Background: Metabolic acidosis is a common acid-base imbalance in critically ill patients. Whether sodium bicarbonate (SB) can improve clinical outcomes in the treatment of metabolic acidosis is still controversial. The aim of this study was to determine the factors influencing the clinical efficacy of SB in the treatment of metabolic acidosis and the potential benefit to patients.
Methods: Patients with metabolic acidosis who were treated with or without SB were identified and grouped from a retrospective cohort (Pediatric Intensive Care Unit [PICU] database), from which the clinical data were extracted. The in-hospital mortality curves of the acid-base balance parameters of patients in the two groups were drawn and fitted using the locally-weighted scatter plot smoothing (LOWESS) method. The prevalence ratios (PRs) of in-hospital mortality were estimated by log-binomial regression based on the maximum likelihood method, and the potential confounders, such as age and disease category, were adjusted.
Results: A total of 6,167 children with metabolic acidosis were enrolled, of whom 2,626 (42.58%) were treated with SB. The overall analysis showed that there was no significant difference in the in-hospital mortality rates (9.71% vs. 10.56%, p = 0.275) between children in the SB treatment and non-treatment groups, adjusted PR = 0.929 (95% CI, 0.802-1.072). There was no significant difference in the in-hospital mortality rates as a function of pH and HCO3- between the two groups. The in-hospital mortality rate as a function of chloride was significantly different; specifically, the curve of the untreated group was U-shaped and the curve of the treated group was L-shaped. The curves of the two groups crossed at 110 mmol/L of chloride after LOWESS fitting. There was no statistically significant difference in the risk of death between the SB treatment and non-treatment groups at a chloride < 107 mmol/L and a chloride >113 mmol/L. In the chloride < 107 mmol/L subgroup, SB treatment had a 41.7% increased risk of in-hospital death (adjusted PR=1.417, 95% CI, 1.069−1.481) and a 35.9% increased risk of 28-day death (adjusted PR=1.359, 95% CI, 1.315−1.474). In the chloride≥113 mmol/L subgroup, SB treatment had a 61.1% reduced risk of in-hospital death (adjusted PR=0.389, 95% CI, 0.268−0.553) and a 56.4% reduced risk of 28-day death (adjusted PR=0.436, 95% CI, 0.295−0.631). The median length of stay in the PICU of children in the SB group was also shorter than children in the non-treatment group when the chloride concentration was ≥110 mmol/L.
Conclusions: The clinical outcomes of SB in the treatment of metabolic acidosis are associated with chloride. When the chloride concentration was high (> 110 mmol/L), children benefited from SB treatment and when the chloride concentration was low (< 107 mmol/L), the risk of death increased.