Evaluation of the expected ventilatory response to metabolic acidosis in chronic hemodialysis patients

INTRODUCTION: Expected pCO2 during metabolic acidosis should be computed to rule out superimposing respiratory acid-base disorders, the latter being featured by too high or too low pCO2 than expected. At least 4 different and inconsistent formulas exist. Based on the common practical rule, the reduction of pCO2 equals 1.2 multiplied by the reduction of serum bicarbonate ([HCO3-]), whereas pCO2 equals to the 2 digit after pH decimal point according to Fulop. Instead, based on Winters' formula, pCO2 equals to 1.5 times [HCO3-] + 8. Finally, a very simple formula has been proposed, it reads pCO2  = [HCO3-] + 15. Beyond the evident simplicity, the latter has been effective in a small hemodialysis cohort but validation in a larger dataset is highly desirable.

METHODS: Formulas have been applied to 180 hemodialysis patients' blood samples dataset featured by mild metabolic acidosis (HCO3  ≥ 14 mEq/L) and root mean square errors (RMSE) associated with each formula were computed. pCO2 reference range was computed as expected pCO2  ± 2 RMSE.

FINDINGS: Fulop's rule and Winters' formula are associated with large prediction errors. Conversely, the common practical rule and the very simple formula are both featured by the same low error (1.7 mmHg). They further show good agreement and pick out the same reference range.

DISCUSSION: Superimposing respiratory acid-base disorders can be promptly and effectively ruled out by computing expected pCO2 as [HCO3-] + 15, a very simple formula proved to be interchangeable with the common practical rule that requires computes and assumptions, but leads to same results.