Hypotension Prediction Index Software to Prevent Intraoperative Hypotension during Major Non-Cardiac Surgery: Protocol for a European Multicenter Prospective Observational Registry (EU-HYPROTECT)

Manuel Ignacio Monge García; Daniel García-López; Étienne Gayat; Michael Sander; Peter Bramlage; Elisabetta Cerutti; Simon James Davies; Abele Donati; Gaetano Draisci; Ulrich H Frey; Eric Noll; Javier Ripollés-Melchor; Hinnerk Wulf; Bernd Saugel

doi:10.3390/jcm11195585

Hypotension Prediction Index Software to Prevent Intraoperative Hypotension during Major Non-Cardiac Surgery: Protocol for a European Multicenter Prospective Observational Registry (EU-HYPROTECT)

J Clin Med. 2022 Sep 23;11(19):5585. doi: 10.3390/jcm11195585.

Authors

Manuel Ignacio Monge García¹, Daniel García-López², Étienne Gayat^{3

4}, Michael Sander⁵, Peter Bramlage⁶, Elisabetta Cerutti⁷, Simon James Davies^{8

9}, Abele Donati¹⁰, Gaetano Draisci¹¹, Ulrich H Frey¹², Eric Noll¹³, Javier Ripollés-Melchor¹⁴, Hinnerk Wulf¹⁵, Bernd Saugel^{16

17}

Affiliations

¹ Medical Affairs Department, Critical Care Europe, Edwards Lifesciences, 1260 Nyon, Switzerland.
² Department of Anaesthesiology and Reanimation, University Hospital Marqués de Valdecilla, 39008 Santander, Spain.
³ Université Paris Cité, INSERM, 75006 Paris, France.
⁴ Department of Anesthesia and Critical Care Medicine, Hôpital Lariboisière, 75475 Paris, France.
⁵ Department of Anaesthesiology, Intensive Care Medicine and Pain Medicine, University Hospital Giessen, Justus-Liebig University Giessen, 35392 Giessen, Germany.
⁶ Institute for Pharmacology and Preventive Medicine, 49661 Cloppenburg, Germany.
⁷ Department of Anesthesia, Transplant and Surgical Intensive Care, Azienda Ospedaliero Universitaria Ancona, 60126 Ancona, Italy.
⁸ York and Scarborough Teaching Hospitals NHS Foundation Trust, York YO31 8HE, UK.
⁹ Centre for Population and Health Studies, Hull York Medical School, York HU6 7RU, UK.
¹⁰ Department of Biomedical Sciences and Public Health, Università Politecnica delle Marche, 60126 Ancona, Italy.
¹¹ Unit of Obstetric and Gynecologic Anesthesia, IRCCS Fondazione Policlinico Universitario Agostino Gemelli, 00168 Rome, Italy.
¹² Department of Anesthesiology, Intensive Care, Pain and Palliative Care, Marien Hospital Herne, Ruhr-University Bochum, 44801 Bochum, Germany.
¹³ Department of Anesthesiology and Intensive Care, Les Hôpitaux Universitaires de Strasbourg, 67098 Strasbourg, France.
¹⁴ Anesthesia and Critical Care Department, Hospital Universitario Infanta Leonor, 28031 Madrid, Spain.
¹⁵ Department of Anaesthesiology-Intensive Care, University Hospital Marburg, 35043 Marburg, Germany.
¹⁶ Department of Anesthesiology, Center of Anesthesiology and Intensive Care Medicine, University Medical Center Hamburg-Eppendorf, 20251 Hamburg, Germany.
¹⁷ Outcomes Research Consortium, Cleveland, OH 44195, USA.

Abstract

Background: Intraoperative hypotension is common in patients having non-cardiac surgery and associated with postoperative acute myocardial injury, acute kidney injury, and mortality. Avoiding intraoperative hypotension is a complex task for anesthesiologists. Using artificial intelligence to predict hypotension from clinical and hemodynamic data is an innovative and intriguing approach. The AcumenTM Hypotension Prediction Index (HPI) software (Edwards Lifesciences; Irvine, CA, USA) was developed using artificial intelligence—specifically machine learning—and predicts hypotension from blood pressure waveform features. We aimed to describe the incidence, duration, severity, and causes of intraoperative hypotension when using HPI monitoring in patients having elective major non-cardiac surgery. Methods: We built up a European, multicenter, prospective, observational registry including at least 700 evaluable patients from five European countries. The registry includes consenting adults (≥18 years) who were scheduled for elective major non-cardiac surgery under general anesthesia that was expected to last at least 120 min and in whom arterial catheter placement and HPI monitoring was planned. The major objectives are to quantify and characterize intraoperative hypotension (defined as a mean arterial pressure [MAP] < 65 mmHg) when using HPI monitoring. This includes the time-weighted average (TWA) MAP < 65 mmHg, area under a MAP of 65 mmHg, the number of episodes of a MAP < 65 mmHg, the proportion of patients with at least one episode (1 min or more) of a MAP < 65 mmHg, and the absolute maximum decrease below a MAP of 65 mmHg. In addition, we will assess causes of intraoperative hypotension and investigate associations between intraoperative hypotension and postoperative outcomes. Discussion: There are only sparse data on the effect of using HPI monitoring on intraoperative hypotension in patients having elective major non-cardiac surgery. Therefore, we built up a European, multicenter, prospective, observational registry to describe the incidence, duration, severity, and causes of intraoperative hypotension when using HPI monitoring in patients having elective major non-cardiac surgery.

Keywords: advanced hemodynamic monitoring; artificial intelligence; blood pressure; hemodynamic instability; machine learning; postoperative complications.

Grants and funding

N/A/Edwards Lifesciences (Switzerland)