RESUMO
BACKGROUND AND AIMS: Use of extracorporeal membrane oxygenation (ECMO) in children is increasing. Yet, little is known about optimal nutritional practices in these children. We aim to describe the nutritional adequacy, factors associated with enteral nutrition, and the association between nutritional adequacy and mortality in children supported on ECMO. METHODS: We conducted a retrospective review of all children (1 month-18 years) requiring ECMO between 2010 and 2016. Data on enteral and parenteral energy and protein intake in the first 7 days of ECMO were collected. Adequacy of nutrition intake was defined as total intake vs. total requirements, expressed as a percentage. RESULTS: 51 patients were included, of which 43 (84.3%) were supported on veno-arterial ECMO. Median ECMO duration was 8.6 days [interquartile range (IQR) 6.1-16.2]. Overall energy and protein adequacy across the first 7 days of ECMO were 48.3% (IQR 28.0-67.4) and 44.8% (IQR 26.9-67.0) respectively. Parenteral nutrition provided majority of calories [median 88.0% (IQR 62.9-100)] and protein [median 91.0% (IQR 62.3-100)] intake. Enteral nutrition (EN) was initiated in 33 (64.7%) patients. Time to EN initiation, vasoactive-inotropic score just before ECMO initiation, veno-arterial ECMO mode and continuous renal replacement therapy in the first week of ECMO were factors associated with EN energy adequacy. Hospital mortality rate was 55% (28/51). Compared to survivors, non-survivors had lower adequacy of EN energy intake [0.5% (IQR 0-4.4) vs. 11.8% (IQR 0-24.5), p = 0.034]. After correcting for ECMO duration, need for continuous renal replacement therapy and number of vasoactive drugs required on ECMO, greater EN energy adequacy remained associated with lower risk of mortality [adjusted odds ratio 0.93 (95% confidence interval: 0.86-0.99), p = 0.048]. CONCLUSIONS: Nutritional adequacy, especially that of EN, remains low in children supported on ECMO. EN energy adequacy was found to be associated with lower mortality. Further studies on nutritional adequacy in pediatric ECMO, as well as strategies to optimize EN in these children, are warranted.