Importance: There are no well-established noninvasive biomarkers for identifying patients at risk for poor outcome after surgery for congenital heart disease. Few studies have assessed prognostic accuracy of cerebral tissue oxygenation index (cTOI) measured by near infrared spectroscopy (NIRS).
Objective: To assess the utility of noninvasive NIRS monitoring as a predictor of outcomes after neonatal cardiac surgery through measurement of cTOI. To examine the utility of noninvasive NIRS monitoring in combination with lactate concentration and inotropic score in prediction of outcomes after neonatal cardiac surgery.
Design: Prospective longitudinal cohort study.
Setting: Operating room and cardiac intensive care unit, Children's National Heart Institute.
Participants: Seventy-five patients with complex congenital heart disease undergoing surgical repair within first month of life.
Exposure: Cerebral TOI, blood lactate, and inotropic scores were measured preoperative, intraoperative and up to 24 hours postoperative.
Main outcome measures: Postoperative mortality and neurodevelopmental outcome assessed by the Bayley Scales of Infant Development (BSID II). Mental and motor scores were obtained at 6, 15, and 21 months. Good outcome was defined as survival and BSID mental and motor scores ≥70 points. Poor outcome was defined as death or BSID scores <70 at most recent follow-up.
Results: Cohort of 75 patients prospectively followed including 40 patients with single ventricle and 35 with two ventricles. Four patients died before discharge and ten died within 21 months. Seven patients were lost to follow-up. Among survivors with follow-up (n = 54), BSID was abnormal in 25 (46%). Patients with poor outcome (n = 39) had lower mean cTOI 60 minutes off-CPB (48% vs. 58%, P = .003) and 24 hours postoperative (49% vs. 59%, P < .001), higher lactate (8.2 vs. 5.0 mmol/L, P = .005) and higher inotropic scores (10 vs. 6, P = .02) at 24 hours postoperative. ROC analysis indicated that cTOI had moderate predictive accuracy of outcome (AUC = 0.751, P < .001). Multivariable regression analysis confirmed that predictive accuracy was improved using both cTOI and lactate at 24 hours postoperative (AUC = 0.813, 95% CI: 0.705-0.921, P < .001) with optimal cutoff values <58% and >7.4 mmol/L, respectively (sensitivity = 95%).
Conclusion: Cerebral TOI combined with lactate at 24 hours postoperative are accurate non-invasive predictive biomarkers of patient survival and neurodevelopmental outcome in neonates with CHD undergoing cardiac surgery.
Keywords: congenital heart disease; lactate; near infrared spectroscopy; neurodevelopmental outcome; pediatric cardiology; tissue oxygenation.
© 2016 Wiley Periodicals, Inc.