Personalized predictive hemodynamic management for major oncologic surgery: effect of progressive implementation of monitoring of digital medical devices with artificial intelligence-based algorithms.
Gilda Pasta, Luciano Frassanito, Mariangela Calabria, Francesco Vassalli, Andrea Belli, Giulia Torricella, Arturo Cuomo, Francesca Bifulco
{"title":"Personalized predictive hemodynamic management for major oncologic surgery: effect of progressive implementation of monitoring of digital medical devices with artificial intelligence-based algorithms.","authors":"Gilda Pasta, Luciano Frassanito, Mariangela Calabria, Francesco Vassalli, Andrea Belli, Giulia Torricella, Arturo Cuomo, Francesca Bifulco","doi":"10.23736/S0375-9393.25.18937-2","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Avoiding intraoperative hypotension and maintaining an adequate cardiac output (CO) during surgery is crucial to ensure tissue oxygen delivery to the tissues and avoid excessive fluid administration. Assisted fluid management (AFM) is a \"decision-support\" system based on artificial intelligence (AI) that helps the clinician to manage fluids during major surgeries. Hypotension Prediction Index (HPI) is a predictive parameter of intraoperative hypotension. The aim of this study was to assess the relative contribution of different levels of technologic assistance (CO monitoring, CO plus HPI, and CO plus HPI and AFM) to the improvement of hemodynamic management, applied to comparable cohorts of non-cardiac surgical patients.</p><p><strong>Methods: </strong>We conducted a retrospective analysis of consecutive patients undergoing major oncologic abdominal surgery, monitored with an arterial radial catheter progressively upgraded with three increasing levels of forecasting event technology. A personalized goal directed fluid therapy (GDT) protocol was applied in all groups. The primary outcome was the time-weighted average (TWA) mean arterial pressure (MAP) <65 mmHg among the three cohorts. Secondary outcomes were the percentage of monitoring time spent with stroke volume variation > 12% and with Cardiac Index (CI) <2 L/min/m<sup>2</sup>.</p><p><strong>Results: </strong>Eighty-two consecutive patients were enrolled: 26 patients in the GDT group, 28 in the HPI group and 28 in the AFM group. TWA-MAP<65 mmHg was 1.13 (0.13-1.83) mmHg in the GDT group, 0.96 (0.26-1.85) mmHg in the HPI group, 0.42 (0.07-0.93) mmHg in the AFM group. Patients with AFM spent roughly 30% of the monitoring time with a CI<2 L/min/m<sup>2</sup> compared to less than 10% in the other two groups (Kruskal-Wallis P value 0.013).</p><p><strong>Conclusions: </strong>An increasing levels of artificial intelligence-based hemodynamic monitoring and decision-support tools shows a trend towards decreasing IOH, but it did not reach statistical significance.</p>","PeriodicalId":18522,"journal":{"name":"Minerva anestesiologica","volume":" ","pages":""},"PeriodicalIF":2.9000,"publicationDate":"2025-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Minerva anestesiologica","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.23736/S0375-9393.25.18937-2","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ANESTHESIOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Avoiding intraoperative hypotension and maintaining an adequate cardiac output (CO) during surgery is crucial to ensure tissue oxygen delivery to the tissues and avoid excessive fluid administration. Assisted fluid management (AFM) is a "decision-support" system based on artificial intelligence (AI) that helps the clinician to manage fluids during major surgeries. Hypotension Prediction Index (HPI) is a predictive parameter of intraoperative hypotension. The aim of this study was to assess the relative contribution of different levels of technologic assistance (CO monitoring, CO plus HPI, and CO plus HPI and AFM) to the improvement of hemodynamic management, applied to comparable cohorts of non-cardiac surgical patients.
Methods: We conducted a retrospective analysis of consecutive patients undergoing major oncologic abdominal surgery, monitored with an arterial radial catheter progressively upgraded with three increasing levels of forecasting event technology. A personalized goal directed fluid therapy (GDT) protocol was applied in all groups. The primary outcome was the time-weighted average (TWA) mean arterial pressure (MAP) <65 mmHg among the three cohorts. Secondary outcomes were the percentage of monitoring time spent with stroke volume variation > 12% and with Cardiac Index (CI) <2 L/min/m2.
Results: Eighty-two consecutive patients were enrolled: 26 patients in the GDT group, 28 in the HPI group and 28 in the AFM group. TWA-MAP<65 mmHg was 1.13 (0.13-1.83) mmHg in the GDT group, 0.96 (0.26-1.85) mmHg in the HPI group, 0.42 (0.07-0.93) mmHg in the AFM group. Patients with AFM spent roughly 30% of the monitoring time with a CI<2 L/min/m2 compared to less than 10% in the other two groups (Kruskal-Wallis P value 0.013).
Conclusions: An increasing levels of artificial intelligence-based hemodynamic monitoring and decision-support tools shows a trend towards decreasing IOH, but it did not reach statistical significance.
期刊介绍：
Minerva Anestesiologica is the journal of the Italian National Society of Anaesthesia, Analgesia, Resuscitation, and Intensive Care. Minerva Anestesiologica publishes scientific papers on Anesthesiology, Intensive care, Analgesia, Perioperative Medicine and related fields.
Manuscripts are expected to comply with the instructions to authors which conform to the Uniform Requirements for Manuscripts Submitted to Biomedical Editors by the International Committee of Medical Journal Editors.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
