Systemic inflammatory response syndrome (SIRS) is a common complication following cardiopulmonary bypass (CPB), associated with increased mortality. We assessed the relationship between indexed oxygen delivery (DO₂i) during CPB and SIRS.

We conducted a prospective observational study at 2 institutions. The primary end-point was clinically-defined SIRS 12 hours after surgery. The secondary end-point was a composite outcome comprising death, TIA/stroke, renal replacement therapy, bleeding, mechanical circulatory support, or intensive care unit (ICU) stay >96 hours. The primary analysis modelled DO₂i in multivariable logistic regression. Linearity was assessed with restricted cubic splines, knot-sensitivity, quintiles, and piecewise fits. The optimal ROC-derived threshold was explored. Patients above and below the threshold were matched 1:1 by propensity score. Structural equation modelling (SEM) assessed mediation between DO₂i, SIRS, and outcomes.

Of 1154 patients screened, 908 were analysed; 221 (24.3%) developed SIRS. Median DO₂i was lower in the SIRS group (274 vs 302 mL/min/m², P < .001). DO₂i was inversely associated with SIRS (aOR 0.798; 95% CI 0.750-0.850, P < .001, per +10 mL/min/m²). In the secondary analysis, DO₂i ≤293 mL/min/m² was identified as threshold (sensitivity 62%, specificity 74%). Propensity score matched 390 patient pairs, with higher SIRS incidence in the low-DO₂i group (33.3% vs 14.4%; P < .001). The composite outcome occurred more frequently in the low-DO₂i cohort (17.7% vs 8.2%; P < .001). SEM showed mediation by SIRS, accounting for 46.2% of the DO₂i effect on outcomes (OR 1.052; 95% CI 1.036-1.067; P < .001).

Low DO₂i during CPB predicts SIRS and adverse outcomes. Goal-directed perfusion to reduce inflammation warrants evaluation in randomized trials.