Blood transfusion is associated with prolonged duration of mechanical ventilation in infants undergoing reparative cardiac surgery

Abstract

Objective: Perioperative transfusion has adverse effects in adults undergoing cardiac surgery. We sought to investigate whether greater use of blood and blood products might be an independent predictor of prolonged postoperative recovery, indicated by duration of mechanical ventilation (DMV), after reparative infant heart surgery.

Design: Secondary analysis of prospectively collected data from two randomized trials of hematocrit strategy during cardiopulmonary bypass in infant heart surgery to explore the association of DMV with perioperative transfusion and other variables.

Setting: Tertiary pediatric hospital.

Patients: Two hundred seventy infants undergoing two ventricle corrective cardiac surgery without aortic arch reconstruction.

Measurements and main results: In univariable analyses, longer DMV was associated with younger age and lower weight at surgery, diagnostic group, and higher intraoperative and postoperative blood product transfusion (each p < .001). In multivariable proportional hazard regression, longer total support time and greater intraoperative and early postoperative blood products per kg were the strongest predictors of longer DMV. Patients in the highest tertile of intraoperative blood products per kg had an instantaneous risk of being extubated approximately half that of patients in the lowest tertile (hazard ratio, 0.51; 95% confidence interval, 0.35, 0.73). Patients who received any blood products on postoperative day 1, compared with those who did not, had a hazard ratio for extubation of 0.65 (95% confidence interval, 0.50, 0.85).

Conclusions: In this exploratory secondary analysis of infants undergoing two ventricular repair of congenital heart disease without aortic arch obstruction, greater intraoperative and early postoperative blood transfusion emerged as potential important risk factors for longer DMV. Future prospective clinical trials are needed to determine whether reduction in blood product administration hastens postoperative recovery after infant heart surgery.

Trial registration: ClinicalTrials.gov NCT00006183.

FIGURE 1

Percent of patients ventilated over time by intraoperative blood product tertile, estimated through Kaplan-Meier survival curves. P-value comparing intraoperative blood product tertiles was based on the Kruskal-Wallis test.