Temporal Physiologic Profiles Associated with Return of Spontaneous Circulation in a Swine Model of Cardiopulmonary...

Background: Physiology-guided cardiopulmonary resuscitation (CPR) is increasingly recognized as a strategy to move beyond mechanical targets, but evidence remains limited and dominated by isolated metrics. We characterized multimodal physiologic profiles associated with return of spontaneous circulation (ROSC) in a swine model. Methods: We analyzed 24 swine undergoing standardized cardiac arrest followed by CPR. Variables included end-tidal carbon dioxide (ETCO₂), systolic arterial pressure (SAP), diastolic arterial pressure (DAP), right ventricular systolic pressure (RVSP), carotid artery blood flow, stroke volume, and cerebral regional oxygen saturation. The primary early analysis used per-animal medians from minutes 1-5 of CPR, before any animal achieved ROSC. Later prespecified 5 min windows were evaluated exploratorily because they increasingly included post-ROSC measurements. A pre-ROSC sensitivity analysis, correlation analysis, and exploratory principal component analysis (PCA) with bootstrap assessment of loading stability were also performed. Results: Nine animals (37.5%) achieved ROSC, with a median time to ROSC of 10 min (range 6-18 min). During minutes 1-5, before any ROSC occurred, animals subsequently achieving ROSC had higher carotid flow, DAP, ETCO₂, right ventricular pressure signal, and SAP; all differences remained significant after false discovery rate correction. In the pre-ROSC sensitivity analysis, ETCO₂, SAP, DAP, right ventricular pressure signal, carotid flow, and stroke volume differed between animals subsequently achieving ROSC and those that did not (all p < 0.001), whereas cerebral regional oxygen saturation did not (p = 0.474). Later analytical windows increasingly included post-ROSC measurements and were therefore interpreted exploratorily. PCA summarized the shared early physiologic variation, with PC1 explaining 69.8% of variance; bootstrap analysis showed consistent positive contributions of the included variables to PC1. Conclusions: Successful resuscitation was associated with a distinct pressure-flow-gas-exchange profile during early and pre-ROSC CPR. These findings support further investigation of multimodal physiology-guided resuscitation.