Effectiveness of hypotension prediction index software in reducing intraoperative hypotension in prolonged prone-position spine surgery: a single-center clinical trial.

Abstract

Intraoperative hypotension (IOH) is associated with morbidity and mortality. The Hypotension Prediction Index (HPI), a machine learning-based tool, offers the opportunity for a proactive approach by predicting hypotensive events. This single center, single blind randomized clinical trial aimed to evaluate the hypothesis that an HPI software-guided approach to IOH management during prone position spine surgery could reduce its incidence compared to our standard care practices. 85 adult patients undergoing spine fusion surgery in the prone position were enrolled. Patients were randomized with a 1:1 allocation ratio. Participants were blinded to their group allocation. In the intervention group, the HPI software was actively used to guide IOH management. In the control group, HPI software readings were blinded, and standard care was administered. The primary outcome was the comparison of time-weighted average (TWA) of IOH between the two groups. Secondary outcomes included a comparison of the incidence of postoperative in-hospital events related to IOH between groups. 77 patients were included in the final analysis (39 in the intervention group), as 8 patients were excluded due to technical issues. No statistically significant difference was found between the intervention and control groups in the TWA of IOH (0.10 mmHg [0.05, 0.23] vs. 0.15 mmHg [0.09, 0.37], p-value 0.088). However, the total duration of hypotensive events per patient was significantly lower in the intervention group (4 min [0.5, 12.2] vs. 11.2 min [2.6, 20.1]; p-value 0.019). Postoperative complication rates did not differ significantly between the two groups. HPI-guided management did not significantly reduce the TWA of IOH compared to standard care in patients undergoing prone-position spine surgery. Complication rates were similar between the two groups.Clinical Trial Registration: This trial was registered with ClinicalTrials.gov (registration number: NCT05341167).