Background: Patients with sepsis admitted to the intensive care unit often present with acid-base disorders. As the traditional interpretation might be clinically misleading, an alternative approach described by Stewart may allow one to quantify the individual components of acid-base abnormalities and provide an insight into their pathogenesis. The aim of our study was to compare the traditional and Stewart approaches in the analysis of acid-base disturbance.
Methods: We analyzed arterial blood gases (ABG) taken from 43 ICU septic patients from admission to discharge categorising them according to SBE values. The traditional concept analysis was compared with the physicochemical approach using the Stewart equations.
Results: 990 ABGs were analysed. In the SBE < -2 mEq L⁻¹ group, hyperlactatemia was observed in 34.7% ABG, hypoalbuminemia in 100% and SIG acidosis in 42% ABG. Moreover, a Cl/Na ratio > 0.75 was present in 96.9% ABG. In the normal range SBE group, elevated lactates were present in 21.3% ABG, SIG acidosis in 14.9%, elevated Cl/Na ratio in 98.4% and hypoalbuminemia in all 324 ABG. In the metabolic alkalosis group (SBE > +2 mEq L⁻¹), hyperlactatemia was observed in 18.4% ABG, SIG acidosis in 5% ABG, Cl/Na ratio> 0.75 in 88.8%, while 99.1% samples revealed hypoalbuminemia.
Conclusion: The use of the Stewart model may improve our understanding of the underlying pathophysiological mechanism and the true etiology of the derangements of acid-base disorders. Indeed, it proves that patients may suffer from mixed arterial blood gas disorders hidden under normal values of SBE and pH.
Keywords: Stewart approach; acid-base disorders; critical care; sepsis.