Public Mobility and Social Media Attention in Response to COVID-19 in Sweden and Denmark | Media and Youth | JAMA Network Open | JAMA Network
[Skip to Navigation]
Sign In
Figure.  Mobility Change, Twitter Volume, and Spread of Coronavirus Disease 2019 (COVID-19) in Denmark and Sweden
Mobility Change, Twitter Volume, and Spread of Coronavirus Disease 2019 (COVID-19) in Denmark and Sweden

Public mobility was compared with the prepandemic baseline period (January 3 to February 6, 2020) across retail and recreational spaces (eg, restaurants, cafes, shopping centers, theme parks, museums, libraries, and movie theaters). Public holidays are marked in red.

Table.  Associations of COVID-19 Interventions With Mobility Change in Denmark and Sweden
Associations of COVID-19 Interventions With Mobility Change in Denmark and Sweden
1.
Flaxman  S, Mishra  S, Gandy  A,  et al; Imperial College COVID-19 Response Team.  Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe.   Nature. 2020;584(7820):257-261. doi:10.1038/s41586-020-2405-7PubMedGoogle ScholarCrossref
2.
Google LLC. Google COVID-19 community mobility reports. Accessed June 14, 2020. https://www.google.com/covid19/mobility/
3.
Drake  TM, Docherty  AB, Weiser  TG, Yule  S, Sheikh  A, Harrison  EM.  The effects of physical distancing on population mobility during the COVID-19 pandemic in the UK.   Lancet Digit Health. 2020;2(8):e385-e387. doi:10.1016/S2589-7500(20)30134-5PubMedGoogle ScholarCrossref
4.
Dong  E, Du  H, Gardner  L.  An interactive web-based dashboard to track COVID-19 in real time.   Lancet Infect Dis. 2020;20(5):533-534. doi:10.1016/S1473-3099(20)30120-1PubMedGoogle ScholarCrossref
5.
Bernal  JL, Cummins  S, Gasparrini  A.  Interrupted time series regression for the evaluation of public health interventions: a tutorial.   Int J Epidemiol. 2017;46(1):348-355.PubMedGoogle Scholar
6.
Li  Y, Campbell  H, Kulkarni  D,  et al. The temporal association of introducing and lifting non-pharmaceutical interventions with the time-varying reproduction (R) number of SARS-CoV-2: a modelling study across 131 countries.  The Lancet Infect Dis. Published online October 20, 2020. doi:10.1016/S1473-3099(20)30785-4
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 10,457
    Citations 0
    Research Letter
    Public Health
    January 4, 2021

    Public Mobility and Social Media Attention in Response to COVID-19 in Sweden and Denmark

    Author Affiliations
    • 1Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden
    • 2National Centre for Register-based Research, Department of Economics and Business Economics, Aarhus University School of Business and Social Sciences Aarhus, Denmark
    JAMA Netw Open. 2021;4(1):e2033478. doi:10.1001/jamanetworkopen.2020.33478
    Introduction

    To reduce the spread of coronavirus disease 2019 (COVID-19), countries have implemented different societal interventions, and evaluation of the effect on public response is needed.1 Sweden and Denmark are comparable countries in terms of health care and sociodemographic characteristics; however, Denmark was one of the first countries to enforce lockdown and subsequent gradual reopening, whereas Sweden has had few restrictions, largely limited to public recommendations. We assessed public mobility and social media attention associated with COVID-19 spread and societal interventions from February 15 to June 14, 2020, in Denmark and Sweden.

    Methods

    For this ecological study, public mobility was measured by Google mobility reports, providing daily percentage change in the number of visitors to public spaces compared with baseline (January 3 to February 6, 2020).2 We focused on mobility across retail and recreational spaces based on previous findings.3 Daily volume of tweets including COVID-19–related hashtags and keywords (eTable in the Supplement) were collected via Sprout Social to study public attention to the disease, capturing 732 634 tweets in Sweden and 324 730 in Denmark. Because this study used summary statistics rather than individual-level data, an institutional ethical review permit was not required according to the rules at Karolinska Institutet, Sweden. This study followed the part of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline that are relevant to ecological studies.

    The exposures that we focused on were disease spread (ie, daily new COVID-19 cases)4 and key societal interventions and announcements (ie, addresses by the prime minister or head of state) targeting public mobility. Interrupted time series analyses were conducted to evaluate the association between societal interventions and mobility, including both a level (β1) and a slope change (β2) that can be interpreted as the modeled percentage change in mobility from before to after the intervention.5 Statistical analyses were performed using R, version 3.6.3 (R Project for Statistical Computing). Details on measures and interrupted time series analyses are provided in the eMethods in the Supplement.

    Results

    The percentage of mobility change, the number of COVID-19–related tweets, and the number of confirmed cases of COVID-19 are plotted with key societal interventions in the Figure. Daily volumes of COVID-19–related tweets followed a similar pattern in Denmark and Sweden, with volumes increasing exponentially during the beginning of disease spread and peaking between March 13 and 17 (from 746 on February 14 to 16 851 on March 13 in Sweden and from 25 to 8398 in Denmark). During this period, the World Health Organization declared a COVID-19 pandemic, Denmark announced a lockdown, and Sweden banned public gatherings or more than 500 people and recommended remote higher education and work. Twitter volumes then decreased with time in both countries. Public mobility to retail and recreation decreased in both countries during the study period, with a greater reduction in Denmark during lockdown. The maximum mean (SD) mobility reduction (apart from public holidays) in Denmark was 38% (2.6%) (from March 23 to 29) and 24% (2.1%) in Sweden (from March 30 to April 5). By the last week of the study period (June 8 to 14), mobility had returned to baseline levels in both countries, with a mean (SD) change of 2.14% (4.4%) in Denmark and −0.14% (5.1%) in Sweden (Figure).

    The interrupted time series analyses showed that the Danish lockdown was associated with a 20% immediate decrease in mobility (β1, –20.19; 95% CI, –27.79 to –12.59) (Table). The ban of gatherings of more than 10 people coincided with the Queen’s address to the nation and was associated with a 12% decrease in mobility (β1, –12.20; 95% CI, –19.74 to –4.65). Phase 1 and 2 of the Danish reopening were associated with immediate reductions in mobility but increased mobility in the subsequent days. In Sweden, only the ban on public gatherings of more than 500 people was associated with a significant decrease in mobility (β1,–11.04; 95% CI, –14.66 to –7.42). Neither the prime minister’s nor the King’s address to the nation showed a statistically significant association with mobility. In both countries, higher numbers of daily confirmed COVID-19 cases were associated with reduced mobility (Denmark: β1,–2.66 [95% CI, –4.51 to –0.82]; Sweden: β1, –0.62 [95% CI, –0.84 to –0.40]).

    Discussion

    In this cross-country comparison, stricter interventions were associated with larger reductions in mobility. Bans on public gatherings showed stronger associations with mobility in both countries compared with national announcements by country leaders and public recommendations, which were the primary intervention types in Sweden. We also found a delay in reaching maximum reduction in mobility after restrictions and for normalization of mobility after relaxation of restrictions, in line with prior research.6 Social media attention to COVID-19 on Twitter was greater for early interventions and disease spread than for later interventions and disease spread, with nearly identical patterns of decreasing COVID-19–related Twitter volumes across Denmark and Sweden despite marked difference in number of cases and the type of societal interventions. Better understanding of how interventions are associated with public mobility and attention may guide policy in relation to the resurgence of COVID-19 currently observed in many countries. Our findings are limited to short-term effects of specific interventions and do not capture interactions between these nor other important factors likely to be associated with public mobility and attention (eg, global media, guidelines from international health care organizations, and treatment and vaccine developments). Any conclusions are limited by the representativeness of Twitter and Goggle mobility data.

    Back to top
    Article Information

    Accepted for Publication: November 22, 2020.

    Published: January 4, 2021. doi:10.1001/jamanetworkopen.2020.33478

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Zhang L et al. JAMA Network Open.

    Corresponding Author: Isabell Brikell, PhD, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Nobels väg 12 A, 171 77 Stockholm, Sweden (isabell.brikell@ki.se).

    Author Contributions: Drs Brikell and Zhang contributed equally to this work. Drs Brikell and Zhang had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Brikell, Chang.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Brikell.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Zhang, Chang.

    Obtained funding: Chang.

    Administrative, technical, or material support: Zhang, Chang.

    Supervision: Brikell, Dalsgaard, Chang.

    Conflict of Interest Disclosures: Dr Brikell reported receiving grants from the Swedish Brain Foundation. Dr Dalsgaard reported receiving grants from The Lundbeck Foundation, the Novo Nordisk Foundation, the European Commission, Helsefonden, and the European Union’s Horizon 2020 research and innovation programme. Dr Chang reported receiving grants from the Swedish Research Council and the Swedish Research Council for Health, Working Life and Welfare. No other disclosures were reported.

    References
    1.
    Flaxman  S, Mishra  S, Gandy  A,  et al; Imperial College COVID-19 Response Team.  Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe.   Nature. 2020;584(7820):257-261. doi:10.1038/s41586-020-2405-7PubMedGoogle ScholarCrossref
    2.
    Google LLC. Google COVID-19 community mobility reports. Accessed June 14, 2020. https://www.google.com/covid19/mobility/
    3.
    Drake  TM, Docherty  AB, Weiser  TG, Yule  S, Sheikh  A, Harrison  EM.  The effects of physical distancing on population mobility during the COVID-19 pandemic in the UK.   Lancet Digit Health. 2020;2(8):e385-e387. doi:10.1016/S2589-7500(20)30134-5PubMedGoogle ScholarCrossref
    4.
    Dong  E, Du  H, Gardner  L.  An interactive web-based dashboard to track COVID-19 in real time.   Lancet Infect Dis. 2020;20(5):533-534. doi:10.1016/S1473-3099(20)30120-1PubMedGoogle ScholarCrossref
    5.
    Bernal  JL, Cummins  S, Gasparrini  A.  Interrupted time series regression for the evaluation of public health interventions: a tutorial.   Int J Epidemiol. 2017;46(1):348-355.PubMedGoogle Scholar
    6.
    Li  Y, Campbell  H, Kulkarni  D,  et al. The temporal association of introducing and lifting non-pharmaceutical interventions with the time-varying reproduction (R) number of SARS-CoV-2: a modelling study across 131 countries.  The Lancet Infect Dis. Published online October 20, 2020. doi:10.1016/S1473-3099(20)30785-4
    ×