The Hypotension Prediction Index versus standard advanced haemodynamic monitoring in patients undergoing major aortic surgery (HYPE-AORTA): study protocol for a prospective randomised controlled trial

Technical Summary: The HYPE-AORTA Trial Protocol

Problem Statement
Intraoperative hypotension is a well-established risk factor for postoperative organ dysfunction and mortality, particularly in high-risk vascular surgery. While Goal-Directed Therapy (GDT) utilizing reactive monitoring (e.g., cardiac output derived from arterial pressure) has shown benefits in some surgical cohorts, evidence remains inconsistent for major aortic procedures. These patients face unique physiological stresses, including abrupt hemodynamic shifts during aortic cross-clamping and declamping.

Current monitoring strategies are fundamentally reactive, meaning clinicians intervene only after hypotension has occurred. This lag may contribute to a cumulative "hypotensive burden" (depth and duration of low blood pressure) that correlates with organ injury. Conversely, the Hypotension Prediction Index (HPI), a machine learning algorithm analyzing arterial waveform characteristics, can predict hypotensive events up to 15 minutes in advance. However, critical knowledge gaps persist regarding HPI's efficacy in major aortic surgery. Previous retrospective data from the authors' institution suggested that while HPI might reduce hypotension time, it could inadvertently increase intraoperative hypertension (overcorrection), a known risk factor for adverse outcomes. Furthermore, no randomized controlled trial (RCT) has definitively demonstrated that predictive monitoring reduces clinically meaningful postoperative outcomes (e.g., acute kidney injury, myocardial injury) in this specific population.

Methodology
The HYPE-AORTA study is a single-centre, prospective, parallel-group, randomized controlled trial designed to address these gaps.

• Population: The trial will enroll 200 patients (ASA physical status III–IV) undergoing elective major abdominal aortic surgery (open or endovascular) with an expected duration exceeding two hours.
• Intervention vs. Comparator: Participants are randomized 1:1 to one of two hemodynamic management strategies:
  1. HPI Group: Management guided by the Hypotension Prediction Index (Acumen IQ sensor). This group utilizes a protocol that triggers preemptive interventions (fluids, vasopressors, or anesthetic depth reduction) when HPI ≥85, alongside reactive management for established hypotension.
  2. Standard Group: Management guided by standard arterial pressure-derived cardiac output (APCO) monitoring (FloTrac sensor). This group follows a reactive protocol based on observed Mean Arterial Pressure (MAP) and derived parameters (SVV, SVRI, CI).
• Protocols: Both groups adhere to structured algorithms targeting a MAP ≥75 mmHg while avoiding excessive hypertension (MAP ≤100 mmHg). The HPI protocol specifically incorporates lessons from the authors' retrospective analysis to mitigate hypertensive overcorrection.
• Primary Outcome: The Time-Weighted Average of MAP below 65 mmHg (TWA-MAP <65 mmHg) from induction to departure from the operating theater. This metric integrates both the depth and duration of hypotension.
• Secondary Outcomes:
  • A composite of postoperative organ complications within 7 days (Acute Kidney Injury, Myocardial Injury after Non-Cardiac Surgery [MINS], stroke, respiratory failure, circulatory failure).
  • Intraoperative hypertension metrics (time and TWA-MAP above 90 and 100 mmHg).
  • 90-day all-cause mortality.
  • Resource utilization (fluids, vasopressors, length of stay).
• Statistical Approach: The study employs an estimation-focused framework rather than a purely hypothesis-testing one. The primary analysis uses a generalized linear model with a Gamma distribution and log-link to estimate the ratio of mean TWA-MAP between groups. The sample size (n=200) was calculated to provide sufficient precision for effect size estimation (targeting a 30% relative reduction) rather than achieving 80% power for a binary significance test. Sensitivity analyses include hurdle models for zero-inflated data and per-protocol analyses. Blinding is maintained for outcome assessors and data analysts, though the anesthesia team cannot be blinded to the monitoring modality.

Key Contributions

• Protocol Refinement: The study design explicitly addresses the "overcorrection" risk identified in prior retrospective work by incorporating balanced hemodynamic targets and monitoring hypertensive exposure as a safety outcome.
• Population Specificity: Unlike previous HPI trials in heterogeneous surgical populations, this trial isolates major aortic surgery, a high-risk cohort with distinct hemodynamic challenges (e.g., cross-clamping).
• Predictive vs. Reactive: It provides a direct, prospective comparison of machine learning-driven predictive monitoring against standard reactive monitoring within a rigorous RCT framework.
• Safety Focus: The protocol places significant emphasis on safety, specifically monitoring for intraoperative hypertension, acknowledging that aggressive hypotension avoidance may have unintended adverse consequences.

Results
As this document is a study protocol, it does not contain experimental results. The paper outlines the planned recruitment timeline (commencing April 2026, concluding March 2028) and the anticipated analysis of data expected by December 2028. The statistical power analysis indicates that with 200 participants, the study will have approximately 65–70% power to detect a 30% reduction in hypotensive burden, with the primary goal being the generation of a precise effect size estimate to inform future multicenter confirmatory trials.

Significance
The authors position the HYPE-AORTA trial as a necessary step to validate whether predictive hemodynamic monitoring offers clinically meaningful advantages over current reactive approaches in high-risk vascular surgery. The study aims to resolve uncertainty regarding the efficacy and safety of the Hypotension Prediction Index in a population where physiological stressors are extreme. Regardless of the outcome, the results are intended to inform optimal perioperative blood pressure management strategies. By incorporating safety monitoring for hypertension and refining the intervention protocol based on prior retrospective findings, the trial seeks to provide robust evidence that can guide the adoption of predictive algorithms in major aortic surgery, potentially reducing organ injury and mortality while avoiding the pitfalls of overcorrection.