Association Between the Dialysate Bicarbonate and the Pre-dialysis Serum Bicarbonate Concentration in Maintenance Hemodialysis: A Retrospective Cohort Study

Background: It is unclear whether the use of higher dialysate bicarbonate concentrations is associated with clinically relevant changes in the pre-dialysis serum bicarbonate concentration. Objective: The objective is to examine the association between the dialysate bicarbonate prescription and the pre-dialysis serum bicarbonate concentration. Design: This is a retrospective cohort study. Setting: The study was performed using linked administrative health care databases in Ontario, Canada. Patients: Prevalent adults receiving maintenance in-center hemodialysis as of April 1, 2020 (n = 5414) were included. Measurements: Patients were grouped into the following dialysate bicarbonate categories at the dialysis center-level: individualized (adjustment based on pre-dialysis serum bicarbonate concentration) or standardized (>90% of patients received the same dialysate bicarbonate concentration). The standardized category was stratified by concentration: 35, 36 to 37, and ≥38 mmol/L. The primary outcome was the mean outpatient pre-dialysis serum bicarbonate concentration at the patient level. Methods: We examined the association between dialysate bicarbonate category and pre-dialysis serum bicarbonate using an adjusted linear mixed model. Results: All dialysate bicarbonate categories had a mean pre-dialysis serum bicarbonate concentration within the normal range. In the individualized category, 91% achieved a pre-dialysis serum bicarbonate ≥22 mmol/L, compared to 87% in the standardized category. Patients in the standardized category tended to have a serum bicarbonate that was 0.25 (95% confidence interval [CI] = −0.93, 0.43) mmol/L lower than patients in the individualized category. Relative to patients in the 35 mmol/L category, patients in the 36 to 37 and ≥38 mmol/L categories tended to have a serum bicarbonate that was 0.70 (95% CI = −0.30, 1.70) mmol/L and 0.87 (95% CI = 0.14, 1.60) mmol/L higher, respectively. There was no effect modification by age, sex, or history of chronic lung disease. Limitations: We could not directly confirm that all laboratory measurements were pre-dialysis. Data on prescribed dialysate bicarbonate concentrations for individual dialysis sessions were not available, which may have led to some misclassification, and adherence to a practice of individualization could not be measured. Residual confounding is possible. Conclusions: We found no significant difference in the pre-dialysis serum bicarbonate concentration irrespective of whe