Bioelectrical impedance analysis of albumin versus succinylated gelatin priming during cardiopulmonary bypass: a propensity-matched retrospective cohort study

Fluid selection for cardiopulmonary bypass (CPB) priming may influence intravascular volume support and perioperative fluid distribution. Bioelectrical impedance analysis (BIA) can detect short-term changes in body-water distribution and phase angle (PA), but its response to different colloid-containing primes remains uncertain. We compared albumin-containing and succinylated gelatin-containing priming, with perioperative change in PA as the primary outcome. This single-centre retrospective cohort was drawn from a 400-patient prematching cohort and included 128 adults in 64 propensity-score-matched pairs. Propensity-score matching used seven clinical covariates. The primary model adjusted for baseline PA, CPB duration, and preoperative SOFA score and used HC3 robust inference. Pair-clustered inference, an extended model, and an augmented inverse probability-weighted (AIPW) analysis assessed matched-design uncertainty and residual confounding. Residual imbalance remained for APACHE II score, CPB duration, NYHA class, and several surgery-type categories. Median change in PA was −0.40° with gelatin and 0.00° with albumin (unadjusted P = 0.052). In the primary model, albumin priming was associated with a 0.63° higher change in PA (95% CI 0.03–1.23; P = 0.041). Pair-clustered inference gave the same point estimate. The extended model (0.60°, 95% CI −0.06 to 1.26) and AIPW analysis (0.54°, pair-bootstrap 95% CI −0.06 to 1.19) retained a similar direction but included no difference. Fluid-distribution and short-term recovery outcomes did not differ materially, and no subgroup interaction was statistically significant. Albumin-containing priming was associated with less perioperative PA decline in the primary model, but the estimate was sensitive to residual-confounding specifications. The results support a physiological signal that warrants prospective study, not established causal or patient-centred benefit.