Hematocrit Levels, Blood Testing, and Blood Transfusion in Infants After Heart Surgery

Abstract

Objective: To determine whether judicious blood testing impacts timing or amount of packed RBC transfusions in infants after heart surgery.

Design: A retrospective study comparing before and after initiation of a quality improvement process.

Setting: A university-affiliated cardiac ICU at a tertiary care children's hospital.

Patients: Infants less than 1 year old with Risk Adjustment for Congenital Heart Surgery category 4, 5, 6, or d-transposition of great arteries (Risk Adjustment for Congenital Heart Surgery 3) consecutively treated during 2010 through 2013.

Intervention: A quality improvement process implemented in 2011 to decrease routine laboratory testing after surgery.

Measurements and main results: Fifty-two infants preintervention and 214 postintervention had similar age, weight, proportion of cyanotic lesions, and surgical complexity. Infants with single versus biventricular physiology were compared separately. The number of laboratory tests per patient adjusted for cardiac ICU length of stay (laboratory tests/patient/day) was significantly lower in postintervention populations for single and biventricular groups (9 vs 15 and 10 vs 15, respectively; p < 0.001). The proportion of single ventricle patients transfused post- and preintervention was not statistically different (72% vs 90%; p = 0.130). Transfusion in the biventricular groups was the same over time (65% vs 65%). Time to first transfusion was significantly longer in the postintervention single ventricle group (4 vs 1 d; p < 0.001), and was not statistically different in the biventricular patients (4 vs 7 d; p = 0.058). The median hematocrit level at first transfusion was significantly lower (37% vs 40%; p = 0.004) postintervention in the cyanotic population, but did not differ in the biventricular group (31% vs 31%; p = 0.840).

Conclusion: In infants after heart surgery, blood testing targeted to individual needs significantly decreased the number of blood tests, but did not significantly decrease postoperative blood transfusion.