difficulties . A survey conducted in March, just as lockdown rules were coming into place in the USA, also highlighted higher levels of psychological distress among lower income households .
There is a need to obtain more granular and real-time information to help us understand the nature and scale of the mental health crisis. A possible source of this information is the large number of digital mental health services providers used by millions of people globally. These include patient to clinician communication tools, digitally enabled treatments, self-managed care solutions, mental health and wellbeing apps, online forums, support networks and digital communities. In addition to this, given the established links between health, social, and economic factors [e.g., 13], insights should also be obtained from financial services providers, and other relevant digital data sources.
In this study, we set out to collect observations from multiple digital services providers (Table 1). To our knowledge, this has never been done at scale before, and we did not know how many providers would respond, or what the nature of their data insights might be. With rapid turnaround, a diverse range of providers came forward with collective information sourced from a user base of at least 10 million people, but possibly reaching upwards of 50 million. The information is summarised in this paper and detailed in Tables 2 and 3.
We used email and social media to announce an urgent call for support to investigate the scale and nature of the mental health impact of COVID-19 . Starting 6 April 2020, BI sent emails to all speakers who had presented at previous ‘Digital Innovation in Mental Health’ conferences , to members of the FinHealthTech Consortium , and then to a much wider digital community via LinkedIn, Twitter, and Facebook.
Respondents were asked to provide qualitative and/or quantitative insights with no exchange of data or identifiable information. We did not provide a framework for insights or any analytic specifications. Therefore, all insights should be considered illustrative examples, not primary research.
We asked providers to be compliant with General Data Protection Regulation (GDPR) and Data Protection Act 2018 if their users were within Europe. To set a good example of responsible innovation, this document only accepted data from providers with clear and accessible privacy policies.
Data insights and draft versions of the paper were shared amongst all co-authors for feedback, including from people with a range of lived experiences of distress and service use.
Given the urgency in sharing initial provider insights, we are not able to draw conclusions from the content provided in Tables 2-3. Instead, we summarise some of the more frequent observations reported by providers.
Insights suggest changes in the type of information individuals are seeking or presenting. From Google Trends data, searches for ‘anxiety symptoms’ doubled between the weeks beginning 8 March and 22 March 2020. In a similar timeframe, Mental Health America (MHA) witnessed a 22% increase in numbers of GAD7 anxiety screens (N=11,033) taken in March 2020 compared to February 2020. Qualitative insights suggest that individuals are seeking practical resources and coping strategies. Themes emerging from It’s Ok To Talk deal with worry and anxiety, strategies to manage work, studies, sleep, dealing with domestic violence and difficult home relationships.
Babylon reports many patients are seeking advice on information about local council support services, seeking advice for activities to keep busy and how to remain healthy, and how to get support to access food and financial concerns. Ieso Digital Health reports up to a third of patients mentioning COVID-19 as a reason for presenting for mental health treatment and also report a rise in patient worries about viruses, with up to 15% of in-session worries about coronavirus and COVID-19.
Papa reported that 53% of users felt less lonely and that virtual companions have performed a range of tasks with elderly users (e.g., obtaining medications, online grocery shopping). Peer support specialists are being rapidly trained. Digital Peer Support trained 750 peer support specialists between 10 March and mid April 2020. Wisdo reported a 283% increase in the numbers of people replying to other people's messages and an increase of 115% in the number of people signing up for roles to provide support for others.
Many providers are experiencing increased support seeking behaviours. For example, Ieso Digital Health reports an 84% increase in referrals. Vala Health reports a doubled volume of mental health-related consultations with GPs during the period 10 March to 8 April 2020. By week four of the UK lockdown, general health enquiries had returned to almost pre-COVID levels, but mental health consultations continued to rise. NAMI reports a 41% increase in demand for HelpLine resources and information.
Ieso Digital Health reported an 84% increase in referrals to their 1-1 online CBT service in the weeks since the lockdown was announced in the UK, relative to the same time period in 2019. Wysa witnessed a 77% increase in new users during February-to-March 2020, as compared to the same period in 2019. Qualitative insights from Orygen (Australia) revealed that young people report privacy concerns in having telehealth consults with family members in the background.
Many providers report observations suggesting increased anxiety, uncertainty, loneliness, and loss. MHA reports that 45% of people who took an anxiety screen in March (N=11,033) scored for severe anxiety. In a self-reported questionnaire to members of The Mighty, 89% of members reported that their daily life has been at least somewhat impacted by increased anxiety; 43% say it has been extremely impacted. This is consistent with reports from XenZone demonstrating increases in sadness (up 161%), health anxiety (up 155%), sleep difficulties (up 90%), concerns over body image (up 43%), eating difficulties (up 31%), loneliness (up 23%), and bereavement (up 20%). The Mental Health Foundation survey reported that respondents felt increasingly lonely, and that this was most pronounced for people aged 18-24 (44%) and 25-34 (35%). Multiple providers report users mentioning their loss of access to care and human support (The Mighty, MeeTwo, Wysa, consultant NHS nurse).
Qntfy’s observations suggest decreased well-being in the general public, and that at times this has been greater amongst those who identify as healthcare providers. Unmind and Wysa reported higher anxiety levels in health staff as compared to the rest of their populations. Sangath reports that community health workers face “fears and insecurities among their patients, as well as added anxieties about the health and wellbeing of their own children and family members.” CBTClinics report a rise of anxiety and depressive type disorders from people emotionally close to frontline health staff (e.g. Parents, spouses, children).
Other outcomes include increases in reporting of unsafe domestic settings (Babylon, Wysa, Teen Line, XenZone), suicidal risk/ideation (MeeTwo, Qntfy, Mental chat, Beyond Blue, Mumsnet), and sleep disturbances (It’s OK to Talk, XenZone, Mumsnet, Qare, BioBeats, Wysa). There have also been increases in the prescription of anti-depressant medications (Jasvinder Kandola), increased requests for pain killers via telehealth (Vala Health), and increased activities on darknet markets mentioning psychiatric medications (The TellFinder Alliance).
A Money and Mental Health Policy Institute survey (N=568) reported a range of concerns by respondents with lived experience of mental health problems about how changes, as a result of COVID-19, might affect their finances: 62% worried about having to access the benefits system, 57% worried about losing their job and 56% worried about creditors chasing them for money. Tully and OpenWrks Group reported that 81% of self-employed customers (N=650) have declared that they do not have any work coming in due to COVID-19. 50% of their wider sample (N=1822) have had income reduced and 19% have lost their income. A Turn2us survey showed that 70% of respondents (N= 6,198) who have had employment affected are unable to afford rent or mortgages. An anonymous financial services provider also shared concerns that their on-site cashiers may be vulnerable and distressed by customer behaviours.
To our knowledge, this study is the first of its kind to bring together large numbers of researchers and private organisations, including financial services providers, to share digital data insights about the mental health concerns of millions of people online. A recent paper  has called for mental health monitoring to move beyond NHS linkage, in order to capture real incidence in the community and embrace new technologies measuring moment-to-moment change. This study provides a proof-of-concept for the viability of this. The information that we have quickly compiled has been sourced from different geographies, demographics, and types of digital interaction, and provides insight into the diversity of individual mental health needs.
The potential value of healthcare insights in financial data is already recognised [18,19] and financial services firms are not only a source of uniquely constructive data on household economies  but can also offer possible mechanisms of direct and indirect health interventions. Given this unique dual positioning, we feel it is important to include the insights of financial services providers to document and support the population’s mental health through this pandemic and beyond. More broadly, alongside public initiatives, we feel there is a possibility for different private sectors to play a beneficial role in the wider social and economic recovery.
Examining the finance sector provides just one example of the interconnectedness of today’s world and the potential benefits of interdisciplinary approaches to better manage the bigger picture of mental health.
We recognise that this study is not rigorous in terms of data collection and methodology. We did not choose these providers in a systematic way. Using data from digital service providers limits our population to people who have access to those digital platforms and many ‘hidden’ populations are not registering in digital spaces. Furthermore, we do not know whether our demographic is representative of any larger population, or whether whole-population impacts can be inferred from digital service impacts. In addition, we did not verify the insights shared by providers. This avoided privacy issues, but has the potential to have introduced inaccuracies or biases in the reported information. This study is also unable to characterise mental health problems at a clinical level because most digital providers did not report clinically-validated measurements. The use of digital measurements to monitor distress is still a work in progress.
Following this study, we will continue to work with providers to capture follow-up insights at later time points (for more information see ) and on-board new providers to address issues of data representativeness. It will also be important to capture insights that relate to resilience and recovery. An important next step will be to develop rigorous means to bring together public and private sector data to monitor mental health needs in real-time (just as contact tracing is used to manage the viral epidemic). This can fuel research and understanding, and help to inform high quality responses that can be delivered remotely to those in need.
Full institutional correspondence address for the corresponding author: Dr Becky Inkster, Wolfson College, University of Cambridge, Cambridge, UK, CB3 9LG, firstname.lastname@example.org
Contributors Statement: Dr Inkster formulated the notion to write about this topic and invited co-authors to join, all having different professional and/or lived experiences who have made important contributions in various ways, such performing literature searches, writing, helping us to connect with digital providers, idea generation, editing, interpretation etc.