The ‘optimal’ hematocrit to which patients should be resuscitated after shock and trauma is controversial. To test the hypothesis that sufficient oxygen delivery can be provided at a lower hematocrit without impairing oxygen consumption or hemodynamic function, 25 patients were prospectively studied immediately following injury and/or acute hemorrhage. Patients were randomized to have their hematacrits (HCT) maintained near 30%. (29.7 ± 0.4% (M ± SEM); n = 12) or 40% (38.4 ± 0.6%, n = 13). Cardiopulmonary parameters were measured twice a day for 3 days. Statistical analysis used a repeated measures analysis of variance with patient age, and ventilator parameters (FIO2, PEEP, and ventilator mode) as covariates. Arterial and venous O2 saturations were not significantly different at different hematocrits, although arterial and venous O2 contents were lower at 30% HCT (a = 14.1 ± 0.2 mlO2/dl, v = 10.1 ± 0.3 mlO2/dl; vs. a = 17.4 ± 0.4 mlO2/dl, v = 13.6 ± 0.6 mlO2/dl; p < 0.05). This resulted in a lower oxygen delivery at the lower HCT. Between the two groups, there also was no significant difference in cardiac index (overall mean, 3.64 ± 0.16 ml/min/m2]), heart rate (99 ± 4 bpm), systemic vascular resistance (1,058 ± 55 dyne-sec/cm5]), or left ventricular stroke work index (4.3 ± 0.3 x 106] dyne-cm/m2]). Intrapulmonary shunt was higher with higher hematocrit (22.6 ± 2.4% at 40% HCT vs. 14.6 ± 1.6% at 30% HCT; p < 0.05) with no difference in end-expiratory pressure.
Our data demonstrate that the determinant of O2 delivery was the hemoglobin concentration. Within the 30% HCT group, oxygen consumption was correlated with oxygen delivery; in the 40% HCT group, there was no similar correlation. However, in both groups tissue oxygenation and perfusion were adequate since the O2 extraction was always less than 30%. In the postinjury state, we found there is no advantage to raising the hematocrit above 30%. In addition, the potential increase in intrapulmonary shunt with a higher hematocrit may be detrimental.