As the COVID-19 spread through Europe in March 2020, the mental health consequences of the pandemic started to receive broad public attention. Policymakers were confronted with the challenge to respond rapidly to an unprecedented situation, where little was known about the potential side-effects of strict containment measures. Experts warned that the combination of adverse factors, including grief and fears of infection, social isolation during shelter-in-place orders and school closures, as well as economic uncertainty, could be detrimental for population mental health. 

In Lausanne, Marius Brülhart and Rafael Lalive faced questions on the effects as members of the Swiss government’s COVID-19 Science Task Force. Their idea was to turn to the national telephone helpline, Die Dargebotene Hand, to obtain real-time information on the mental and social problems expressed by the population. In a pilot study published in May 2020, they concluded that the worst fears probably had not come to pass: the (relatively mild) lockdown in Switzerland had not triggered a significant increase in helpline calls, and any additional calls were mainly due to people worrying about infection and health risks rather than to deeper mental and social concerns.

Around the same time in Germany, a discussion between my friend and co-author Stephanie Reich and me led us to investigate trends in online-search behavior, where we discovered a large spike in search volumes for the country’s largest helpline Telefonseelsorge. Curious about whether a similar increase was also recorded for actual calls, we decided to reach out to the helpline and were excited to learn that in January 2019, they had started to systematically collect information and that they were willing to make the data available to us for a more in-depth analysis. A first look at the data revealed indeed a noticeable increase in call volumes around the pandemic outbreak, particularly for calls related to fears and loneliness.

Shortly before we published our preprint in May 2020, we discovered the study by Marius Brülhart and Rafael Lalive. Soon after sharing our results, we scheduled a first joint Zoom meeting in which we explored the potential for collaboration and further research. We quickly realized that in order to better understand the different patterns observed, we would need data from other countries, with a different timeline of the pandemic and where stricter containment measures were implemented.

In the following months, we reached out to crisis helplines around the world, explored alternative empirical strategies, and sequentially included more helplines in the analyzes as we received and cleaned additional data. Although the individual helplines do not collect information in a unified framework, we found sufficient consistency in the classification of conversation topics and caller characteristics to combine the data in our pooled analyses. During this exciting time, we benefited from helpful discussions with helpline staff that helped us to better understand the nature of the available information. In particular, several highly insightful meetings with Alena Goldstein, Jonathan Higgins, and Sean Murphy of Vibrant Emotional Health allowed us to make sense of the data on Lifeline call volumes across US states. Moreover, we received valuable feedback as we presented some of our results online at the annual meeting of the Telefonseelsorge, as well as in seminars at the Universities of Basel and Freiburg. Although this feels natural by now, it is worth noting that during the entire project, all communication between Lausanne and Freiburg took place online in Zoom.

In early 2021, we had documented a number of consistent patterns across different helplines and countries during the pandemic outbreak, as well as subsequent dynamics during the second and third waves of infections in France, Germany, and the USA. As we felt that our results from this under-explored data source were of interest to a large audience, we decided to submit our manuscript to Nature. The highly constructive comments we received from reviewers helped to substantially improve the paper.