A discrete model for the evaluation of public policies: The case of Colombia during the COVID-19 pandemic

• Published in: PloS one 2023-02, Vol.18 (2), p.e0275546
• Main Authors: Catano-Lopez, Alexandra, Rojas-Diaz, Daniel, Lizarralde-Bejarano, Diana Paola, Puerta Yepes, María Eugenia
• Format: Article
• Language: English
• Subjects: Asymptomatic; Behavior; Biology and Life Sciences; Colombia; Colombia - epidemiology; Communicable Disease Control; Control; COVID-19; COVID-19 - epidemiology; COVID-19 - prevention & control; Decision making; Diffusion processes; Disease control; Disease transmission; Epidemics; Epidemiology; Evaluation; Forecasts and trends; Humans; Infections; Mathematical models; Medical research; Medicine and Health Sciences; Medicine, Experimental; Modelling; Natural history; Numerical analysis; Pandemics; Pandemics - prevention & control; Parameter estimation; Parameters; Partial differential equations; People and places; Probability; Public health administration; Public Policy; Quarantine; Research and Analysis Methods; Severe acute respiratory syndrome coronavirus 2; Social Sciences; Time series; Vaccination; Viral diseases; Viral infections
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0275546

In mathematical epidemiology, it is usual to implement compartmental models to study the transmission of diseases, allowing comprehension of the outbreak dynamics. Thus, it is necessary to identify the natural history of the disease and to establish promissory relations between the structure of a mathematical model, as well as its parameters, with control-related strategies (real interventions) and relevant socio-cultural behaviors. However, we identified gaps between the model creation and its implementation for the use of decision-makers for policy design. We aim to cover these gaps by proposing a discrete mathematical model with parameters having intuitive meaning to be implemented to help decision-makers in control policy design. The model considers novel contagion probabilities, quarantine, and diffusion processes to represent the recovery and mortality dynamics. We applied mathematical model for COVID-19 to Colombia and some of its localities; moreover, the model structure could be adapted for other diseases. Subsequently, we implemented it on a web platform (MathCOVID) for the usage of decision-makers to simulate the effect of policies such as lock-downs, social distancing, identification in the contagion network, and connectivity among populations. Furthermore, it was possible to assess the effects of migration and vaccination strategies as time-dependent inputs. Finally, the platform was capable of simulating the effects of applying one or more policies simultaneously.