Mitigating Covid-19 outbreaks in workplaces and schools by hybrid telecommuting

Simon Mauras, Vincent Cohen-Addad, Guillaume Duboc, Max Dupré la Tour, Paolo Frasca, Claire Mathieu, Lulla Opatowski, Laurent Viennot

Abstract: The COVID-19 pandemic necessitated unprecedented contact-limiting restrictions across workplaces and educational institutions. However, the quantitative impact of specific hybrid telecommuting strategies remained largely unexplored. This study employs a simulation-based approach to analyze outbreak magnitudes on three real-life contact networks: a workplace, a primary school, and a high school in France. Using an SEIR transmission model enriched with COVID-19-specific parameters—including super-spreaders, asymptomatic infectiousness, and pre-symptomatic periods—we evaluate two primary strategies: "Rotating" (splitting cohorts into alternating shifts) and "On-Off" (alternating the entire group between on-site and remote work). Our results consistently rank Rotating week-by-week as the most effective mitigation strategy across all network topologies. Furthermore, we identify a critical threshold: when the baseline reproduction number $R_0$ is below 1.38, these hybrid strategies can effectively suppress outbreaks by reducing the effective reproduction number $R_e$ to less than 1.

Introduction

Early pandemic responses relied heavily on large-scale lockdowns, which, while effective at curbing viral spread, imposed severe economic and social burdens. Hybrid telecommuting—alternating between on-site presence and remote work—emerged as a potential middle ground. The objective of this research is to quantify the relative effectiveness of different scheduling patterns (Daily vs. Weekly, Rotating vs. On-Off) in suppressing viral transmission within localized social environments like offices and classrooms.

Methodology

To simulate the spread of the virus, researchers utilized a stochastic SEIR (Susceptible-Exposed-Infectious-Recovered) model. The model was specifically tuned to reflect the biological realities of COVID-19:

  • Super-spreaders: Modeled via a Gamma distribution to account for individuals with high infectiousness.
  • Asymptomatic/Pre-symptomatic Transmission: The model accounts for the significant portion of spread occurring before or in the absence of symptoms.
  • Imperfect Compliance: Recognizing that social interactions do not cease entirely during remote periods, the model assumes individuals maintain 25% of their usual interactions through a "background" graph.

The study tested five primary strategies against a "No Strategy" baseline: On-Off Daily, On-Off Weekly, Rotating Daily, Rotating Weekly, and Full Telecommuting.

Datasets

The simulations were grounded in high-resolution empirical contact data from the SocioPatterns project. This data, collected via RFID sensors, tracked face-to-face interactions (within 1.5m) at 20-second intervals in three distinct French environments: a primary school (242 individuals), a workplace (217 individuals), and a high school (327 individuals). A synthetic random graph was also used as a control to ensure results were not merely artifacts of specific community structures.

Key Results

The study found a consistent ranking of strategies across all tested environments. "Rotating" strategies, which split the population into two cohorts that never meet on-site, significantly outperformed "On-Off" strategies, where the entire group alternates together.

Figure 4: Comparison of outbreak probability, time to outbreak, and total infected across different strategies for a baseline R0 of 1.25.

As shown in the figure above, for a high school environment with a baseline $R_0$ of 1.25, the "No Strategy" scenario results in a 27% probability of an outbreak. Implementing a Rotating Weekly schedule reduces this probability to approximately 12%, nearly approaching the safety levels of Full Telecommuting (~7%), which still carries risk due to persistent outside-work contacts.

Insights and Discussion

Two primary factors drive the success of the Rotating Weekly strategy:

  1. Structural Cohorting: By ensuring Group A and Group B never overlap on-site, the network is effectively halved, breaking numerous potential transmission chains.
  2. Biological Synchrony: Weekly rotations align more effectively with the virus's incubation and infectious periods. An individual infected at the end of their on-site week will likely spend their most infectious period remote during the following week.
Figure 5: Effective reproduction number (Re) as a function of the baseline R0 for various strategies.

A critical finding is the "1.38 Threshold." As illustrated in the $R_e$ vs $R_0$ plot, if the baseline $R_0$ within a network is below 1.38, all four hybrid strategies are capable of pushing the effective reproduction number $R_e$ below the critical value of 1.0, thereby preventing a sustained outbreak.

Conclusion and Limitations

This study provides a clear hierarchy for organizational planning: Rotating Weekly is the superior hybrid model for mitigating localized COVID-19 spread. However, the authors note that these simulations model "closed systems" and do not account for external infection vectors like public transport or family members. Additionally, while based on real-world data, the small scale of these networks (hundreds of people) suggests that future research should explore how these localized "blocks" interact within larger city-wide dynamics.

Summary updated

Abstract

The COVID-19 outbreak has forced most countries to impose new contact-limiting restrictions at workplaces, universities, schools, and more roadly in our societies. Yet, the power and imitations of these unprecedented strategies or containing virus spread within the populations remain unquantified. Here, we develop a simulation study to analyze COVID-19 outbreak magnitudes on three real-life contact networks stemming from a workplace, a primary school and a high school in France. Our study provides the first fine-grained analysis of the impact of contact-limiting strategies at work-places, schools and high schools, including (1) Rotating, in which workers are evenly split into two shifts that alternate on a daily or weekly basis; and (2) On-Off, where the whole group alternates periods of normal work interactions with complete telecommuting. We model epidemics spread in these different setups using an SEIR transmission model enriched with the coronavirus most salient specificities: super-spreaders, infec-tious asymptomatic individuals, and pre-symptomatic infectious periods. Our study yields clear results: The ranking of the strategies based on their ability to mitigate epidemic propagation in the network from a first index case is the same for all network topologies (work place, primary school and high school). Namely, from best to worst: Rotating week-by-week, Rotating day-by-day, On-Off week-by-week, and On-Off day-by-day. Moreover, our results show that when the baseline reproduction number R0 within the network is < 1:38, all four strategies efficiently control outbreak by decreasing effective Re to <1. These results can support public health decisions and telecommuting organization locally.
BibTeX
@article{mauras2021mitigating,
  author = {Simon Mauras and Vincent Cohen-Addad and Guillaume Duboc and Max Dupré la Tour and Paolo Frasca and Claire Mathieu and Lulla Opatowski and Laurent Viennot},
  title = {{Mitigating Covid-19 outbreaks in workplaces and schools by hybrid telecommuting}},
  journal = {PLOS Computational Biology},
  year = {2021},
  doi = {10.1371/journal.pcbi.1009264},
}
Figure