Alexandr A. Bryzgalov, Johannes Ponge, Janik Suer, Tyll Krueger, Beryl Musundi, Chao Xu, Wolfgang Böck, Johannes Horn, Mahreen Kahkashan, André Karch, Julian Patzner, Mirjam Kretzschmar, Rafael Mikolajczyk, Carla Hartmann, Alexandr A. Bryzgalov, Beryl Musundi, Chao Xu, Mahreen Kahkashan, Myka Sarajan, Julian Patzner, Hannah Derwanz, Johannes Ponge, Bernd Hellingrath, André Karch, Veronika K . Jaeger, Janik Suer, Huynh Thi Phuong, Mirjam Kretzschmar, Vitaly Belik, Andrzej K. Jarynowski, Marlli Zambrano, Steven Schulz, Richard Pastor, Alejandra Rincón, Ashish Thampi, Alexander Kuhlmann, André Calero Valdez, Lilian Kojan, Markus Schölz, Jan Pablo Burgard, Soheil Shams, João Vitor Pamplona, Wolfgang Böck, Lukas Bayer, Sudarshan Tiwari, Berit Lange, Isti Rodiah, Wolfgang Greiner, Maren Steinmann, Sebastian Gruhn, Uwe Siebert, Beate Jahn, Tyll Krueger
Most models for infectious disease spread simplify contact heterogeneity by assuming constant rates within a week. However, empirical studies show clear variation, such as reduced workplace contacts on weekends. In this work, we investigate the effects of daily variation in workplace contacts on the spread of respiratory infections using the individual-based framework GEMS (German Epidemic Micro-Simulation System) with a synthetic population of 5 million individuals. We compare the scenario with uniform daily contacts to the scenario with more contacts on workdays and fewer on weekends, keeping weekly totals constant. Simulations reveal that uniform contact rates yield higher prevalence for short latent periods (1-2 days) and lower prevalence for longer latencies (5-6 days). The effect diminishes for long infectious periods (≥ 7 days) but is more pronounced at lower basic reproduction numbers. Depending on the disease specification and the differences in weekday contacts, the impact of weekday heterogeneity can be very strong. These findings open new possibilities for weekday-dependent mitigation measures. We conclude that weekly contact dynamics should be explicitly incorporated into epidemic models to avoid systematic errors in reproduction number estimation.