Hindsight in 2020: Modeling Preventative Measures for Select Diseases in Ontario using Probabilistic Dynamic Programming and SEIR Compartmental Frameworks

Abstract

In this work, we provide a “rear-view mirror” analysis for approaches to two different diseases that were relevant to Ontario in 2020. Using probabilistic dynamic programming, we simulate two versions of an optimized publicly funded shingles vaccination program for seniors; the first using existing vaccination data under the 2016-2019 program, and the second by building on a population with existing coverage and a new vaccine with improved efficacy. Then, using an SEIR compartmental model with ordinary differential equations, we use a derivative-free optimization method to model age-stratified case development of COVID-19 during the first wave of the pandemic, as well as determining the relative significance of specific NPI measures in slowing transmission. Lessons learned from the early handling of these diseases may inform better healthcare planning and prevent unnecessary future costs and suffering.