BACKGROUND: Point-of-care (POC) devices measuring hematocrit rely on determination of electrical conductivity of whole blood. We hypothesized that some frequently administered IV fluids independently alter blood conductivity and confound hematocrit determination. METHODS: Whole human blood was diluted to predetermined hematocrit values with normal saline, lactated Ringer solution, hetastarch, or plasma. Electrical conductivity and hematocrit (i-STAT® and spun methods) were measured at each dilution. In separate experiments, the effects of propofol and heparin were noted on these variables. RESULTS: Greater dilution significantly increased conductivity irrespective of diluent type. The magnitude of the conductivity slopes increased in order for plasma, hetastarch, lactated Ringer solution, and normal saline dilution. Moreover, each slope varied from every other slope (all P < 0.0001), and 94.2% of hematocrit values measured by i-STAT (n = 211 of 224) were less than those for the spun method. Dilution with plasma, normal saline, lactated Ringer solution, and hetastarch caused bias (Bland-Altman limits of agreement) of -2.7% (-6.9/1.4), -4.6% (-7.3/-2.0), -4.8% (-7.8/-1.7), and -2.0% (-5.6/1.9), respectively. The Cohen κ agreement values (5th-95th confidence interval) for a transfusion trigger of 30% were 0.90 (all values, 0.85-0.95), 0.25 (hematocrit <30%, 0.02-0.48), and 0.21 (hematocrit 18%-30%, 0.01-0.42). Clinically relevant concentrations of propofol and heparin had minimal effects on electrical conductivity or hematocrit determination. CONCLUSIONS: Dilution of blood with frequently used IV solutions affects whole blood conductivity determinations and thereby decreases hematocrits measured by a POC device relying on this method as compared with spun hematocrit. Conductivity-based hematocrit POC devices should be cautiously interpreted when hemodilution is present.
BACKGROUND: Point-of-care (POC) devices measuring hematocrit rely on determination of electrical conductivity of whole blood. We hypothesized that some frequently administered IV fluids independently alter blood conductivity and confound hematocrit determination. METHODS: Whole human blood was diluted to predetermined hematocrit values with normal saline, lactated Ringer solution, hetastarch, or plasma. Electrical conductivity and hematocrit (i-STAT® and spun methods) were measured at each dilution. In separate experiments, the effects of propofol and heparin were noted on these variables. RESULTS: Greater dilution significantly increased conductivity irrespective of diluent type. The magnitude of the conductivity slopes increased in order for plasma, hetastarch, lactated Ringer solution, and normal saline dilution. Moreover, each slope varied from every other slope (all P < 0.0001), and 94.2% of hematocrit values measured by i-STAT (n = 211 of 224) were less than those for the spun method. Dilution with plasma, normal saline, lactated Ringer solution, and hetastarch caused bias (Bland-Altman limits of agreement) of -2.7% (-6.9/1.4), -4.6% (-7.3/-2.0), -4.8% (-7.8/-1.7), and -2.0% (-5.6/1.9), respectively. The Cohen κ agreement values (5th-95th confidence interval) for a transfusion trigger of 30% were 0.90 (all values, 0.85-0.95), 0.25 (hematocrit <30%, 0.02-0.48), and 0.21 (hematocrit 18%-30%, 0.01-0.42). Clinically relevant concentrations of propofol and heparin had minimal effects on electrical conductivity or hematocrit determination. CONCLUSIONS: Dilution of blood with frequently used IV solutions affects whole blood conductivity determinations and thereby decreases hematocrits measured by a POC device relying on this method as compared with spun hematocrit. Conductivity-based hematocrit POC devices should be cautiously interpreted when hemodilution is present.
Authors: Michael Hawkes; Andrea L Conroy; Robert O Opoka; Sophie Namasopo; W Conrad Liles; Chandy C John; Kevin C Kain Journal: Am J Trop Med Hyg Date: 2014-03-03 Impact factor: 2.345