Determinants of oxygenation during hemodialysis and related procedures. A report of data acquired under varying conditions and a review of the literature.

Abstract

A decrease in arterial oxygen tension during hemodialysis has been attributed to a number of factors. In order to more completely define these factors, we studied respiratory gas exchange, arterial blood gases and pH, and dialyzer flux of CO2 during pure ultrafiltration, three types of acetate dialysis, and sorbent regenerated bicarbonate dialysis in which the dialysate concentration of bicarbonate varies. Changes due to position and extracorporeal circulation of a 300-ml volume of blood (sham dialysis) were studied for any effect contributing to the hypoxemia noted with circulation through the membrane and variation in dialysate. Alveolar oxygen tension (PAO2) is calculated by the equation PAO2 = PIO2-PaCO2 (FIO2 + 1-FIO2/RE). RE is the ratio of CO2 excretion by the lung (VCO2) to oxygen consumption (VO2). RE equals RQ (metabolic quotient) when no extrapulmonary CO2 losses occur. Normals in a lounge chair had no change in RE and PAO2. RE decreased to 0.75 during sham dialysis and PAO2 decreased. During pure ultrafiltration RE decreased due to a decrease in VO2 and VCO2 with proportionately greater decrease in VCO2. PAO2 decreased accordingly. Acetate dialysis produced an increase in oxygen consumption without a proportional increase in CO2 excretion and both RQ and RE decreased. When PAO2 decreased during any of these procedures, arterial oxygen tension (PaO2) decreased without a change in A-aO2 gradient. No changes in PaCO2 were noted. RQ did not change during bicarbonate dialysis. At high bicarbonate dialysate concentrations, however, PaCO2 increased and PAO2 decreased. The major reason for hypoxemia during acetate dialysis is a decrease in alveolar oxygen tension due to changes in metabolism and a decrease in pulmonary CO2 excretion when CO2 is lost from the dialyzer. The increasing pH may contribute to the metabolic change during acetate dialysis and the hypoventilation during bicarbonate dialysis. There is little evidence to support an effect of pulmonary capillary obstruction or changes in oxyhemoglobin association on the decrease in arterial oxygen tension observed.