Application of new calculation algorithms to oxygenation parameters of cirrhotic patients. A comparison between calculated and measured data.

Oxygen status from both arterial and mixed venous blood was analyzed by reliable methods in 39 cirrhotic patients. These measured data were checked with computed oxygen parameters by new calculation algorithms. Calculated oxygen contents were higher than directly measured values but there was a highly significant correlation between them. Calculated and measured 2,3-DPG mean values were not significantly different but there was no correlation between them. A large difference was observed between measured and computed evaluation of oxygen-hemoglobin affinity without correlation between P50. In hyperkinetic patients, no correlation was observed between 'compensation factor' and the increase of cardiac index. So, in these patients the new iterative equations were not valid to determine traditional oxygen parameters from only Po2 and So2 arterial measurements. Moreover the new oxygenation parameters appeared frankly inadequate.

Is P50 the most representative P(SO2) to evaluate HbO2 affinity?

Off-line computer assistance allowed a correct visualization of the actual data included in any experimental whole oxygen-hemoglobin association curve. The affinity of hemoglobin for oxygen (characterized by the successive PO2 corresponding to any saturation value: P(SO2)), the shape of the association curve (Hill's curve and Hill's numbers) and the Bohr coefficients all along the oxygenation process, were determined with great accuracy. Results agreed with literature values obtained for successive steps of oxygenation. A batch of 78 computerized curves was divided into 3 groups (normal: NL, right deviated: RD and left deviated: LD) to which the principal component (P.C.) method was applied. It was therefore possible not only to study correlations between any curve summarized by its P.C. and any external variable but also to define for any association curve, whatever its eventual shift, the most representative value of P(SO2). For all mixed groups, the most significant parameter would be P44. If the characterization of the HbO2 dissociation curve is to be represented by a single point, then it should be P48 for the NL curves, P52 for the LD group, and P24 for the RD group. Then, the common use of P50 appeared to be illegitimate and inadequate for the right shifted curves, a very frequent circumstance in clinical practice.