A secondary analysis of the SODa-BIC randomized clinical trial using multiple hierarchical composite outcomes.

BACKGROUND

The Sodium Bicarbonate for Metabolic Acidosis in the ICU (SODa-BIC) trial assessed the effect of sodium bicarbonate on Major Adverse Kidney Events at 30 days (MAKE30) in critically ill patients with metabolic acidosis. Although the trial demonstrated no significant difference in the primary outcome, the individual components showed differing magnitudes and directions of effect; we undertook an exploratory analysis of the SODa-BIC trial using different hierarchical outcomes to better understand potential treatment effects of sodium bicarbonate.

METHODS

Post-hoc analysis of the SODa-BIC trial. All patients included in the SODa-BIC trial were eligible. The primary outcome was a win ratio analysis with a hierarchy consisting of death, requirement for renal replacement therapy, persistent renal dysfunction, and hospital length of stay. Several sensitivity and subgroup analyses were conducted.

RESULTS

The primary analysis resulted in 59,250 pairs from 488 patients. Consistent with the SODa-BIC analysis, there was no overall evidence of a treatment effect of sodium bicarbonate, with a win ratio of 0.97 (95%CI 0.79 - 1.19) and a Desirability of Outcome Ranking probability of 0.49 (95%CI 0.45 - 0.54). A majority of pairwise comparisons were resolved by mortality and hospital length of stay. Evidence of significant treatment effect modification was observed across subgroups defined by acute kidney injury and severity of acidemia.

CONCLUSION

In this post-hoc analysis of the SODa-BIC trial, the use of hierarchical composite outcomes did not alter the neutral overall interpretation of the original trial. However, this analysis provides additional insight into the relative contribution of individual components of the primary outcome.