Bedside hemoglobin measurements: sensitivity to changes in serum protein and electrolytes

Our objective was to compare the effect of protein and electrolyte changes associated with hemodilution on the accuracy of photometric and conductivity hemoglobin determination methods.

Blood samples from 10 patients with normal preoperative serum electrolytes and total protein levels were studied. From an indwelling arterial line, 20 ml of blood were removed; hemoglobin values were measured pre-(Baseline) and postdilution by Coulter counter, conductivity, and photometric methods. Blood samples were diluted by placing 4 ml of blood into three test tubes, and adding 1 ml of either 25% albumin, 0.9% sodium chloride, or 5% dextrose in water.

Blood sample dilution resulted in a reported conductivity hemoglobin that was significantly different from the Coulter value (p = 0.0004) when 25% albumin, 0.9% sodium chloride, and 5% dextrose in water solution was used. Using the same dilutions, the photometric method accurately reflected Coulter hemoglobin values. The correlation between photometric and Coulter hemoglobin measurements was R2 = 0.97, p = 0.0001. Correcting the conductivity hemoglobin values for changes in total protein, chloride and sodium significantly improved correlation with Coulter hemoglobin (R2 of uncorrected versus corrected = 0.37 and 0.72; p = 0.0001).

In the range of electrolyte and protein concentrations found in this study, the photometric method of hemoglobin assessment was more accurate than either corrected or uncorrected conductivity hemoglobin determinations, as compared to Coulter-based measurements.