Long term outcome, late mortality and causes of late death in 741 ECMO patients treated at a UK centre since 1992

Abstract

Introduction ECMO is a resource intense treatment offered to the sickest patients with significant short term mortality for certain diagnoses. We aimed to investigate longer term survival rates and late deaths. Methods This study was registered as an audit. A list of all patients receiving ECMO support since the service was established in 1992 was downloaded from the local ECMO database. The UK National Health Service number was used to trace patients9 survival status via the UK Care Record Service, which routinely provides this information for hospital information services. Patients were classified according to diagnosis. Late death was defined as death >90 days after ECMO, since many patients were discharged to other intensive care units. Medical records of all 48 ‘late death’ patients were accessed to establish the cause of death. Results Included were 741 children of which 274 (36.9%) died early, and 48 (6.5%) died late, >90 days after ECMO. Overall 5 year actuarial survival rates were: meconium (MAS) (n=122) 87%, congenital diaphragmatic hernia (CDH) (n=49) 42%, other neonatal (n=120) 52%, paediatric viral (n=68) 67%, paediatric bacterial (n=54) 61%, other paediatric respiratory (n=38) 54%, congenital heart disease (CDH) (n=169) 32% and cardiomyopathy/myocarditis (CM) (n=121) 57%. Late deaths occurred in all groups (1.8% up to 14.3%) depending on diagnosis. The odds (OR) of late death was highest for: CDH 14.3% (n=8) OR 6.6 (95% CI 1.6, 26.7) (p Conclusion Although the majority of deaths were early, a late attrition was observed in ECMO patients. Late deaths were more prevalent in complex children related to underlying long term conditions such as CDH, prematurity, cardiac transplant and congenital heart disease.