HPI Smart Alerts and Smart Trends to Improve Compliance with an Intra-operative Hemodynamic Management Algorithm and Reduce Renal Injury Following Major Elective Noncardiac Surgery: The SMART TRENDS Feasibility Study

Document Type

Conference Proceeding

Publication Date

5-1-2025

Publication Title

Anesth Analg

Keywords

Anesthesiology

Abstract

Introduction: Avoidance of intraoperative hypotension is a well accepted goal in clinical practice.1 In major non-cardiac surgery, goal-directed hemodynamic therapy (GDHT) is used to improve outcomes.2 Trials of GDHT have complex algorithms with variable degrees of compliance, reported to be around 50%.3,4 The Hypotension Prediction Index (HPI, Edwards Lifesciences, Irvine, CA) is a machine learning-derived early warning system that predicts hypotension and analyzes hemodynamic mechanisms. Recently, HPI Software incorporated smart trends and alerts (HPI Smart Alerts) to highlight the mechanism(s) of hypotension and facilitate the interpretation of the hemodynamic profile and adherence to decision support algorithms. We hypothesized that prompts from HPI Smart Alerts, triggered by HPI ≥ 85 (indicating high risk of hypotension), would result in improved compliance with an intraoperative GDHT algorithm and consequently a lower incidence of acute kidney injury (AKI) than previously reported.3: Methods: This was an IRB-approved, multicenter, mixed‐: methods, single‐arm feasibility study enrolling participants at 8 centers in the United States. It was registered at ClinicalTrials.gov (NCT05957406). We included adults, ASA ≥ 2, undergoing elective, noncardiac, moderate-to-high-risk surgery with an expected duration of ≥ 3 hours, planned intraarterial blood pressure monitoring, and an expected postoperative length of stay ≥ 3 days. Following induction of general anesthesia and institution of mechanical ventilation, the GDHT protocol began with stroke volume optimization via intravenous fluid challenge. Subsequent goals of the GDHT algorithm included actionable trends at HPI >50 and automated HPI Smart Alert notifications at HPI ≥ 85 displaying the most likely mechanism in common physiologic language (i.e., preload, afterload, contractility). All providers received in-person teaching and plastic cards with easy-to-understand language and appropriate treatment choices. The primary endpoint was the proportion of HPI Smart Alert notifications resulting in timely (<5 min) initiation of intervention consistent with the algorithm. Secondary endpoints included AKI per the modified KDIGO definition and the incidence and magnitude of hypotension (MAP <65mmHg). Sample size estimation was based on achieving a compliance rate of at least 75%, where the lower limit for the 95% confidence interval needed to be greater than 60%. A bootstrap approach was taken to calculate the 95% confidence interval and the p‐value for the test because the observations for each subject were not necessarily independent (N=122, ~12 participants per center). Our a priori sample size was also estimated to have 90% power to demonstrate a rate of AKI below 15% as previously reported.3: Results: A total of 122 participants were included in the study. 113 completed the GDHT protocol. The enrollment among the centers was equal. Table 1 presents the baseline characteristics. The mean and standard deviation (SD) of age and BMI were 62.7 years (12.18) and 29.9 Kg/M2 (6.49), respectively. The proportion of females was 45.9%. The most common ASA classification was III (74.6%). An open surgical approach was used in 76.9% of the cases. The overall compliance rate in the entire cohort was 77.1% (95% CI: 74.5-79.6%). The median (interquartile range, IQR) total minutes under MAP 65mmHg was 2.9 (0.4-8.7) min. Similarly, the median (IQR) time-weighted average of MAP <65mmHg was 0.0 (0.0 - 0.1) mmHg. The median (IQR) area under the curve <65mmHg was 8.8 (0.3 – 33.7) mmHg*min. The incidence of postoperative AKI in this study was significantly lower (4.4%; 95% CI: 1.4-9.9, p<0.001) compared to a previously reported rate of 15%3 (Table 2). No device-related adverse events were reported.Conclusions: In a well monitored setting, this trial achieved high compliance with guidance provided by an HPI Smart Alerts intraoperative GDHT algorithm in adults undergoing moderate to high-risk non-cardiac surgery. The duration and degree of hypotension, as well as the incidence of postoperative AKI, were lower than previously reported. These results support larger scale studies of GDHT using HPI Smart Alerts and its effect on clinical outcomes.

Volume

140

Issue

5

First Page

1026

Last Page

1027

Find It @ Sladen

Share

COinS