Antecedentes: el sistema de salud peruano se encuentra en riesgo durante periodos de aumento de casos de COVID-19. Las herramientas de triaje ayudan a identificar a pacientes con riesgo de muerte o complicaciones para tomar mejores decisiones clínicas y optimizar recursos. Objetivo: validar el score PRIEST para la predicción de eventos adversos en pacientes hospitalizados con diagnóstico de COVID-19 en el Hospital Cayetano Heredia en Lima-Perú. Materiales y métodos: estudio de validación de cohorte retrospectiva. Se recopilaron datos demográficos y clínicos de 336 pacientes con diagnóstico de COVID-19 entre abril 2020 y enero 2021. Se calculó el puntaje PRIEST de cada participante, se construyó una curva ROC y se determinó el AUC para el desenlace primario combinado (fallecimiento o ingreso a UCI), y para los desenlaces secundarios independientes de fallecimiento e ingreso a UCI. Se realizó un análisis bivariado y multivariado con regresión logística de las variables del score PRIEST. Resultados: 195 (58.0%) participantes presentaron el desenlace primario combinado, 88 (26.2%) fallecieron y 138 (41.4%) ingresaron a UCI. El score PRIEST no mostró capacidad discriminativa para el desenlace primario combinado: AUC 0.556 (0.494-0.618, CI 95%), pero si mostró una capacidad discriminativa aceptable para el desenlace secundario independiente de fallecimiento: AUC 0.738 (0.682-0.793, CI 95%). Conclusiones: el score PRIEST no presenta buena capacidad discriminativa para el desenlace combinado de fallecimiento o ingreso a UCI, pero sí presenta una capacidad discriminativa aceptable para el desenlace independiente de fallecimiento.
Background: The Peruvian healthcare system is at risk during periods of increased COVID-19 contagion. Triage tools help identify patients at risk of death or complications, enabling better clinical decision-making and resource optimization. Objective: to validate the PRIEST score for predicting adverse events in hospitalized patients diagnosed with COVID-19 at Cayetano Heredia Hospital in Lima-Peru. Materials and Methods: a retrospective validation cohort study was conducted. Demographic and clinical data were collected from 336 patients hospitalized with COVID-19 between April 2020 and January 2021. The PRIEST score was calculated for each participant, and a ROC curve was constructed. The area under the curve (AUC) was calculated for the primary outcome of combined adverse events (death or ICU admission), as well as the secondary outcomes of death and ICU admission. Bivariate and multivariate analyses were performed using logistic regression of the PRIEST score variables. Results: 195 (58.0%) participants experienced the primary outcome, with 88 (26.2%) deaths and 138 (41.4%) ICU admissions. The PRIEST score did not show discriminative capacity for the primary outcome: AUC 0.556 (0.494-0.618, 95% CI). However, it did demonstrate acceptable discriminative capacity for the outcome of death: AUC 0.738 (0.682-0.793, 95% CI). Conclusions: The PRIEST score does not exhibit good discriminative capacity for the combined outcome of death or ICU admission, but it does demonstrate good discriminative capacity for the outcome of death.
