Medicine (Baltimore). 2026 Jun 12;105(24):e49257. doi: 10.1097/MD.0000000000049257.
ABSTRACT
BACKGROUND: The Hypotension Prediction Index (HPI) is a machine-learning-derived early-warning algorithm that uses real-time arterial pressure waveform features to prospectively identify episodes of intraoperative hypotension (IOH) - defined as mean arterial pressure < 65 mm Hg sustained for ≥ 1 minute. IOH is a common hemodynamic complication in surgical patients and is associated with impaired end-organ perfusion and an increased risk of postoperative morbidity. To rigorously evaluate the predictive performance of HPI and its impact on IOH, we conducted a systematic review and meta-analysis of randomized controlled trials, adhering strictly to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 statement.
METHODS: A systematic literature search was conducted across multiple electronic databases from their inception to April 10, 2026. The search strategy employed a combination of keyword terms, including "hypotension," "prediction," "index," "surgery," and "randomized controlled trial." Primary outcomes were the time-weighted average, area under the threshold, incidence, and duration of IOH. Secondary outcomes included intraoperative fluid administration, estimated blood loss, urine output, use of vasopressors and inotropes, and postoperative acute kidney injury (AKI). Pooled effect estimates were calculated using odds ratios for dichotomous outcomes and weighted mean differences for continuous outcomes, with corresponding 95% confidence intervals.
RESULTS: Eighteen randomized controlled trials (RCTs) involving 2279 noncardiac surgical patients and 1 RCT enrolling 130 cardiac surgical patients were included in the analysis. Following standardized data conversion, meta-analytic results demonstrated that HPI-guided hemodynamic management significantly reduced IOH across multiple complementary metrics: time-weighted average (WMD = -0.19, 95% CI: -0.25 to - 0.12), area under the threshold (WMD = -55.17, 95% CI: -73.62 to - 36.71), incidence (WMD = -2.42, 95% CI: -3.36 to - 1.49), cumulative duration (WMD = -10.61, 95% CI: -14.11 to - 7.11), all P < .00001. Similarly, the HPI-guided group reduced intraoperative fluid administration and phenylephrine utilization. In contrast, the incidence of AKI showed no statistically significant difference (OR = 0.83, 95% CI: 0.64 to 1.07, P = .16).
CONCLUSION: Current evidence shows that HPI-directed hemodynamic management effectively reduces IOH and decreases intraoperative fluid administration and phenylephrine utilization. However, these physiological improvements did not translate into clinical benefits for postoperative AKI.
PMID:42299539 | DOI:10.1097/MD.0000000000049257