Prospective crossover study of the effect of phlebotomy and intravenous crystalloid on hematocrit.

OBJECTIVE To compare the changes in hematocrit (Hct) between phlebotomized and nonphlebotomized individuals given IV crystalloid. METHODS A prospective, crossover volunteer study was performed comparing Hct changes immediately and 30 minutes after IV crystalloid bolus in 20 healthy adults with and without prebolus phlebotomy. In the control portion, volunteers were given a 15-mL/kg bolus of normal saline over 30 minutes with Hct determination before (H1), immediately after (H2), and 30 minutes after (H3) crystalloid infusion. At least 7 days later, the same subjects were phlebotomized 1 unit of blood and then administered a 15-mL/kg IV bolus of normal saline 30 minutes later. Hcts were obtained before (H4) and 30 minutes after (H5) phlebotomy (immediately prior to crystalloid infusion). Hcts were also obtained immediately after (H6) and 30 minutes after (H7) crystalloid infusion. A post-hoc test performance analysis was then performed to determine the Hct drop thresholds that would yield the maximal sensitivity and specificity for 500 mL of blood loss (via phlebotomy) in this population. RESULTS The Hct (%) drops in the nonphlebotomized individuals receiving IV fluids averaged 4.5 +/- 1.3 immediately and 3.2 +/- 1.3 30 minutes after infusion. These drops were different (p < 0.05) from the Hct drop in individuals receiving IV fluids after phlebotomy, which averaged 6.6 +/- 1.5 and 5.7 +/- 1.1, respectively. Post-hoc analysis revealed that Hct drops of 5.4 immediately, or 4.3 at 30 minutes after infusion, had a sensitivity of > 90% and a specificity of 75% for identification of patients in the phlebotomy group. CONCLUSIONS The practice of measuring serial Hcts may be helpful to identify trauma patients with occult blood loss. A prospective clinical trial is needed to validate these Hct drop thresholds (immediate and 30 minutes postinfusion) in crystalloid-resuscitated trauma patients.