Changes in social contact patterns in Germany during the SARS-CoV-2 pandemic – an analysis based on the COVIMOD study

Abstract

Background

During the SARS-CoV-2 pandemic, Germany employed several nonpharmaceutical interventions (NPIs) to reduce social contacts and decelerate the virus’s spread. Associations between demographics and other factors, e.g. perceived pandemic threat level, might help explain variations in social contact behaviours. We aimed to estimate contact numbers during the pandemic in Germany and assess factors associated with changes therein.

Methods

Between 04/2020 and 12/2021, we conducted an online contact survey (COVIMOD) with 33 waves in Germany. We calculated the mean and 95% confidence interval of daily reported contacts (“people who you met in person and with whom you exchanged at least a few words, or with whom you had physical contact”) using bootstrapping. The effects of different factors on the number of contacts were determined by fitting generalized additive models (GAMs).

Results

The COVIMOD survey recorded 59,585 responses from 7,851 participants across Germany. The overall mean number of daily social contacts during the study period was 3.30 (95%CI: 3.23–3.38), with the number of non-household contacts being twice as high as the number of contacts with household members. The lowest overall number of contacts (2.11, 95%CI: 2.01–2.22) was reported during Germany's strongest contact reduction campaigns (end of 04/2020), when the number of household contacts was three times higher than non-household contacts. The highest number of contacts (6.38, 95%CI: 5.67–7.15) was observed during periods of relaxed measures (June 2020), when household contacts were four times fewer than non-household contacts. The work and school contacts shaped the overall variation of contact patterns in Germany during the pandemic. In participants under 18 years, partially/fully closing schools reduced school contacts by 83% (95%CI: 80–85%) and overall contacts by 39% (95%CI: 36–42%). Higher risk perceptions regarding COVID-19 were associated with 11% (95% CI: 2–17%) more social contacts among all participants and 66% (95%CI: 32–108%) more work contacts in the adult participants.

Conclusions

Our study revealed fluctuations in the number of social contacts during the SARS-CoV-2 pandemic in Germany, with substantial variations influenced by NPIs and individual factors. Understanding these factors affecting social contacts is vital for refining disease transmission models and informing future pandemic response strategies.