AN OPTIMIZATION MODEL TO REDUCE THE HUMAN FATALITIES IN SRI LANKA DUE TO COVID-19 CONSIDERING DIFFERENT NON- PHARMACEUTICAL INTERVENTION PLANS

Abstract

By considering Sri Lanka's limited healthcare resources and economic productivity in the absence of a vaccine, this study aims to control an epidemic outbreak while balancing its sanitary and economic repercussions to reduce the effects of COVID-19 and investigate its dynamics. The World Health Organization (WHO) recommended implementing a number of non-pharmaceutical interventions (NPIs). As a result, certain combinations of NPIs together with lockdowns were used to control the outbreak in Sri Lanka. These control measures are not only economically expensive but also have the potential to cause political instability, societal exhaustion, and annoyance. NPIs' stringency has reduced the number of deaths due to COVID-19, but it has also had negative consequences for the public and commercial sectors, hindering economic growth and significantly affecting people’s mental health. In order to assist policymakers in assessing the level of funding that the nation can afford to reduce the spread of disease and, as a result, the number of fatalities, we propose an optimization model after carefully analyzing the Sri Lankan epidemic context. By empowering policymakers to execute a series of NPIs in the relevant district using the available but constrained health care resources, we aim to identify a method to reduce the number of deaths within the various budget options. The first step is to apply a Mixed Integer Non-Linear Programming epidemic model to determine the optimal NPI sequence for each of the 25 districts over various planning horizons. Non-linear terms in the model are linearized to transform this Mixed Integer Non-Linear Programming model to a Mixed Integer Linear Programming model. This Mixed Integer Linear Programming model is thus transformed into an Integer Linear Programming model using the decreasing severity property of the NPI sequence. For each district, three plans, namely P_1, P_2, and P_(3,) are taken into consideration. There are no limitations on lockdowns in the plan P_1, but plans P_2 and P_3 included relaxations on the implementation of lockdowns. By altering the budget, the fatality is estimated over the three plans. The IBM ILOG Optimization Studio is used to solve the Mixed Integer Linear Programming model. This study shows that the NPI sequence, which has stringent lockdown conditions, is a better implementation plan to minimize the infection and, consequently, minimize the fatality in the country.