Predictive Ability of Stroke Volume Changes Following Positive End-Expiratory Pressure Reduction on Fluid Responsiveness During One-Lung Ventilation.

Abstract

During one-lung ventilation (OLV) with low tidal volumes, asymmetric intrathoracic pressures may blunt the cardiopulmonary interactions on which conventional dynamic indices rely, limiting their usefulness for guiding fluid therapy. The positive end-expiratory pressure (PEEP) test, a reduction in PEEP to augment venous return, has been proposed as a fluid-free assessment of preload responsiveness in mechanically ventilated patients. However, whether its diagnostic performance is retained during OLV is unknown. We therefore hypothesized that the stroke-volume change (ΔSVPEEP test) induced by the PEEP test would predict fluid responsiveness during OLV. We conducted a single-center prospective interventional study (June to September 2024) in adults undergoing elective lung resection requiring OLV. The primary objective was to evaluate the diagnostic accuracy of the ΔSVPEEP test for predicting fluid responsiveness. Secondary objectives were to assess the diagnostic performance of the PEEP test-induced change in mean arterial pressure (ΔMAPPEEP test) and baseline stroke volume variation (SVV), and to compare their predictive abilities. After stabilization using OLV (PEEP 10 cm H2O) but before skin incision (closed-chest conditions), baseline hemodynamics were recorded; stroke volume (SV) was measured by arterial pulse-contour analysis. PEEP was then reduced from 10 to 0 cm H2O, and variables were re-measured. Subsequently, a 15° head-down tilt (Trendelenburg) served as an autotransfusion challenge; patients with ≥10% SV increase were classified as responders. Receiver operating characteristic (ROC) analysis was used to assess the predictive ability of each index. A gray-zone analysis was performed for the ΔSVPEEP test to delineate the range of diagnostic uncertainty. Of 43 patients screened, 38 were analyzed; 18 (47%) were responders. ΔSVPEEP test was larger in responders than in nonresponders (16.7% ± 8.9 vs 4.6% ± 4.3; P < .0001). ΔSVPEEP test predicted fluid responsiveness with an area under the ROC curve (AUC) of 0.95 (95% confidence interval [CI], 0.82-0.99; P < .0001), and the optimal cutoff of 8.6% yielded 94.4% sensitivity (95% CI, 72.7-99.9) and 80.0% specificity (95% CI, 56.3-94.3). ΔMAPPEEP test showed an AUC of 0.84 (95% CI, 0.68-0.94; P < .0001), whereas SVV was not predictive (AUC 0.64, 95% CI, 0.47-0.79; P = .12). The gray zone for the ΔSVPEEP test was 3.6% to 12.5%, comprising 14 of 38 (36.8%) of patients. During OLV, the relative increase in SV elicited by the PEEP test predicted fluid responsiveness as defined by the Trendelenburg maneuver. This simple, fluid-free maneuver may assist intraoperative fluid management during thoracic anesthesia.