The accuracy of calculated base excess in blood.

Most equations used for calculation of the base excess (BE, mmol/l) in human blood are based on the fundamental equation derived by Siggaard-Andersen and called the Van Slyke equation: BE = Z x [[cHCO3-(P) - C7.4 HCO3-(P)] + beta x (pH -7.4)]. In simple approximation, where Z is a constant which depends only on total hemoglobin concentration (cHb, g/dl) in blood, three equations were tested: the ones proposed by Siggaard-Andersen (SA), the National Committee for Clinical Laboratory Standards (NCCLS) or Zander (ZA). They differ only slightly in the solubility factor for carbon dioxide (alphaCO2, mmol/l x mmHg) and in the apparent pK(pK'), but more significantly in the plasma bicarbonate concentration at reference pH (C7.4HCO3-(P), mmol/l) and in beta, the slope of the CO2-buffer line (mmol/l) for whole blood. Furthermore, the approximation was improved either by variation in Z (r(c)), or in the apparent pK (pK) with changing pH. Thus, from a total of seven equations and from a reference set for pH, pCO2 and BE taken from the literature (n=148), the base excess was calculated. Over the whole range of base excess (-30 to +30 mmol/l) and PCO2 (12 to 96 mmHg), mean accuracy (deltaBE, mmol/l) was greatest in the simple equation according to Zander and decreased in the following order: +/-0.86 (ZA); +/-0.94 (ZA, r(c)); +/-0.96 (SA, r(c)); +/-1.03 (NCCLS, r(c)); +/-1.40 (NCCLS); +/-1.48 (SA); and +/-1.50 (pK'). For all clinical purposes, the Van Slyke equation according to Zander is the best choice and can be recommended in the following form: BE= (1 -0.0143 x cHb) x [[0.0304 x PCO2 x 10pH-6.1-24.26] + (9.5+1.63 x cHb) x (pH -7.4)] - 0.2 x cHb x (1-sO2), where the last term is a correction for oxygen saturation (sO2). Hence, base excess can be obtained with high accuracy (<1 mmol/l) from the measured quantities of pH, pCO2, cHb, and SO2 in any sample, irrespective of whether venous or arterial blood is used.