Central nervous system complications during pediatric extracorporeal life support: incidence and risk factors

Abstract

Objective: Identify the incidence and risk factors for development of acute, severe central nervous system (CNS) complications of pediatric extracorporeal life support (ECLS).

Design: Retrospective review of Extracorporeal Life Support Organization (ELSO) registry database.

Setting: Pediatric intensive care units of 115 tertiary centers internationally.

Patients: Pediatric patients, 1 month to 18 yrs of age, who had ECLS between the years 1981-2002.

Measurements and main results: Data concerning 4,942 patients who underwent one run of ECLS were analyzed. Six hundred thirty-six patients (12.9%) developed acute, severe CNS complications. Patients who required ECLS during extracorporeal cardiopulmonary resuscitation (n = 161; 3.3%) were more likely to develop CNS complications (n = 42; 26.1%) than patients who did not have extracorporeal cardiopulmonary resuscitation (p < .001; odds ratio [OR], 2.48; 95% confidence interval [CI], 1.73-3.57). Stepwise logistic regression analysis of therapies patients received before initiation of ECLS showed that the use of a left ventricular assist device (p = .001; OR, 3.45; 95% CI, 1.64-7.22), bicarbonate (p < .001; OR, 1.61; 95% CI, 1.26-2.05), and vasopressor/inotropic medications (p = .035; OR, 1.22; 95% CI, 1.01-1.48) were significant independent predictors of development of CNS complications. Among patients who had pulmonary failure as an indication for ECLS, the CNS complication rate was significantly higher for those treated with venoarterial ECLS than those who had venovenous ECLS (13.5% vs. 5.7%; p < .001; OR, 0.43; 95% CI, 0.34-0.67). Multiple logistic regression analysis of the complications other than CNS complications associated with the use of ECLS showed that pH <7.20, creatinine concentration >3.0 mg/dL, use of inotropes, presence of myocardial stun, and requirement of cardiopulmonary resuscitation during ECLS independently predicted development of CNS complications.

Conclusion: Patients who have metabolic acidosis, a bicarbonate or inotrope/vasopressor requirement, cardiopulmonary resuscitation, or a left ventricular assist device before initiation of ECLS are at greater risk for development of CNS complications. After initiation of ECLS, patients who develop renal failure or metabolic acidosis or undergo venoarterial ECLS should be closely monitored for development of CNS complications.