A dynamic recovery failure index after acute type A aortic dissection surgery

Acute type A aortic dissection (ATAAD) is associated with a high risk of postoperative severe organ dysfunction. Preoperative risk scores alone may not capture early postoperative pathophysiological recovery. This study aimed to develop and temporally validate a dynamic inflammation-perfusion-coagulation recovery failure index (dIPC-RFI), based on routinely available preoperative, immediate postoperative, and 24-hour postoperative laboratory and blood gas variables, for risk restratification of severe organ dysfunction beyond 24 h after ATAAD surgery.

Patients with ATAAD who underwent emergency surgical repair at The University of Hong Kong-Shenzhen Hospital between August 2021 and December 2025 were retrospectively included. Postoperative 24 h was used as the fixed risk-assessment starting point; only patients who were alive, free from the primary outcome, and had complete variables required for dIPC-RFI assessment at 24 h were included. Patients were split chronologically into a derivation cohort (first 70%) and a temporal validation cohort (last 30%). The primary outcome was severe organ dysfunction occurring after 24 h until discharge or within 30 days, defined as any of the following: continuous renal replacement therapy, mechanical ventilation for ≥72 h, neurological complications, multiple organ dysfunction syndrome, in-hospital death after 24 h or 30-day death, intra-aortic balloon pump support, extracorporeal membrane oxygenation, or tracheostomy. dIPC-RFI consisted of three domains-inflammation, perfusion, and coagulation-each including one static 24-hour variable and one dynamic change variable. Cutoffs were derived using receiver operating characteristic curves and the Youden index in the derivation cohort and then fixed for the validation cohort. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), Brier score, calibration curve, calibration intercept and slope, and decision curve analysis.

A total of 330 patients were included, of whom 134 (40.6%) developed severe organ dysfunction after 24 h. Event rates were 39.0% in the derivation cohort and 44.4% in the temporal validation cohort (p = 0.392). In the validation cohort, dIPC-RFI achieved an AUC of 0.873 (95% CI, 0.794-0.935) and a Brier score of 0.1387, outperforming the simplified clinical model (AUC = 0.762) and the extended clinical model (AUC = 0.735). The combined model integrating the simplified clinical model and dIPC-RFI further improved discrimination (AUC = 0.908) and reduced the Brier score to 0.1243. Component-specific analyses showed acceptable discrimination for major components with sufficient validation events, including CRRT, prolonged mechanical ventilation, neurological complications, and MODS component. The no-CRRT sensitivity analysis and the death-or-MODS sensitivity analysis supported the robustness of the main findings.

dIPC-RFI provides a simple 24-hour postoperative framework for identifying incomplete recovery across inflammation, perfusion, and coagulation after ATAAD surgery. It may help ICU teams identify patients who require intensified monitoring, repeated lactate and coagulation assessment, early multidisciplinary review, and preparation for organ support. dIPC-RFI should be used as a risk-restratification aid rather than a stand-alone treatment trigger, and prospective multicenter validation is required before routine implementation.