Clinical risk factors for adverse outcomes in emergency department patients with acute decompensated heart failure: An ADHERE registry analysis

Abstract

Study objectives: Acute decompensated heart failure (ADHF) is a leading cause of emergency department (ED) visits, subsequent inpatient admission, and substantial inhospital morbidity. Although retrospective data analyses have suggested that several clinical variables that can be assessed while a patient is in the ED are associated with inpatient adverse events, emergency physicians have no standard method of risk stratification for patients with ADHF. Methods: ED data from 65,180 patients in the Acute Decompensated Heart Failure National Registry (ADHERE) were analyzed. ADHERE is an observational registry of patients hospitalized with ADHF and in which more than 250 community, tertiary, and academic medical centers participate. Patients were included if they were admitted to an inpatient bed from the ED, had an initial systolic blood pressure of greater than 90 mm Hg, and did not have positive cardiac markers. Adverse events were defined as any 1 or more of the following: inpatient death, procedure (coronary artery bypass graft surgery, percutaneous coronary intervention, new dialysis, intra-aortic balloon catheter, cardiac transplantation, cardioversion, defibrillation, cardiopulmonary resuscitation, pacemaker, ventricular assist device, or mechanical ventilation), transfer to an ICU after initial admission to a floor bed (ie, condition worsened), or length of stay greater than third quartile. Details of medical management and outcomes were collected through discharge record review. Thirty dichotomous explanatory variables were used; these variables were available to the treating emergency physician and selected from a previous Classification and Regression Tree (CART) analysis. Stepwise logistic regression was used to select variables, and odds ratio estimates (with 95% confidence interval [CI]) were calculated. Results: Of 45,651 patients who met inclusion criteria, 14,031 (31%) had at least 1 adverse event (death 3%, procedure 11%, ICU transfer 4%, LOS >third quartile 24%). Of 10 significant ED variables selected by the model, 4 had clinically useful odds ratio (OR >1.40) estimates with acceptably narrow 95% CI: blood urea nitrogen level greater than 40 mg/dL (1.77; 95% CI 1.70 to 1.86 mg/dL), sodium level less than 130 mEq/L (1.49; 95% CI 1.36 to 1.64 mEq/L), systolic blood pressure less than 110 mm Hg (1.47; 95% CI 1.38 to 1.56 mm Hg), and pulse rate greater than 100 beats/min (1.44; 95% CI 1.38 to 1.51 beats/min). The other 6 variables were race (white), age (≥65 years), dyspnea at rest, peripheral vascular disease, chronic obstructive pulmonary disease, and fatigue. Conclusion: Clinical variables that are available to emergency physicians treating ADHF patients in the ED may be used to identify patients at risk for subsequent inpatient adverse events. Early, aggressive treatment should be considered for patients fulfilling 1 or more of these criteria. Study objectives: Acute decompensated heart failure (ADHF) is a leading cause of emergency department (ED) visits, subsequent inpatient admission, and substantial inhospital morbidity. Although retrospective data analyses have suggested that several clinical variables that can be assessed while a patient is in the ED are associated with inpatient adverse events, emergency physicians have no standard method of risk stratification for patients with ADHF. Methods: ED data from 65,180 patients in the Acute Decompensated Heart Failure National Registry (ADHERE) were analyzed. ADHERE is an observational registry of patients hospitalized with ADHF and in which more than 250 community, tertiary, and academic medical centers participate. Patients were included if they were admitted to an inpatient bed from the ED, had an initial systolic blood pressure of greater than 90 mm Hg, and did not have positive cardiac markers. Adverse events were defined as any 1 or more of the following: inpatient death, procedure (coronary artery bypass graft surgery, percutaneous coronary intervention, new dialysis, intra-aortic balloon catheter, cardiac transplantation, cardioversion, defibrillation, cardiopulmonary resuscitation, pacemaker, ventricular assist device, or mechanical ventilation), transfer to an ICU after initial admission to a floor bed (ie, condition worsened), or length of stay greater than third quartile. Details of medical management and outcomes were collected through discharge record review. Thirty dichotomous explanatory variables were used; these variables were available to the treating emergency physician and selected from a previous Classification and Regression Tree (CART) analysis. Stepwise logistic regression was used to select variables, and odds ratio estimates (with 95% confidence interval [CI]) were calculated. Results: Of 45,651 patients who met inclusion criteria, 14,031 (31%) had at least 1 adverse event (death 3%, procedure 11%, ICU transfer 4%, LOS >third quartile 24%). Of 10 significant ED variables selected by the model, 4 had clinically useful odds ratio (OR >1.40) estimates with acceptably narrow 95% CI: blood urea nitrogen level greater than 40 mg/dL (1.77; 95% CI 1.70 to 1.86 mg/dL), sodium level less than 130 mEq/L (1.49; 95% CI 1.36 to 1.64 mEq/L), systolic blood pressure less than 110 mm Hg (1.47; 95% CI 1.38 to 1.56 mm Hg), and pulse rate greater than 100 beats/min (1.44; 95% CI 1.38 to 1.51 beats/min). The other 6 variables were race (white), age (≥65 years), dyspnea at rest, peripheral vascular disease, chronic obstructive pulmonary disease, and fatigue. Conclusion: Clinical variables that are available to emergency physicians treating ADHF patients in the ED may be used to identify patients at risk for subsequent inpatient adverse events. Early, aggressive treatment should be considered for patients fulfilling 1 or more of these criteria.