The impact of non-pharmaceutical interventions on SARS-CoV-2 transmission across 130 countries and territories

Abstract

Background: Non-pharmaceutical interventions (NPIs) are used to reduce transmission of SARS coronavirus 2 (SARS-CoV-2) that causes coronavirus disease 2019 (COVID-19). However, empirical evidence of the effectiveness of specific NPIs has been inconsistent. We assessed the effectiveness of NPIs around internal containment and closure, international travel restrictions, economic measures, and health system actions on SARS-CoV-2 transmission in 130 countries and territories.

Methods: We used panel (longitudinal) regression to estimate the effectiveness of 13 categories of NPIs in reducing SARS-CoV-2 transmission using data from January to June 2020. First, we examined the temporal association between NPIs using hierarchical cluster analyses. We then regressed the time-varying reproduction number (R_t) of COVID-19 against different NPIs. We examined different model specifications to account for the temporal lag between NPIs and changes in R_t, levels of NPI intensity, time-varying changes in NPI effect, and variable selection criteria. Results were interpreted taking into account both the range of model specifications and temporal clustering of NPIs.

Results: There was strong evidence for an association between two NPIs (school closure, internal movement restrictions) and reduced R_t. Another three NPIs (workplace closure, income support, and debt/contract relief) had strong evidence of effectiveness when ignoring their level of intensity, while two NPIs (public events cancellation, restriction on gatherings) had strong evidence of their effectiveness only when evaluating their implementation at maximum capacity (e.g. restrictions on 1000+ people gathering were not effective, restrictions on < 10 people gathering were). Evidence about the effectiveness of the remaining NPIs (stay-at-home requirements, public information campaigns, public transport closure, international travel controls, testing, contact tracing) was inconsistent and inconclusive. We found temporal clustering between many of the NPIs. Effect sizes varied depending on whether or not we included data after peak NPI intensity.

Conclusion: Understanding the impact that specific NPIs have had on SARS-CoV-2 transmission is complicated by temporal clustering, time-dependent variation in effects, and differences in NPI intensity. However, the effectiveness of school closure and internal movement restrictions appears robust across different model specifications, with some evidence that other NPIs may also be effective under particular conditions. This provides empirical evidence for the potential effectiveness of many, although not all, actions policy-makers are taking to respond to the COVID-19 pandemic.

Keywords: COVID-19; Health impact assessment; Longitudinal analysis; Non-pharmaceutical interventions; Pandemic; Policy evaluation; Public health intervention; Quantitative; SARS-CoV-2.

Fig. 1

Temporal changes in NPI stringency index (range = 0-100) by region. Countries with available data are assigned corresponding geographical regions based on the World Bank classification.

Fig. 2

The proportions of countries implementing NPIs in each group by region. Colours indicate different NPI groups: dark blue—internal containment and closure, red—international travel restrictions, green—economic measures, and light blue—health system policies; n values represent the total number of countries in each World Bank region; the y-axis shows the proportion of countries implementing NPIs per category, e.g. one means every country in the region is implementing some measures from the NPI group. The turning point of 13 April is shown as a dotted-and-dashed vertical line.

Fig. 3

Dendrogram from the hierarchical cluster analysis of NPI time-series by scenario. The height of the node connecting two NPIs on the dendrogram represents the degree of similarity between their time-series. For example, under the Maximum Effort Scenario, the time-series of “restrictions on internal movement” is more similar to that of “restrictions on gatherings” (linked at point A), compared to that of “close public transport” (linked at point B). The hierarchical clustering analysis relies on Ward’s method and Euclidean distances. The colour of the text boxes corresponds to the group each NPI is in; red dashed boxes indicate statistically significantly temporal clusters identified through bootstrapping.

Fig. 4

Variable selection results. Optimal models are based on a backward selection process using AIC/BIC. NPIs are colour coded based on their respective NPI categories. Cell-content represents corresponding p values: > 0.05—ns (i.e. non-significant); ≤ 0.05 and > 0.01—*; ≤ 0.01 and > 0.001—**; ≤ 0.001 —***.

Fig. 5

Effect sizes for each NPI from the selected models. Points and lines indicate mean and 95% confidence intervals.