Extracorporeal life support for severe respiratory failure in children with immune compromised conditions.

OBJECTIVES

To examine a large cohort of children treated with extracorporeal life support (ECLS) for severe respiratory failure to investigate the hypothesis that patients with an immune compromise condition (ICC) would have reduced survival to hospital discharge compared with patients without this classification.

DESIGN

Retrospective cohort study.

SETTING

Extracorporeal Life Support Organization (ELSO) data registry.

PATIENTS: All nonneonatal pediatric patients receiving ECLS for respiratory failure.

INTERVENTIONS: None.

MEASUREMENTS AND MAIN

RESULTS

As of February 4, 2004, the ELSO registry contained 2,879 pediatric patients between 1 month and 19 yrs of age who were treated with ECLS for respiratory failure. Overall, 183 patients had at least one International Classification of Diseases (Ninth Revision) or Current Procedural Terminology code associated with an ICC; ICC status was associated with lower hospital survival (31 vs. 57 %; p < .001). Six ICC subgroups were also examined: immune deficiency, leukemia-lymphoma, cancer, opportunistic infection, solid organ transplant, and bone marrow transplant. Each ICC subgroup was also associated with reduced hospital survival, varying from a high of 34.6% (solid organ transplants) to a low of 0% (bone marrow transplant). In a multivariate logistic regression model that controlled for factors reported to be associated with survival and other respiratory interventions (high-frequency ventilation, inhaled nitric oxide, and surfactant), the presence of an ICC remained associated with reduced hospital survival (odds ratio 0.20-0.45; p < .001). In this multivariate model, an unexpected strong association between inhaled nitric oxide therapy and lower ECLS survival was observed (odds ratio 0.49-0.80; p < .001).

CONCLUSIONS

In this cohort of pediatric patients receiving ECLS for respiratory failure, survival to hospital discharge was reduced for each ICC subgroup examined and was approximately one in three for the overall group. Further study of the association of lower survival rates for patients who received inhaled nitric oxide prior to ECLS is needed.