Estratificação de fragilidade na predição de desfechos desfavoráveis de curto prazo em idosos admitidos em um serviço hospitalar de emergência

Abstract

Introduction: Several risk factors, especially advanced age, and the presence of chronic diseases, have been identified as predictors of mortality and poor outcomes in older adults admitted to the Emergency Department (ED). However, frailty studies have shown these factors, alone, are not sufficient predictors of poor outcomes in older adults since this is an heterogeneous population. Frailty is a state of increased risk and, the degree of baseline frailty may be an individual predictor of poor outcomes more accurate than age and presence of diseases. Objective: To investigate the association between frailty and 90 and 180-day mortality after admission to the ED and to assess the ability of three frailty instruments to predict mortality and other poor outcomes in older adults admitted at a public ED. Method: Prospective cohort study, which included older adults admitted at the ED of a Public Hospital who spent at least one night in it. The degree of baseline frailty was assessed by the Clinical Frailty Scale (CFS), the Visual Scale of Frailty (VS-Frailty) and the Clinical-Functional Vulnerability Index (IVCF-20). The predictive capacity of the instruments was compared using the Receiver Operator Characteristics (ROC) curve analysis. Additionally, association of frailty degree and time to death within 180 days was analyzed. The primary outcome was 90 and 180-day mortality. The secondary outcomes were functional decline, reattendance to ED, readmission, institutionalization, need for home care and time to death within 180 days of admission. Results: 206 participants were included. Frailty prevalence ranged between 53.9 and 61.7%, depending on the frailty instrument used. Of the 61 patients (29.6%) who died by the end of follow-up, 45 deaths (21.8%) occurred within 90 days of admission to the ED. Adjusted for demographic and clinical variables, frailty remained as an independent predictor of 90 and 180-day mortality after admission. Other strong predictors were advanced-stage cancer, severe liver disease and admission to intensive care unit in last year. The CFS, compared to the VS-Frailty and the IVCF-20, showed greater accuracy by the ROC curve (AUROC) to predict mortality considering both studied periods. Frailty measured by the three instruments was associated with the need for home care and readmission, with good and regular accuracy, respectively. Accuracy of other outcomes was not significant. Conclusion: Baseline frailty measured by the CFS, VS-Frailty and IVCF-20 is a good predictor of 90 and 180-day mortality, the need for home care and readmission in older adults admitted to ED. We propose the use of these instruments to screen frailty in older adults admitted to ED, with the aim of helping in clinical decision-making. The use of frailty stratification instruments in older adults at ED may help to identify older adults who derive less benefit from receiving invasive treatment and become clinical decisions more appropriate to the baseline clinical and functional profile of these patients. It can also be a starting point for timely discussions with patients and families concerning preferences and goals of care.