Clinical Use of an Automated SBP-Responsive Pneumatic Tourniquet: A Single-Center Retrospective Study.

Conventional pneumatic tourniquets typically rely on fixed empirical pressures that do not account for patient-specific vascular characteristics or dynamic hemodynamic changes, potentially exposing tissues to unnecessary mechanical and ischemic stress. This single-center retrospective cohort study evaluated the clinical performance and safety of an automated systolic blood pressure (SBP)-responsive pneumatic tourniquet in 203 consecutive eligible patients who met the predefined inclusion and exclusion criteria and underwent upper-extremity surgery between January 2023 and September 2024. Continuous or high-frequency SBP measurements were synchronized with cuff-pressure waveforms to characterize real-time closed-loop pressure modulation. Primary outcomes included hemostatic effectiveness and breakthrough bleeding, while secondary outcomes included cuff-pressure behavior, the relationship between SBP and cuff pressure, perioperative hemodynamic stability, and postoperative neurologic or cutaneous complications. The mean tourniquet cuff pressure was 39.61 ± 2.41 kPa (approximately 297 ± 18 mmHg), with values ranging from 34 to 46 kPa. A significant positive correlation was observed between preoperative SBP and cuff pressure (r = 0.706, p < 0.001). Hemostasis was consistently effective, with no breakthrough bleeding observed (0%; 95% CI, 0%-1.8%). Perioperative blood pressure showed only modest fluctuations, and no postoperative neurologic deficits, paresthesia, skin injury, or distal perfusion abnormalities were identified. These findings suggest that a fully automated SBP-responsive tourniquet can provide reliable hemostasis, dynamically maintain physiologically appropriate occlusion pressures, and achieve a favorable short-term safety profile without Doppler-based calibration or manual pressure adjustment, supporting its potential as an individualized and workflow-efficient approach to surgical tourniquet management.