Análisis del Efecto de la Pandemia de Covid-19 en los Servicios Hospitalarios de una Póliza de Salud Privada Mediante Metodologías de Series de Tiempo con Intervención e Interrumpidas

Abstract

Time series are often affected by unusual or outlier values, making it necessary to measure the influence of infrequent events or to determine and consider the intervention point affecting the time series for accurate forecasting of future values and model parameter estimation. The 2019-2020 COVID-19 pandemic, an unexpected event in the 21st century, was an intervention event that impacted the normal provision of healthcare services. This study applied time series models with intervention and interrupted time series analysis to assess how the COVID-19 pandemic affected hospitalization services. Forecast scenarios were created to compare the observed reality with what would have been expected without the pandemic. The results demonstrate that the inclusion of time series methodologies with interventions and interruptions significantly improves the accuracy of forecasts, highlighting the importance of these methods in effective healthcare sector planning during crises.