Development of a two-buffer model which simulates the acid-base properties of blood and allows comparison of the different acidbase concepts according to Stewart and to Siggaard-Andersen. The two-buffer model consisted of different aqueous solutions of bicarbonate/CO(2) (pCO(2), sCO(2), pK(1)), HEPES buffer (A(tot), pK(a)) and electrolytes. These were used to calculate the pH from the independent variables according to Stewart - strong ion difference (SID), pCO(2) and total concentration of the weak acids (A(tot)) - from which all other dependent variables (cHCO(3)(-), cA(-), BB, BE) were obtained and compared with the measured values. The normal pH (7.408) was calculated from the normal values for SID (48 mmol/l), pCO(2) (40 mmHg) and A(tot) (45.2 mmol/l) and agreed perfectly with the measured value (7.409+/-0.001). This was also valid for all calculated and measured pH values when the SID was varied: non-respiratory alkalosis ( upward arrow) or acidosis ( downward arrow), pCO(2):respiratory acidosis ( upward arrow) or alkalosis ( downward arrow) and A(tot):hyperproteinemic acidosis ( upward arrow) or hypoproteinemic alkalosis ( downward arrow) were varied and the sum of the buffer bases (BB) was always equal to the SID. All changes and hence BE were also equal, providing that A(tot) was normal. This was not the case, however, if A(tot) was outside the normal range, when BE was then the difference from the normal BB at the respective reference point. Whereas the deviation of the measured pCO(2) was acceptable (1.74+/-0.86 mmHg), this was not the case for the SID (-6.18+/-3.58 mmol/l) calculated from the measured ion concentrations (Na, K, Ca, Cl). Despite controversial discussions, both concepts are much closer than might be expected. Whereas in the Stewart approach the focus of analysis is on plasma, with the Siggaard-Andersen approach it is on blood. Hence, a combined analysis of the blood gases (pH, pCO(2), pO(2), sO(2), cHb, BE) and of the strong ion gap (SIG) may be useful.
Read full abstract