Background and aims The incidence of AKI in critically ill neonates is estimated at 6–24%with 10–61% mortality. Whilst the incidence of AKI in neonates undergoing ECMO is unknown, its presence independently predicts mortality. We audited the incidence of AKI in neonates on ECMO in our centre against the published incidence of AKI in a similar cohort of neonates undergoing congenital cardiac disease surgery with cardiac bypass.

Methods All neonates who underwent ECMO due to respiratory disease in one year were included (n = 24). The case notes, fluid balance charts and laboratory data were reviewed. AKI was graded based on published RIFLE criteria.

Results Twenty five percent of neonates developed AKI; 1 (4%), 2 (8%) and 3 (13%) were graded as “Risk”, “Injury” and “Failure” respectively based on creatinine rise alone. If reduced urine output and rise in creatinine were used the number of infants with AKI was 9 (38%). The number of neonates with “Risk”, “Injury” and “Failure” was 3 (13%), 2 (8%) and 4 (17%) respectively.

Conclusions The incidence of AKI in our cohort was lower than the published cohort used as the audit standard. Nevertheless, there was a higher proportion of more severe AKI in our cohort. This may be accounted for by the emergent rather than elective admissions of our cohort. Recent data suggests a threshold of <0.5 mL/kg/hr is to low for the neonatal population, thus our data may represent an underestimate. Nevertheless this data illustrates the need for close monitoring of renal function and urine output.