Hypotension prediction index and postoperative complications: the need for risk-stratified patient selection.

The Hypotension Prediction Index is a validated machine learning algorithm based on numerous characteristics of the arterial waveform and aimed at predicting hypotension up to 15 min in advance.1 The recent special article by Enevoldsen and Vistisen focused on the data selection process for the development of this algorithm, highlighting the overrepresentation of the mean arterial pressure (MAP) within the Hypotension Prediction Index as substantiated with simulated data.2 This led us to perform a detailed comparison between the two signals in clinical practice, and investigate the relationship between Hypotension Predicting Index alarms and different simultaneously assessed MAP alarms, to further clarify the presumed added value of the Hypotension Prediction Index over MAP.In an observational pilot study, the Hypotension Prediction Index was used in addition to routine monitoring in the operating room on a convenience sample of 33 adult, high-risk noncardiac surgery patients requiring invasive blood pressure monitoring (local ethical committee waived approval [#K22-42], Medisch Spectrum Twente, Enschede, The Netherlands). Hemodynamic management was performed according to standard care, with nonblinded addition of the HemoSphere advanced monitor (version 2.1, Edwards Lifesciences, Inc., USA). The Hypotension Prediction Index and averaged MAP values were recorded every 20 s from start of incision until end of surgery, and no additional selection or filtering was applied.A cross-correlation analysis was performed between the normalized Hypotension Prediction Index and MAP signals. This signal analysis technique quantifies the level of similarity between two time series as a function of the time displacement relative to each other. This primary analysis was performed on each individual patient, and the group-level descriptive statistics are reported here (mean ± SD, when normally distributed, or median [25th to 75th interquartile ranges] otherwise). Second, we focused on the association between Hypotension Prediction Index alarms and different concurrent MAP alarms.The study population comprised 15 (45%) men, 18 (55%) American Society of Anesthesiologists (ASA; Schaumburg, Illinois) Physical Status II, and 15 (45%) ASA III patients. A total of 11 (33%) had previous cardiovascular disease, 18 (55%) hypertension, 6 (18%) diabetes, and 3 (9%) chronic obstructive pulmonary disease, with an age of 71 [53 to 76] yr. A total of 8,025.7 min of data were analyzed, with a surgery duration per patient of 244.2 ± 109.6 min. At least one Hypotension Prediction Index alarm (default threshold of 85 or greater) was detected in 28 (85%) patients, with a total alarm duration of 25.8 [8.1 to 41.3] min for this subgroup. Hypotension, defined as MAP less than 65 mmHg for at least 1 min, was recorded in 18 (55%) patients, with a total duration of 7.7 [4.2 to 17.8] min for this subgroup.The MAP and Hypotension Prediction Index signals show an inverted trend, i.e., when MAP decreases, the Hypotension Prediction Index value increases, predicting a low blood pressure (example in fig. 1). The maximum cross-correlation coefficient between MAP and Hypotension Prediction Index signals is –0.91 ± 0.04, with the relationship being negative due to the inverted trend mentioned (fig. 2). The maximum cross-correlation was detected at 0.0 ± 0.0 min time shift for all patients, indicating no time delay between the two signals.The Hypotension Prediction Index alarm started 3.0 [1.0 to 8.9] min before hypotension occurred, with a concurrent MAP of 71 [70 to 73] mmHg. When a Hypotension Prediction Index alarm was present, the MAP was 75 mmHg or less 94% of the time. When a threshold of MAP 70 mmHg or less was chosen, 98% of the time a Hypotension Prediction Index alarm was also present.The cross-correlation analysis shows that the Hypotension Prediction Index and MAP are highly correlated, while exhibiting no time delay. This is surprising, since one would expect the Hypotension Prediction Index to leverage multiple features of the preceding arterial waveforms to predict future MAP values. In this sense, our results substantiate the hypothesis by Enevoldsen and Vistisen that the predictive value of the Hypotension Prediction Index above the concurrent MAP may be limited.1 From our data, it is tempting to hypothesize that setting an alarm at a MAP approximately 70 to 75 mmHg (fig. 1) might yield a prediction of intraoperative hypotension comparable to the Hypotension Prediction Index. In line with our hypothesis, another MAP-based predictive method has recently been reported, using linear extrapolation on current and previous MAP values to predict future MAP.3We realize that the observational nature and limited sample size of this pilot study warrant caution when interpreting our results, including the inferred predictive validity of concurrent MAP alarms. Nevertheless, our results should instigate larger clinical studies to meticulously analyze the clinical performance of the Hypotension Prediction Index in comparison to intraoperative monitoring based solely on MAP.Support was provided solely from institutional and/or departmental sources.M. P. Mulder, Dr. Donker, and Dr. Fresiello provide research consultancy to Maquet Critical Care AB (Solna, Sweden) but do not receive personal fees. The other authors declare no competing interests.

Extreme hemodynamic changes, especially intraoperative hypotension (IOH), are common and often prolonged during Liver Transplant (LT) surgery and during initial hours of recovery. Hypotension Prediction Index (HPI) software is one of the tools which can help in proactive hemodynamic management. The accuracy of the advanced hemodynamic parameters such as Cardiac Output (CO) and Systemic Vascular Resistance (SVR) obtained from HPI software and prediction performance of the HPI in LT surgery remains unknown. This was a retrospective observational study conducted in a tertiary academic center with a large liver transplant program. We enrolled 23 adult LT patients who received both Pulmonary Artery Catheter (PAC) and HPI software monitoring. Primarily, we evaluated agreement between PAC and HPI software measured CO and SVR. A priori, we defined a relative difference of less than 20% between measurements as an adequate agreement for a pair of measurements and estimated the Lin's Concordance Correlation Coefficient and Bland-Altman Limits of Agreement (LOA). Clinically acceptable LOA was defined as ± 1 L.min-1 for CO and ± 200 dynes s.cm-5 for SVR. Secondary outcome was the ability of the HPI to predict future hypotension, defined as Mean Arterial Pressure (MAP) less than 65 mmHg lasting at least one minute. We estimated sensitivity, positive predictive value, and time from alert to hypotensive events for HPI software. Overall, 125 pairs of CO and 122 pairs of SVR records were obtained from 23 patients. Based on our predefined criteria, only 42% (95% CI 30%, 55%) of CO records and 53% (95% CI 28%, 72%) of SVR records from HPI software were considered to agree with those from PAC. Across all patients, there were a total of 1860 HPI alerts (HPI ≥ 85) and 642 hypotensive events (MAP < 65 mmHg). Out of the 642 hypotensive events, 618 events were predicted by HPI alert with sensitivity of 0.96 (95% CI: 0.95). Many times, the HPI value remained above alert level and was followed by multiple hypotensive events. Thus, to evaluate PPV and time to hypotension metric, we considered only the first HPI alert followed by a hypotensive event ("true alerts"). The "true alert" was the first alert when there were several alerts before a hypotension. There were 614 "true alerts" and the PPV for HPI was 0.33 (95% CI 0.31, 0.35). The median time from HPI alert to hypotension was 3.3 [Q1, Q3: 1, 9.3] mins. There was poor agreement between the pulmonary artery catheter and HPI software calculated advanced hemodynamic parameters (CO and SVR), in the patients undergoing LT surgery. HPI software had high sensitivity but poor specificity for hypotension prediction, resulting in a high burden of false alarms.

The "Hypotension Prediction Index (HPI)" represents a newly introduced monitoring-tool that aims to predict episodes of intraoperative hypotension (IOH) before their occurrence. In order to evaluate the feasibility of protocolized care according to HPI monitoring, we hypothesized that HPI predicts the incidence of IOH and reduces the incidence and duration of IOH. This single centre feasibility randomised blinded prospective interventional trial included at total of 99 patients. One group was managed by goal-directed therapy algorithm based on HPI (HPI, n = 25), which was compared to a routine anaesthetic care cohort (CTRL, n = 24) and a third historic control group (hCTRL, n = 50). Primary endpoints included frequency (n)/h, absolute and relative duration (t (min)/% of total anaesthesia time) of IOH. Significant reduction of intraoperative hypotension was recorded in the HPI group compared to the control groups (HPI 48%, CTRL 87.5%, hCTRL 80%; HPI vs. CTRL, respectively hCTRL p < 0.001). Perioperative quantity of IOH was significantly reduced in the interventional group compared to both other study groups (HPI: 0 (0-1), CTRL: 5 (2-6), hCTRL: 2 (1-3); p < 0.001). Same observations were identified for absolute (HPI: 0 (0-140) s, CTRL: 640 (195-1315) s, hCTRL 660 (180-1440) s; p < 0.001) and relative duration of hypotensive episodes (minutes MAP ≤ 65mmHg in % of total anaesthesia time; HPI: 0 (0-1), CTRL: 6 (2-12), hCTRL 7 (2-17); p < 0.001). The HPI algorithm combined with a protocolized treatment was able to reduce the incidence and duration of hypotensive events in patients undergoing primary hip arthroplasty.Trial registration: NCT03663270.

The Hypotension Prediction Index (HPI) is a validated algorithm developed by applying machine learning for predicting intraoperative arterial hypotension (IOH). We evaluated whether the HPI, combined with a personalized treatment protocol, helps to reduce IOH (depth and duration) and perioperative events in real practice. This was a single-center retrospective study including 104 consecutive adults undergoing urgent or elective non-cardiac surgery with moderate-to-high risk of bleeding, requiring invasive blood pressure and continuous cardiac output monitoring. Depending on the sensor, two comparable groups were identified: patients managed following the institutional protocol of personalized goal-directed fluid therapy (GDFT, n = 52), or this GDFT supported by the HPI (HPI, n = 52). The time-weighted average of hypotension for a mean arterial pressure < 65mmHg (TWAMAP<65), postoperative complications and length of hospital stay (LOS) were automatically downloaded from medical records and revised by clinicians blinded to the management received by patients. Differences in preoperative variables (i.e. physical status -ASA class-, acute kidney Injury-AKI- risk) and outcomes were analyzed using non-parametric tests with Hodges-Lehmann estimator for the median of differences. ASA class and AKI risk were similar (p = 0.749 and p = 0.837, respectively). Blood loss was also comparable (p = 0.279). HPI patients had a lower TWAMAP<65 [0.09mmHg (0-0.48mmHg)] vs [0.23mmHg (0.01 to 0.97mmHg)], p = 0.037. Postoperative complications were less prevalent in the HPI patients (0.46 ± 0.98 vs. 0.88 ± 1.20), p = 0.035. Finally, LOS was significantly shorter among HPI patients with a median difference of 2days (p = 0.019). The HPI combined with a GDFT protocol may help to minimize the severity of IOH during non-cardiac surgery.

Intraoperative hypotension during non-cardiac surgery is associated with postoperative complications such as acute kidney injury (AKI), myocardial injury, and stroke, which may increase mortality and severe adverse outcomes. Although the Hypotension Prediction Index (HPI) may help reduce intraoperative hypotension, its clinical value in lowering the incidence of these complications remains uncertain. This meta-analysis evaluates whether HPI-guided hemodynamic management reduces major postoperative complications (including AKI, cardiorenal, and cerebrovascular events) in adult patients undergoing non-cardiac surgery. A systematic search was conducted in PubMed, EMBASE, Cochrane Library, and Web of Science, to identify RCTs assessing HPI in non-cardiac surgery. The outcomes encompassed the incidence of postoperative complications such as AKI, myocardial injury after non-cardiac surgery (MINS), stroke and 30-day mortality. Pooled effect estimates, including odds ratios (ORs) with 95% confidence intervals (95% CIs), were calculated using either fixed-effects or random-effects models based on heterogeneity assessments. Sensitivity analyses were performed by excluding trials with a high or unclear risk of bias to evaluate the robustness of the findings. A total of 10 RCTs involving 1746 participants were included. The results revealed no statistically significant difference in incidence of AKI (OR: 0.85; 95%CI: 0.65 to 1.10; P=0.21), MINS(OR: 0.62; 95%CI: 0.36 to 1.06; P=0.08), stroke (OR: 0.63; 95%CI: 0.20 to 1.98; P=0.42), and 30-day mortality (OR: 0.87; 95%CI: 0.32 to 2.34; P=0.78) between HPI group and control group. Hemodynamic management guided by the HPI in adults undergoing non-cardiac surgery does not significantly reduce the incidence of major postoperative complications compared to standard care.

BackgroundIntraoperative hypotension (IOH), defined as a mean arterial pressure (MAP) below 65 mmHg, is a common complication during surgery and is associated with significant postoperative morbidity, including acute kidney injury, myocardial injury, stroke, and increased mortality. Despite the availability of traditional monitoring techniques, predicting and preventing IOH remains a challenge. The Hypotension Prediction Index (HPI), a machine-learning algorithm developed by Edwards Lifesciences, aims to predict IOH by analyzing real-time arterial waveform data, offering an opportunity for proactive management.ObjectiveThis systematic review and meta-analysis evaluate the efficacy of the HPI in predicting and preventing IOH in cardiac and non-cardiac surgeries compared to standard blood pressure monitoring techniques.MethodsA comprehensive search was conducted in PubMed, Scopus, Embase, and Web of Science databases for studies published from January 2019 to May 2024. Studies were included if they utilized machine learning algorithms, including HPI, to predict or detect IOH in adult surgical patients. Sensitivity, specificity, area under the curve (AUC), and time-weighted average (TWA) of hypotension were the primary outcomes. Subgroup analyses were performed to evaluate differences between cardiac and non-cardiac surgeries. Meta-analytic methods were applied using random-effects models to account for study variability.ResultsA total of 22 studies were included, encompassing both cardiac and non-cardiac procedures. The HPI demonstrated an overall sensitivity of 83% and specificity of 83% in predicting IOH. The pooled AUC for all surgeries was 0.90. However, subgroup analysis revealed variability in HPI performance between cardiac and non-cardiac surgeries, with lower diagnostic odds ratios (DOR) in cardiac settings. HPI combined with invasive arterial blood pressure monitoring reduced the TWA of hypotension more effectively than either invasive or non-invasive methods alone. The comparison of HPI and MAP for diagnostic accuracy showed minimal differences across all time frames, with SMD values close to zero.ConclusionOur study shows that the HPI has high sensitivity and specificity in predicting intraoperative hypotension, but its clinical advantage over standard MAP-based monitoring is uncertain. While HPI reduces hypotension duration, this may not improve cardiovascular or renal outcomes. Further independent trials are needed to validate its effectiveness before widespread adoption, and it should be considered alongside simpler interventions like staff education and MAP targeting in the meantime.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12871-025-03250-4.

Background: Hypotensive events may occur during surgical interventions and are associated with major postoperative complications, depending on their duration and severity. Intraoperative hemodynamic goal-directed therapy can reduce postoperative complications and mortality in high-risk surgeries and high-risk patients. The study hypothesis was that a proactive approach by hypotension predictive index (HPI) is more effective than a reactive goal-directed therapy (GDT) in reducing the number of hypotensive events during radical cystectomy and that this is associated with improved postoperative outcomes. Methods: The study was a single-center prospective observational study conducted at Galliera Hospital, from November 2019 to February 2025, with a before-after population of sixty-seven patients with reactive approach (GDT group) and sixty-five patients with a proactive approach (HPI group) undergoing radical cystectomy, managed with a standardized ERAS protocol and invasive or non-invasive hemodynamic monitoring. The aim of the study was to compare the incidence, duration, and severity of intraoperative hypotensive episodes between a proactive approach guided by the Hypotension Prediction Index (HPI) and a reactive goal-directed therapy (GDT) strategy guided by an advanced hemodynamic monitoring system. Results: The HPI group had a 65% reduction in hypotensive events (225 vs. 633, p < 0.001), with a 72% reduction in their duration (14 vs. 49 min, p < 0.001) and an 85% reduction in their severity (time-weighted average MAP < 65 mmHg 0.11 vs. 0.76, p < 0.001) compared to the GDT group. The HPI-guided group showed a reduction in postoperative infectious complications (10 vs. 26) and in-hospital length of stay (8 ± 4 versus 13 ± 8 days). Conclusions: A proactive approach may allow attenuating the occurrence and severity of hypotensive events more than a reactive goal-directed approach during radical cystectomy.

BACKGROUNDRobot-assisted radical prostatectomy (RARP) represents the leading application of robotic surgery in the treatment for prostate cancer with faster recovery and reduced invasiveness. Maintaining stable blood pressure during RARP is crucial to avoid complications. The approach used is goal-directed therapy (GDT); however, the Hypotension Prediction Index (HPI), a machine learning algorithm that analyses arterial waveforms, may provide additional benefits.OBJECTIVETo evaluate the cumulative amount, frequency and duration of intraoperative hypotension episodes in patients undergoing RARP, comparing those managed with a GDT protocol guided by the HPI algorithm versus those managed without it.DESIGNProspective randomised study.SETTINGSingle-centre university hospital. Recruitment from January 2022 to April 2023.PARTICIPANTSEighty-two patients.INTERVENTIONSPatients undergoing RARP were randomly assigned to either a GDT protocol alone (control) or a GDT protocol guided by the HPI (HPI group). All patients received both general anaesthesia and a single-shot spinal technique.MEAN OUTCOME MEASURESCumulative amount of intraoperative hypotension [measured using the time-weighted average (TWA) of mean arterial pressure (MAP) below 65 mmHg]; frequency of hypotensive events; duration of hypotensive events; postoperative complications; length of stay.RESULTSNo differences were observed in TWA-MAP, or in the frequency and duration of hypotensive events between the groups. Both groups maintained stable haemodynamics with minimal hypotension, and had similar fluid infusion and vasoconstrictor administration. Additionally, there were no differences in postoperative complications or length of stay.CONCLUSIONSIn our study, HPI guidance did not reduce intraoperative hypotension during RARP. Interestingly, the control group experienced fewer hypotensive events than typically reported in the literature, likely because of the high standards of haemodynamic and anaesthesiologic management maintained across both groups.TRIAL REGISTRATIONClinicaltrials.gov identifier: NCT06535464.

Pheochromocytomas and paragangliomas are uncommon neuroendocrine tumors that present notable anesthetic challenges, especially in controlling blood pressure. The Hypotension Prediction Index (HPI) can be used for intraoperative hemodynamic monitoring and management. Here, we report the case of a 46-year-old woman diagnosed with paraganglioma who underwent laparoscopic retroperitoneal resection guided by a monitoring system with HPI. Overall, mean arterial pressure (MAP) and HPI displayed an inverted relationship during the anesthesia induction period and subsequent phases of the procedure. Her MAP, initially around 90 mmHg, reached 120-150 mmHg while intubated and during tumor manipulation, and thus intermittent doses of phentolamine were necessary to control hypertensive events. MAP then declined following ligation of the tumor’s feeding vessels, accompanied by a corresponding rise in HPI values. Intraoperative hypotension was managed by HPI-based protocols, resulting in a very low time-weighted average-MAP (TWA-MAP) below 65 mmHg of 0.007 mmHg. While the minimum cross-correlation value between HPI and MAP occurred at a time lag of zero, indicating no delay between HPI and MAP, we found HPI alerted 3 to 17 minutes before MAP reached below 65 mmHg. We could add a new insight into the interpretation of cross-correlation analysis, that no time delay between HPI and MAP might not necessarily mean the predictive ability of HPI was low. Whereas this case highlights the potential of the HPI in mitigating intraoperative hypotension, future research is necessary to evaluate its predictive accuracy during paragangliomas resection.

The Hypotension Prediction Index (HPI) was recently introduced and clinically validated in different surgical conditions. This prospective observational study evaluated HPI's performance in living donor liver transplant recipients under the hypothesis that HPI would be inferior to the previously reported predictability in major surgery due to the surgical characteristics of liver transplantation. Twenty adult patients undergoing living donor liver transplantation were enrolled. HPI was monitored during the surgery with the attending anesthesiologist blinded to the HPI. The mean arterial pressure and HPI were recorded at 1-minute intervals. The area under the curve (AUC) of the receiver operating characteristic curve was calculated for the whole dataset and at each phase of liver transplantation at 5, 10, and 15 minutes to analyze HPI's performance. A total of 9173 data points were analyzed. The AUC for predicting hypotension at 5 minutes was 0.810 (95% confidence interval [CI]: 0.780-0.840). The AUCs for predicting hypotension at 10 and 15 minutes were 0.726 (95% CI: 0.681-0.772) and 0.689 (95% CI: 0.642-0.737), respectively. The AUCs for predicting hypotension at 5 minutes in the preanhepatic, anhepatic, and neohepatic phase were 0.795 (95% CI: 0.711-0.876), 0.728 (95% CI: 0.638-0.819), and 0.837 (95% CI: 0.802-0.873), respectively. The HPI's performance was inferior to that previously reported in major surgery. HPI in this observational study in living donor liver transplantation predicted hypotension with moderate-to-low accuracy, its predictive value being highest in the neohepatic phase and lowest in the anhepatic phase.

Intraoperative hypotension is associated with worse perioperative outcomes for patients undergoing major noncardiac surgery. The Hypotension Prediction Index is a unitless number that is derived from an arterial pressure waveform trace, and as the number increases, the risk of hypotension occurring in the near future increases. We investigated the diagnostic ability of the Hypotension Prediction Index in predicting impending intraoperative hypotension in comparison to other commonly collected perioperative hemodynamic variables. This is a 2-center retrospective analysis of patients undergoing major surgery. Data were downloaded and analyzed from the Edwards Lifesciences EV1000 platform. Receiver operating characteristic curves were constructed for the Hypotension Prediction Index and other hemodynamic variables as well as event rates and time to event. Two hundred fifty-five patients undergoing major surgery were included in the analysis yielding 292,025 data points. The Hypotension Prediction Index predicted hypotension with a sensitivity and specificity of 85.8% (95% CI, 85.8%-85.9%) and 85.8% (95% CI, 85.8%-85.9%) 5 minutes before a hypotensive event (area under the curve, 0.926 [95% CI, 0.925-0.926]); 81.7% (95% CI, 81.6%-81.8%) and 81.7% (95% CI, 81.6%-81.8%) 10 minutes before a hypotensive event (area under the curve, 0.895 [95% CI, 0.894-0.895]); and 80.6% (95% CI, 80.5%-80.7%) and 80.6% (95% CI, 80.5%-80.7%) 15 minutes before a hypotensive event (area under the curve, 0.879 [95% CI, 0.879-0.880]). The Hypotension Prediction Index performed superior to all other measured hemodynamic variables including mean arterial pressure and change in mean arterial pressure over a 3-minute window. The Hypotension Prediction Index provides an accurate real time and continuous prediction of impending intraoperative hypotension before its occurrence and has superior predictive ability than the commonly measured perioperative hemodynamic variables.

Clinical trials and validation studies demonstrate promising hypotension prediction capability by the Hypotension Prediction Index (HPI). Most studies that evaluate HPI derive it from invasive blood pressure readings, but a direct comparison with the noninvasive alternative remains undetermined. Such a comparison could provide valuable insights for clinicians in deciding between invasive and noninvasive monitoring strategies. Evaluating predictive differences between HPI when obtained through noninvasive versus invasive blood pressure monitoring. Post hoc analysis of a prospective observational study conducted between 2018 and 2020. Single-centre study conducted in an academic hospital in the Netherlands. Adult noncardiac surgery patients scheduled for over 2 h long elective procedures. After obtaining informed consent, 91 out of the 105 patients had sufficient data for analysis. The primary outcome was the difference in area under the receiver-operating characteristics (ROC) curve (AUC) obtained for HPI predictions between the two datasets. Additionally, difference in time-to-event estimations were calculated. AUC (95% confidence interval (CI)) results revealed a nonsignificant difference between invasive and noninvasive HPI, with areas of 94.2% (90.5 to 96.8) and 95.3% (90.4 to 98.2), respectively with an estimated difference of 1.1 (-3.9 to 6.1)%; P = 0.673. However, noninvasive HPI demonstrated significantly longer time-to-event estimations for higher HPI values. Noninvasive HPI is reliably accessible to clinicians during noncardiac surgery, showing comparable accuracy in HPI probabilities and the potential for additional response time. Clinicaltrials.gov (NCT03795831) https://clinicaltrials.gov/study/NCT03795831.

Introduction Percutaneous endovascular valvular interventions can result in profound hemodynamic instability, elevated burden of intraoperative hypotension (IOH) and related postoperative complications: ischemic stroke, acute kidney injury and increased mortality. Machine learning(ML),a branch of Artificial intelligence (AI), can analyze large volumes of data, find associations and allowing predictive rather than reactive interventions. The Hypotension Prediction Index (HPI)a ML derived algorithm,provides a unitless number from 0 to 100, that increases accordingly to the risk of developing a hypotensive event (mean arterial pressure – MAP – &amp;lt; 65 mmHg for more than 1 minute) in the following minutes. The aim of this study is to describe IOH in patients undergoing percutaneous valve repair under general anesthesia treated according to an HPI–based hemodynamic guidance (fig.1). Methods Eligible adult patients undergoing transcatheter valve repair procedures (MitraclipTM, TriclipTM)were included in the study. When HPI value exceeded 85,a proactive individualized treatment protocol to prevent hypotension was provided according to the following modalities (fig. 2). Primary outcome measure was TWA–MAP (time weighted average mean arterial pressure)under the threshold of 65 mmHg. Secondary outcomes were number of patients with at least one hypotensive event, number of events per patient,depth and duration of hypotensive events and area under MAP threshold of 65 mmHg (AUT–MAP &amp;lt; 65). Results Twenty–five consecutive patients were prospectively enrolled and treated. During an average monitoring time per patient of 187 ± 31 minutes, the global burden of hypotension, measured as TWA–MAP &amp;lt; 65 mmHg, was 0.12 [0.02, 0.8] mmHg. Two thirds of the patients(16/25) experienced hypotensive events, with a median number of hypotensive events of 1 [0, 3.25] per patient and about 11% of the time spent &amp;lt; 65 mmHg. Each event lasted 4 [1.7, 8.6] minutes with a MAP of 59 [56, 62] mmHg, leading to a total AUT–MAP &amp;lt; 65 mmHg of 20.3 [3.5, 142.2] mmHg x minutes. The majority of hypotensive events occurred after induction of general anesthesia, while hypotension was rare during the procedure (fig. 3). Conclusions HPI algorithm provides accurate and continuous prediction of impending IOH before its occurrence. Machine learning models,as in the case of HPI, could facilitate the physicians to treat IOH which is a potentially modifiable risk factor for major postoperative complications.

Background: Hypotension is common in the post-anesthesia care unit (PACU) and intensive care unit (ICU), and is associated with adverse patient outcomes. The Hypotension Prediction Index (HPI) algorithm has been shown to accurately predict hypotension in mechanically ventilated patients in the OR and ICU and to reduce intraoperative hypotension (IOH). Since positive pressure ventilation significantly affects patient hemodynamics, we performed this validation study to examine the performance of the HPI algorithm in a non-ventilated PACU and ICU population. Materials & Methods: The performance of the HPI algorithm was assessed using prospectively collected blood pressure (BP) and HPI data from a PACU and a mixed ICU population. Recordings with sufficient time (≥3 h) spent without mechanical ventilation were selected using data from the electronic medical record. All HPI values were evaluated for sensitivity, specificity, predictive value, and time-to-event, and a receiver operating characteristic (ROC) curve was constructed. Results: BP and HPI data from 282 patients were eligible for analysis, of which 242 (86%) were ICU patients. The mean age (standard deviation) was 63 (13.5) years, and 186 (66%) of the patients were male. Overall, the HPI predicted hypotension accurately, with an area under the ROC curve of 0.94. The most used HPI threshold cutoff in research and clinical use, 85, showed a sensitivity of 1.00, specificity of 0.79, median time-to-event of 160 s [60–380], PPV of 0.85, and NPV of 1.00. Conclusion: The absence of positive pressure ventilation and the influence thereof on patient hemodynamics does not negatively affect the performance of the HPI algorithm in predicting hypotension in the PACU and ICU. Future research should evaluate the feasibility and influence on hypotension and outcomes following HPI implementation in non-ventilated patients at risk of hypotension.

The Hypotension Prediction Index is a proprietary prediction model incorporated into a commercially available intraoperative hemodynamic monitoring system. The Hypotension Prediction Index uses multiple features of the arterial blood pressure waveform to predict hypotension. The index publication introducing the Hypotension Prediction Index describes the selection of training and validation data. Although precise details of the Hypotension Prediction Index algorithm are proprietary, the authors describe a selection process whereby a mean arterial pressure (MAP) less than 75 mmHg will always predict hypotension. We hypothesize that the data selection process introduced a systematic bias that resulted in an overestimation of the current MAP value's ability to predict future hypotension. Since current MAP is a predictive variable contributing to Hypotension Prediction Index, this exaggerated predictive performance likely also applies to the corresponding Hypotension Prediction Index value. Other existing validation studies appear similarly problematic, suggesting that additional validation work and, potentially, updates to the Hypotension Prediction Index model may be necessary.