In a recent study posted to the Research Square* preprint server researchers evaluated the effectiveness of social distancing measures adopted to curb the spread of coronavirus disease 2019 (COVID-19).
Most countries introduced public health measures to curtail the transmission of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the etiologic agent of the COVID-19 pandemic.
These restrictions were enforced to minimize social contacts and prevent overwhelmingly large epidemics that could potentially paralyze the national health care systems.
One of the most prominent measures adopted was social distancing that banned mass gatherings school closures, and limited international travel and internal mobility. Prior studies reported the positive effects of these measures and observed that social distancing had saved human lives. However, it is noteworthy that these conclusions were drawn based on the data from the early COVID-19 period when social distancing was stringent and effective.
In a previous report, the present study’s authors showed that if social distancing was not sustainable from the onset of the outbreak until attaining herd immunity, a subsequent outbreak could negate the results of previously enforced measures. This led to speculation whether less stringent measures could yield effective results.
The study
This study evaluated the effectiveness of social distancing in a model-based system analysis with Austria and Slovenia as the model countries. The researchers calibrated a compartmental model against a year-long epidemiological period for Slovenia and Austria. They simulated a hypothetical model without social distancing to assess the added value of decreasing social contacts. Additionally, they simulated scenarios where social distancing was less stringent in the early phases of the COVID-19 pandemic.
The standard susceptible-infected-resistant (SIR) model was extended to include compartments: susceptible (S), infected (I), infected in isolation (II), resistant symptomatic (RS), resistant asymptomatic (RA), deaths (D), and deaths due to denial of intensive care treatment (DL). They assumed that the entire population was initially susceptible in order to measure the proportion of people in each compartment, with susceptible individuals contracting infection over time.
Results
The compartment model, after calibration, generated close trajectories of deaths (daily and cumulative) in Austria and Slovenia. A higher basic reproduction number (R0) was observed for Slovenia as the daily number of contacts was larger than that in Austria. Cumulative infections were > 2.1 million in Austria and 1.2 million in Slovenia, far higher than national records, suggesting that 82% and 75% of infections in Slovenia and Austria were undocumented.
The number of fatalities was compared between the data calibrated model and the hypothetical simulation, which assumed no social distancing. Google mobility data from the pre-pandemic period was used as the proxy for social contacts in the hypothetical case. The authors observed significant mitigation of initial outbreaks in Slovenia and Austria with the implementation of social distancing; however, subsequent outbreaks were large.
While social distancing was effective initially in the pandemic with fewer fatalities, this effect was not reproduced over the long term. Implementing social distancing in scenarios with high R0 resulted in multiple outbreaks, whereas those with low R0 resulted in only one outbreak. These findings implied that social distancing was effective when R0 was low and less effective with higher R0. This meant that social distancing could have reduced deaths by 40% in Slovenia and 63% in Austria in the study period.
The research team hypothesized that subsequent outbreaks in Winter 2020-21 would have been smaller if social distancing had been less stringent during the early outbreaks in Spring 2020. To this end, they simulated less intense social distancing in the initial outbreak(s) with gradual increments of mobility volumes in Slovenia and Austria. This simulation suggested a sinusoidal relationship with tipping points that changed from “less social distancing: more cumulative deaths” to “less social distancing: fewer cumulative deaths.” In one effect, less stringent social distancing in Spring 2020 caused a balanced allocation of infections in the two seasons, i.e., more infections in Spring relieved the health care systems in the Winter 2020-21 outbreak, decreasing the overall number of fatalities. In the second effect, potentially high outbreaks in winter were mitigated by the less stringent social distancing in Spring resulting in an overall decrease in cumulative infections.
Conclusions
The current study demonstrated that limiting contacts significantly curbed the number of infections and fatalities during the early pandemic period. But in the long term, these beneficial effects were less prominent due to larger subsequent outbreaks. The falling effectiveness of social distancing could be due to robust containment initially and the high proportion of susceptible people that coincided with the increased transmissibility of the virus in winter.
Moreover, relaxing the restrictions during low transmission seasons to relieve pressure on health care systems in high transmission seasons mitigated winter outbreaks. This was effective in places where seasonality is distinct and long-term containment cannot be implemented due to high R0.
*Important notice
Research Square publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.
