Capacidad potencial de un centro coordinador de urgencias y emergencias para predecir ingresos hospitalarios y en unidades de cuidados intensivos por COVID-19

Abstract

Objective. To analyze the association between the perceived care demand in the emergency call center of Castilla La Mancha (and hospital and ICU admissions for COVID-19, as well as their temporal characteristics, to explore its potential capacity as a predictive tool for COVID hospital admissions. Methodology. Retrospective observational study on the daily calls made to the emergency call center of Castilla La Mancha, both calls to 112 and those made to COVID line, in the period between March 1 and October 14, 2020. The data were analyzed by codes “diarrhea”, “dyspnea”, “fever” and “general discomfort” that were used as predictor variables, and their relationship with hospital admissions and ICU admissions. Results. A total of 831,943 calls were received at the CLM emergency call center through 112, with a maximum on March 13, 2020 with 10,582 calls. On COVID line, a total of 208,803 calls were received in that period, with a maximum on March 15 with 23,744. A statistically significant relationship was found between the regulation codes studied (specific symptoms) and the number of calls with hospital admissions and ICU admissions, with a predictive capacity of 2 weeks in relation to occupancy peaks. The codes with the greatest relationship were “general malaise” and “diarrhea”. Conclusion. We have found an association between the number of calls to a CCUE due to dyspnea, fever, general discomfort, diarrhea and the number of calls with hospital admissions and ICU for COVID-SARS-2 2 weeks in advance, mainly due to general discomfort and diarrhea. The design of predictive expert systems and their automation using artificial intelligence could be part of the preparation, planning and anticipation programs of health systems in the near future in the event of future pandemics..