Adaptive learning in containment measures: evaluation of policy interventions during the 2020 waves of Covid-19 in Italy

Abstract

ABSTRACT The COVID-19 pandemic hit Italy very harshly in two waves, the first in spring 2020 and the second between the autumn and the winter of the same year. Data show some major differences between the two phases; in particular, the first wave caused fewer infections but had a higher fatality rate. These pandemic evolutions, together with modified social conditions, called for a rapid adaptation of containment measures, i.e. stricter and homogeneous in the first wave, flexible and diversified in the second wave. The interrupted time series analysis applied to daily data on new cases yields positive results for both interventions in flattening the infection curve. The policies achieved almost the same percentage of positive cases avoided in the two waves. Adaptive and diversified policies based on learning from previous results seem to be suitable for this kind of decision-making in conditions of uncertainty.