Background: Earlier work on adults undergoing surgery with cardiopulmonary bypass suggests that there is a close relationship between the lower limit of the cerebral and renal autoregulation pressures. Although cerebral autoregulation during bypass in infants has been extensively investigated, the impact of bypass on kidney function is not well known. It is, nevertheless, acknowledged that the main pathophysiological process involved in cardiac surgery-related kidney damage is tubular injury, and that urine neutrophil gelatinase-associated lipocaline (uNGAL) is a reliable biomarker of injury.
Objective: To identify the most predictive bypass variable for the measurement of renal injury, its threshold value and the most predictive time below that threshold.
Design: Observational study linking electronically recorded bypass perfusion pressure and oxygen delivery rate with intra-operative uNGAL excretion. Variations in bypass variables were accounted for by their excursions below several thresholds.
Setting: French tertiary referral paediatric cardiac centre.
Patients: A total of 72 infants in whom uNGAL was measured within 1 h of bypass.
Main outcome measure: Renal injury, identified by a high creatinine normalised uNGAL concentration (>21.2 μg mmol).
Results: At the end of bypass, 43.75% of infants had high uNGAL. A more than 40% pressure drop below the normal age-standardised mean arterial pressure was associated with high uNGAL. Receiver operating curve [interquartile range] areas were 0.626 [0.501 to 0.752] for a more than 40% drop, and 0.679 [0.555 to 0.804] for a more than 50% drop. A more than 40% pressure drop for 19.5 min provided a 0.65 negative predictive value for high uNGAL, and a more than 50% pressure drop for 5.4 min provided a 0.67 negative predictive value. The link between uNGAL and oxygen delivery rate was negligible.
Conclusion: Maintaining the perfusion pressure above 60% of the normal age-standardised mean arterial pressure may provide an effective renal protective strategy.
Trial registration: Registered on October 11, 2010, ClinicalTrials.gov Identifier: NCT01219998.