514

Abstract

Introduction: Extracorporeal cardiopulmonary resuscitation (eCPR) has demonstrated success as a rescue therapy in refractory cardiac arrest among patients with congenital heart disease. Outcome predictors among eCPR-supported patients with structurally normal hearts are less well defined. Methods: Extracorporeal Life Support Organization (ELSO) registry data was retrospectively reviewed from January 1998 through December 2011. Patients with an ICD-9 diagnosis code consistent with structural heart disease were excluded. The remaining subjects were categorized by ICD-9 codes into non-mutually exclusive diagnostic subcategories. Univariate predictors significant for an increase in mortality were included in a multivariate logistic regression model. Results: During the study period, 1431 patients met inclusion criteria. Median age was 16 years (2.4, 51 years) and weight was 50 kg (12.5, 71 kg). Subjects were 60% male. Venoarterial support was instituted in 97% of cases, with 86% categorized as a percutaneous approach. Common diagnostic groups included atraumatic acute respiratory failure (15%, n=219), cardiomyopathy (13%, n=189), sepsis (8%, n=119), and myocarditis (8%, n=113). Overall survival to hospital discharge was 32%. Multivariate logistic regression (Goodness-of-Fit p=0.52) demonstrated an independent survival benefit among younger patients (p=0.04) and those with a shorter duration from intubation to cannulation (p=0.02). A diagnosis of sepsis (odds ratio [OR] 3.0, 95% confidence interval [CI] 1.5-5.9) was independently associated with an increase in odds of mortality, while the diagnosis of myocarditis was protective (OR 0.31, CI 0.18-0.54). While supported with ECMO, neurologic complications (OR 6.2, CI 4.2-9.3), pulmonary hemorrhage (OR 2.0, CI 1.1-3.8), serum pH <7.2 (OR 2.3, CI 1.5-3.7), hyperbilirubinemia (OR 2.2, CI 1.1-4.3), and CPR (OR 2.3, CI 1.4-4.0) were all independently associated with an increased odds of mortality. Conclusions: Among patients with structurally normal hearts receiving eCPR, nearly one-third survived to hospital discharge. Defining independent pericannulation predictors of mortality may assist in identifying populations of patients most likely to benefit from this resource-intensive rescue therapy.