Optimal Models for Distributing Vaccines in a Pandemic

ABSTRACT

Distributing vaccines among a massive population is one of the challenging tasks in a pandemic. Therefore, health care organizations need to optimize the assignment of vaccination appointments for people while considering their priorities and preferences. In this paper, we propose two optimal vaccine distribution models as Integer Linear Programming (ILP) models;namely, Priority-based Model (PM) and Priority & Preference-based Model (PPM), to maximize the distribution of vaccines among a given population. In PM, we divide the people among several priority groups and ensure maximum vaccine distribution among the higher priority groups. However, along with the priority groups, PPM also considers a list of preferred vaccine distribution centers and time slots for each person. Thus, this model maximizes vaccine distribution among the higher priority groups by assigning appointments in their desired location and time. We analyzed the performance of our proposed models on a randomly generated dataset. In addition, we also performed a case study for our proposed models on the COVID-19 vaccination dataset from Thunder Bay, Canada. In both experiments, we show that PPM outperforms PM in full-filling people's preferences while maximizing the distribution of vaccines among the higher priority groups.