Assessing the effect of non-pharmaceutical interventions on COVID-19 transmission in Spain, 30 August 2020 to 31 January 2021

Abstract

Background: After a national lockdown during the first wave of the COVID-19 pandemic in Spain, regional governments implemented different non-pharmaceutical interventions (NPIs) during the second wave. Aim: To analyse which implemented NPIs significantly impacted effective reproduction number (R-t) in seven Spanish provinces during 30 August 2020-31 January 2021. Methods: We coded each NPI and levels of stringency with a 'severity index' (SI) and computed a global SI (mean of SIs per six included interventions). We performed a Bayesian change point analysis on the R-t curve of each province to identify possible associations with global SI variations. We fitted and compared several generalised additive models using multimodel inference, to quantify the statistical effect on R-t of the global SI (stringency) and the individual SIs (separate effect of NPIs). Results: The global SI had a significant lowering effect on the R-t (mean: 0.16 +/- 0.05 units for full stringency). Mandatory closing times for non-essential businesses, limited gatherings, and restricted outdoors seating capacities (negative) as well as curfews (positive) were the only NPIs with a significant effect. Regional mobility restrictions and limited indoors seating capacity showed no effect. Our results were consistent with a 1- to 3-week-delayed R-t as a response variable. Conclusion: While response measures implemented during the second COVID-19 wave contributed substantially to a decreased reproduction number, the effectiveness of measures varied considerably. Our findings should be considered for future interventions, as social and economic consequences could be minimised by considering only measures proven effective.