Denmark, Sweden Reduced Mobility With Different COVID-19 Interventions

Public responses to the coronavirus disease 2019 (COVID-19) pandemic have varied widely throughout the world, and a new study shows both lockdowns and public recommendations are associated with reductions in the mobility of visitors to public spaces.

The study, detailed in a research letter in JAMA Network Open, looked at associations between interventions in Denmark and Sweden and public mobility and social media attention associated with COVID-19 spread from Feb. 15 to June 14.

“Based on data from Denmark and Sweden, two similar Nordic countries but with different COVID-19 intervention strategies and rates of disease spread, our findings suggest that both bans and public recommendations were associated with reduced mobility in both countries (i.e. fewer visitors to retail and recreational spaces) during spring 2020, but stricter bans co-coincided with greater immediate mobility reductions,” corresponding author Isabell Brikell, PhD, Department of Medical Epidemiology and Biostatistics at Karolinska Institutet in Stockholm, Sweden, told Contagion®.

The investigators from Karolinska Institutet and Aarhus University in Denmark, analyzed Google mobility reports that detailed changes in the number of visitors to public places such as retail and recreational spaces. They also looked at the volume of tweets that included hashtags associated with COVID-19 along with new cases of COVID-19 and societal interventions and announcements from government leaders.

Using statistical computing, they plotted mobility change, tweets related to COVID-19 and new cases with societal interventions.

Public mobility decreased in both countries, associated with a lockdown in Denmark and a ban of large public gatherings of more than 500 people and recommendation of remote higher education and work in Sweden. The study reported a maximum mobility reduction of 38% in Denmark from March 23 to 29 and 24% in Sweden from March 30 to April 5. Public addresses from government leaders were not associated with any significant further reductions in mobility.

“We expected to see a greater reduction in Denmark due to the strict lock-down enforced, so I was somewhat surprised that the difference in mobility between Sweden and Denmark was not bigger and that mobility was strongly reduced in Sweden as well (where retail and recreation remained open),” Brikell said. “Another interesting finding was that attention to COVID-19 on Twitter followed a near identical pattern in both countries, with high attention to COVID-19 in the beginning of the pandemic which declined rapidly over time. This was somewhat surprising given the differences in disease spread, number of COVID-19 related deaths (higher in Sweden) and types of societal interventions (lockdown in Denmark, mainly public recommendations in Sweden).”

The study included 732634 COVID-19-related tweets in Sweden and 324730 in Denmark. The volume of tweets related COVID-19 increased the most early in pandemic between Feb. 14 to March 13. During that time the volume of COVID-19-related tweets rose from 746 to 16851 in Sweden and from 25 to 8398 in Denmark.

“Our results may aid in planning societal interventions for the ongoing/coming COVID-19 outbreaks,” Brikell said. “However, any intervention to reduce mobility to public spaces must be weighed against other important aspects, for example economic impact and psychological effect of social isolation, which we did not investigate.

“In addition to the limitations outlined in the article, it is also important to note that we did not study how changes in mobility affected COVID-19 spread, and the mobility data does not contain information about whether public recommendations (e.g. social distancing) were followed when traveling to or spending time in public spaces.”

Other studies have suggested such measure are effective in combatting the spread of COVID-19. A recent study by investigators at the University of Wisconsin-Madison showed that implementing social distance measures a week earlier would have reduced confirmed cases in New York City by 80% at the end of May 2020.

An earlier study of 4 metropolitan areas released by the US Center for Disease Control and Prevention found that the average 3-day percentage change of cases generally declined as social distancing measures were put in place and as more people stayed home in each community.

“We are not planning on further analyses of the Google mobility data (several interesting research articles have used these during fall 2020),” Brikell said. “We are however considering how Twitter data may be used to gauge public opinion across a range of important topics in relation to COVID-19, including vaccines. Given that so much of the debate regarding COVID-19 happens on social media, it is important for researchers, clinicians and policy makers to understand how online debate shapes opinion and behaviors regarding important health issues.”