[Effect of dyshemoglobinemia (methemoglobinemia and carboxyhemoglobinemia) on accuracy of measurement in pulse oximetry in operations of long duration].

In two groups - eleven patients each - who underwent long lasting surgical procedures, arterial oxygen saturation was measured non-invasively (pulse oximeter) and invasively (CO-oximeter). In addition, total haemoglobin (THb), fractions of haemoglobin (metHb) and bloodgases were monitored. All patients of group I received an epidural anaesthesia via catheter with prilocaine combined with general anaesthesia. In group II a modified neurolept analgesia (balanced) was performed. As expepted we found a prilocaine-induced methaemoglobinaemia (metHb up to 11.6 Vol%) whereas in group II physiological values were not exceeded. Carboxyhaemoglobin up to 4 Vol% was found in the smokers of both groups, which decreased continuously - in contrast to metHb - during hyperoxic ventilation. Comparing the non-invasively and the invasively determined saturations no correlation (r = -0.002) was found in group I. However, a positive correlation (r = 0.652) was obtained in group II. After correction of SaO2 (pulse oximeter) using an adjusted formula a fair correlation (0.613) was found in group I, too. Using this equation for correction the correlation in group II increased to r = 0.824. Because of methodological facts (the pulse oximeter registers only desoxygenated haemoglobin) the arterial oxygen desaturation (O2Hb or fractional saturation) accompanying dyshaemoglobinaemia remains undetectable to the pulse oximeter. In the cases of known or expected dyshaemoglobinaemia pulse oximetry can yield only limited information; fractions of inactive haemoglobin should be measured by other means or accounted for by the given equation for correction (SaO2 corr. = SaO2 pulse oximeter - (COHb + metHb]. Pathophysiological effects of dyshaemoglobinaemia are discussed.