Reduced social interaction strategies to flatten COVID-19 infection curve

Venky Krishnan, TIFR-CAM

The innate need for human beings to socially interact (characterized here by physical closeness and face-to-face interactions rather than over long distance using technology) with others on a regular basis is central to having a wholesome and well-rounded life. Limiting this social interaction, by lockdowns for instance, is one crucial strategy to arresting the spread of a pandemic such as the one we are in, given the absence of vaccines or readily available medical interventions. A complete lockdown, however, has severe implications to the economy, and it also affects our mental and social well-being. Furthermore, over time, we develop lockdown fatigue leading to diminishing returns. The authors in this article propose three moderate, social-interaction limiting strategies as an alternative to complete lockdowns:

  • Social interactions based on similarity: In this strategy, individuals interact only with other members who share a similar characteristic. These characteristics could include geographic closeness (interactions only with members of one's neighborhood), being members of an organization (interactions only with members of one's place of work) or age similarity (interactions only within one's age group).
  • Social interactions limited to cohesive communities: In this strategy, people limit their interactions with those individuals who share common contacts. For instance, as the authors describe through a simple example, two people socially interact only if they have several mutual friends in common. This leads to interactions that are localized in nature, limiting any potential spread to within those cohesive communities.
  • Repeated social interactions with only a select set of people: In this strategy, individuals decide with whom they would interact on a regular basis and restrict their meetings to only that set over time. This limits the number of people that one interacts with, while not reducing the urge to connect on a regular basis.

The authors use a social network based model where individuals are denoted by nodes (or vertices in a graph network) and an edge between two nodes denotes interaction between these two individuals. If one node gets infected, then the disease spreads to all the nodes that are connected to the infected node by an edge. With this framework, and using techniques from deterministic and probabilistic infection models, they show that the aforementioned reduced social interaction strategies lead to flattened infection curves.

[Last update 18 June 2020]