Physicochemical effects of bicarbonate-buffered solution versus compound sodium lactate during major abdominal surgery: A randomized, open-label, non-inferiority trial

Introduction Bicarbonate-buffered crystalloids avoid exogenous lactate and have physicochemical properties distinct from compound sodium lactate (CSL), but direct comparative evidence during major abdominal surgery remains limited. Methods This single-center, open-label, randomized non-inferiority trial enrolled high-risk adults undergoing elective major abdominal surgery. Participants received bicarbonate-buffered solution or CSL as the allocated intraoperative crystalloid. The primary estimand was the arithmetic mean difference in standard base excess (SBE) at skin closure (bicarbonate-buffered solution minus CSL). The non-inferiority margin was −1.5 mEq/L. The trial was prospectively registered with the Australian New Zealand Clinical Trials Registry (ACTRN12619001228178). Results Fifty participants were randomized (25 per group) and received their allocated intervention. End-of-surgery arterial blood gas data were missing for two CSL participants. In the intention-to-treat analysis using multiple imputation, the estimated mean difference in end-of-surgery SBE was −0.42 mEq/L (95% confidence interval [CI], −1.92 to 1.09 mEq/L). The complete-case per-protocol estimate was −0.48 mEq/L (95% CI, −2.02 to 1.06 mEq/L). In both analyses, the lower CI bound was below the prespecified non-inferiority margin; non-inferiority was not established. Secondary outcomes were exploratory; observed end-of-surgery lactate was lower with bicarbonate-buffered solution (median 0.9 vs 1.2 mmol/L; unadjusted P = 0.031), without a corresponding difference in SBE or pH. Conclusions In this small trial of hemodynamically stable adults undergoing elective major abdominal surgery, the primary non-inferiority result was inconclusive. The data did not establish interchangeability, equivalence, superiority, or inferiority of the two fluids. As the trial was powered for a physiologic endpoint, no comparative inference can be made about organ dysfunction, postoperative complications, recovery, or mortality. Larger trials with standardized co-interventions and patient-important outcomes are required before routine substitution for CSL can be recommended.