Hemodynamic changes in anemic premature infants: are we allowing the hematocrits to fall too low?

Currently, many nurseries allow hematocrits to fall to <21% in apparently "stable" premature infants before considering a blood transfusion. We evaluated clinical changes and hemodynamic changes by echocardiogram in "stable" anemic premature infants before, during, and after transfusion.

"Stable" premature infants (< or =32 weeks' gestation) who were to receive transfusions (2 aliquots of 10 mL/kg packed red blood cells, 12 hours apart) were eligible for prospective enrollment. Cardiac function by echocardiography and vital signs were measured 4 times: 1 to 3 hours before and 2 to 4 hours after the initial aliquot and 4 to 7 hours and 27 to 34 hours after the second aliquot. Infants were grouped prospectively according to pretransfusion hematocrit ranges for analysis: < or =21% (low), 22% to 26% (mid), and > or =27% (high).

Thirty-two infants were enrolled. No differences were observed between the groups in sex, birth weight, postconceptional age, or postnatal weight at enrollment. Before transfusion, low- and mid-range groups had higher left ventricular end systolic and diastolic diameters, in comparison with high range. Low range had increased stroke volume in comparison with the high-range group. These changes persisted after transfusion. Mean diastolic blood pressure rose and peak velocity in the aorta fell in the low-range group after transfusion. Pretransfusion hematocrit was correlated with but poorly predictive of echocardiographic measurements. Infants with inappropriate weight gain had increased ventricular end diastolic diameters, consistent with congestive heart failure.

Apparently "stable" anemic premature infants may be in a clinically unrecognized high cardiac output state, and some echocardiographic measurements do not improve within 48 hours after transfusion. The benefits of transfusion practices guided by measures of cardiac function should be evaluated.