Evolution of disease transmission during the COVID-19 pandemic: patterns and determinants

Epidemic models are being used by governments to inform public health strategies to reduce the spread of SARS-CoV-2. They simulate potential scenarios by manipulating model parameters that control processes of disease transmission and recovery. However, the validity of these parameters is challenged by the uncertainty of the impact of public health interventions on disease transmission, and the forecasting accuracy of these models is rarely investigated during an outbreak. We fitted a stochastic transmission model on reported cases, recoveries and deaths associated with SARS-CoV-2 infection across 101 countries. The dynamics of disease transmission was represented in terms of the daily effective reproduction number ([Formula: see text] ). The relationship between public health interventions and [Formula: see text] was explored, firstly using a hierarchical clustering algorithm on initial [Formula: see text] patterns, and secondly computing the time-lagged cross correlation among the daily number of policies implemented, [Formula: see text] , and daily incidence counts in subsequent months. The impact of updating [Formula: see text] every time a prediction is made on the forecasting accuracy of the model was investigated. We identified 5 groups of countries with distinct transmission patterns during the first 6 months of the pandemic. Early adoption of social distancing measures and a shorter gap between interventions were associated with a reduction on the duration of outbreaks. The lagged correlation analysis revealed that increased policy volume was associated with lower future [Formula: see text] (75 days lag), while a lower [Formula: see text] was associated with lower future policy volume (102 days lag). Lastly, the outbreak prediction accuracy of the model using dynamically updated [Formula: see text] produced an average AUROC of 0.72 (0.708, 0.723) compared to 0.56 (0.555, 0.568) when [Formula: see text] was kept constant. Monitoring the evolution of [Formula: see text] during an epidemic is an important complementary piece of information to reported daily counts, recoveries and deaths, since it provides an early signal of the efficacy of containment measures. Using updated [Formula: see text] values produces significantly better predictions of future outbreaks. Our results found variation in the effect of early public health interventions on the evolution of [Formula: see text] over time and across countries, which could not be explained solely by the timing and number of the adopted interventions.