Analytical models and methods for the COVID-19 pandemic: a survey focused on progression and mitigation via non-pharmaceutical interventions

The ongoing global impact of the COVID-19 pandemic necessitates a complex, multistage approach involving containment, suppression and mitigation. We review analytical and operations research (OR) work dedicated to forecasting pandemic progression and devising non-pharmaceutical interventions (NPIs), with a particular focus on these topics due to their recent prominence and practical importance. This systematic review aims to provide decision-makers and healthcare planners with the most effective tools and motivate further focused and impactful research. A total of 79 papers were surveyed, 48% of which focus on pandemic prediction and the remainder on NPIs. Diverse predictive methods, including compartmental models, statistical tools and machine learning, showed remarkable accuracy in predicting pandemic spread and in assessing the effectiveness of NPIs. There is, however, no accepted best prediction method even though ensemble forecasting appears to outperform other methods. Furthermore, NPIs are recognized for their effectiveness in mitigating the pandemic and their efficacy is influenced by factors such as location, input parameters, and demographics. These findings underscore the necessity for a comprehensive comparison of predictive methods and for models combining predictive methods and optimization strategies.