Acumen Hypotension Prediction Index software

The Acumen Hypotension Prediction Index (HPI) software is a first-of-its-kind technology that provides you with information regarding the likelihood of a patient trending toward a hypotensive event.*

Multiple studies have shown that Acumen HPI software:

  • Achieves statistically significant reduction of hypotension when combined with a treatment protocol in noncardiac surgery vs. standard of care1,2
  • Demonstrates superior predictive abilities for hypotension than common hemodynamic parameters such as cardiac output (CO), stroke volume (SV), and changes in mean arterial pressure (MAP)3
  • Has proven and reliable accuracy4

See the findings from these clinical studies.

*A hypotensive event is defined as MAP <65 mmHg for a duration of at least one minute.
Acumen Hypotension Prediction Index software

Edwards Lifesciences has a heritage of partnering with clinicians to advance patient care. That heritage is at the core of the Acumen Hypotension Prediction Index software. Developed in partnership with clinicians across the world and the first in a new category of products, Acumen Hypotension Prediction Index software offers the only predictive monitoring parameter for hypotension that is available in Europe.

 Acumen Hypotension Prediction Index Software

This first-of-its-kind predictive decision support software detects the likelihood of a patient trending towards a hypotensive event before the event occurs, and provides you with insights to understand the root cause and inform a potential course of action for your patient.

Know more about the risks of hypotension.

Part of the Acumen intelligent decision support suite, the Acumen Hypotension Prediction Index software is unlocked with the Acumen IQ sensor.

Three key elements of the Acumen Hypotension Prediction Index software

HPI parameter

HPI Parameter

The HPI parameter displays as a value ranging from 0 to 100, with higher values indicating higher likelihood of a hypotensive event.

The proprietary algorithm - developed with machine learning, and using data from almost 59,000 hypotensive events and over 144,000 non-hypotensive events - detects potential of hypotensive trending of a patient's mean arterial pressure (MAP). The HPI parameter value is updated every 20 seconds, providing continuous predictive insights into developing hypotensive events.

The higher the value of the HPI parameter, the greater the likelihood a hypotensive event will occur.

The diagnostic performance of the HPI parameter was assessed through clinical validation studies:

Summary of Clinical Validation Studies

The full table of Results of Clinical Validation studies may be found in the Operator's Manual.

  1. Specificity: ratio of true negatives to total number of non-events (negatives) with a negative defined as a data point that is at least 20 minutes away from any hypotensive event
  2. NPV
  3. Sensitivity: ratio of true positives to total number of events (positives) with positive defined as data point that is at most 5 minutes prior to a hypotensive event.
  4. PPV

TP = True Positive   FP = False Positive
TN = True Negative FN = False Negative

HPI high alert popup

HPI high alert popup

The HPI high alert popup alerts you when your patient is trending toward or experiencing a hypotensive event.

If the HPI parameter value exceeds 85 for two consecutive 20-second updates or reaches 100 at any time, the HPI high alert popup window will appear, prompting you to review the patient hemodynamics using the HPI secondary screen.

HPI secondary screen

If your patient is trending toward a hypotensive event, or is experiencing a hypotensive event, you can investigate the root cause and proactively inform a potential course of action. The advanced hemodynamic pressure and flow parameters provided on the HPI secondary screen allow you an opportunity to investigate and identify the root cause of potentially developing hypotensive events.

HPI secondary screen

The HPI secondary screen is accessed through the HPI high alert popup, by touching the HPI Information Bar (when enabled); by pressing the button on the HPI Key Parameter, or at any time through the Clinical Actions menu on the monitor. Acumen Hypotension Prediction Index software secondary screen displays consolidated values for patient parameters of MAP, CO, SVR, PR, SV, and SVV/PPV as well as two additional indicators of contractility and afterload to provide a complete hemodynamic profile of the patient. The advanced hemodynamic parameters on the secondary screen are arranged visually by preload, contractility, and afterload.

Using these advanced hemodynamic parameters can provide you potential insights into the cause of a hypotensive event.

Preload

Preload

Stroke volume variation (SVV)

Stroke volume variation, the percent difference between minimum and maximum stroke volume (SV), during a respiratory cycle.

Preload

Pulse pressure variation (PPV)

Pulse pressure variation, the percent difference between minimum and maximum pulse pressure (PV), during a respiratory cycle.

Contractility

Contractility

Systolic slope (dP/dt)

Maximal upslope of the arterial pressure waveform from a peripheral artery.

Afterload

Afterload

Dynamic arterial elastance (Eadyn)

The ratio of pulse pressure variation to stroke volume variation (PPV/SVV).

HPI Brochure

*A hypotensive event is defined as MAP <65 mmHg for a duration of at least one minute.
Clinical Studies Summary

Acumen HPI software combined with a treatment protocol achieved statistically significant reduction in hypotension vs. standard of care1,2

Two randomized controlled trials have shown that using Acumen HPI software in combination with a hemodynamic treatment protocol significantly reduced the incidence and duration of hypotensive events in patients undergoing noncardiac surgery.1,2

HYPE trial results featured in JAMA1

Highlights from 2020 Wijnberge, et al.1

Publication in JAMA: “Effect of a Machine Learning–Derived Early Warning System for Intraoperative Hypotension vs Standard Care on Depth and Duration of Intraoperative Hypotension During Elective Noncardiac Surgery: The HYPE Randomized Clinical Trial”

  • Elective, noncardiac surgery patients monitored with Acumen HPI software had a median time of hypotension per patient of 8 minutes compared to 32.7 minutes in the control group
  • Time-weighted average of hypotension combines the duration and the severity of hypotension corrected for the total duration of the procedure. With Acumen HPI software, the study showed a median .38 mmHg difference between the interventional and control group.
  • The Acumen HPI software secondary screen provided insight into the potential root cause of that hypotension, enabling clinicians to identify the appropriate treatment course

HYPE trial: primary and secondary endpoints1

Table Image

HYPE trial: hemodynamic diagnostic guidance and treatment protocol12

The trial protocol is included below for reference purposes only as it relates to this particular study.

Diagram

In the HYPE trial, there was no significant difference in the cumulative dose of vasopressors or fluids, and hypotension was prevented without increasing the number of hypertensive events.1

Results from Acumen HPI implementation in total hip arthroplasty prospective trial2

Highlights from 2019 Schneck, et al.2

Publication in the Journal of Clinical Monitoring and Computing: "Hypotension Prediction Index based protocolized haemodynamic management reduces the incidence and duration of intraoperative hypotension in primary total hip arthroplasty: a single centre feasibility randomised blinded prospective interventional trial"

Acumen Hypotension Prediction Index software combined with protocolized treatment was shown to reduce the relative and absolute duration of hypotensive events in total hip arthroplasty patients, in comparison to a historical and prospective control group.

Trial: primary endpoints2

HPICTRLhCTRL
Number of hypotensive events per hour (n/hr) 0 5 2
Absolute IOH time (sec) 0 640 660
Relative IOH time (IOH time as % of total anesthesia time) 0 6 7
HPI=Hypotension Prediction Index software; CTRL=routine anaesthetic care cohort; hCTRL=historic control group
HPI group, n=25; CTRL, n=24; hCTRL, n=50. Total n=99
Graphs demonstrated the incidence (a), absolute (b), and relative duration (c) of hypotensive events in the three study cohorts; significant differences between the study groups are symbolized by **p < 0.01 and ***p < 0.001.

Total hip arthroplasty prospective trial: protocol2

The trial protocol is included below for reference purposes only as it relates to this particular study.

Diagram

Acumen HPI software had superior ability to predict hypotensive events than common hemodynamic parameters3

Highlights from 2019 Davies, et al.3

Publication in Anesthesia and Analgesia: "Ability of an Arterial Waveform Analysis–Derived Hypotension Prediction Index to Predict Future Hypotensive Events in Surgical Patients"

When compared with hemodynamic parameters such as SV, CO, SVV, and MAP, Acumen HPI software showed a higher predictive performance at 5 and 10 minutes before hypotension in this study3.

Prediction of hypotension at 5 minutes before an event3

Image

Prediction of hypotension at 10 minutes before an event3

Image

Receiver operating characteristic curves for HPI, CO, SV, MAP, PP, HR, SVV, and shock index for prediction hypotension
5 and 10 min before the event.

Acumen HPI software demonstrated high accuracy in predicting hypotension4

Highlights from Hatib, et al4

Publication in Anesthesiology: "Machine-learning Algorithm to Predict Hypotension Based on High-fidelity Arterial Pressure Waveform Analysis"

  • Predictive algorithms are usually assessed via a ROC curve, with the AUC showing the predictive power of the algorithm4,14
  • At 10 minutes before an event, Acumen HPI software predicted hypotension with a specificity and sensitivity of 89% and 90% respectively, and with an AUC of 0.95 in this study.4

ROC closest to the y-axis approaches a perfect model, with fewer false-positive and false-negative values13

Image

ROC at 10 minutes before a hypotensive event4

Image

AUC=area under curve; ROC=receiver operating characteristic.
Risk of hypotension

In noncardiac surgery patients, research findings have revealed strong associations between intraoperative hypotension and elevated risk of both acute kidney injury (AKI) and myocardial injury after noncardiac surgery (MINS).5-7

MINS — the most common cardiovascular complication that occurs after noncardiac surgery — is the leading cause of mortality within one month following surgery.5,8 It is a substantial public health issue.8

 Risk of hypotension

More than 200 million adults across the world will undergo noncardiac surgery annually, and this number continues to increase each year.9,10 Globally, an estimated 8 million patients over 45 years old — more than 1 in 12 patients — suffer myocardial injury after noncardiac surgery per year.8,10,11

Lowest MAP (mmHg)

Once a patient’s mean arterial pressure (MAP) reaches 65 mmHg, it only takes 10 minutes of exposure to see higher associations between intraoperative hypotension and MINS.5

Lowest MAP (mmHg)

Additionally, if a patient's MAP reaches 50 mmHg, it only takes one minute of exposure to see a significant escalation in risk of MINS making early identification of a hypotensive event critical.5

Acumen Hypotension Prediction Index software is the first predictive technology that provides you with information regarding the likelihood of a patient trending toward a hypotensive event* and assists you in understanding the root cause of deteriorating cardiovascular stability.

*A hypotensive event is defined as MAP <65 mmHg for a duration of at least one minute.
The Acumen IQ sensor unlocks the Acumen Hypotension Prediction Index feature
Acumen IQ

The Acumen IQ sensor — part of the minimally invasive family of hemodynamic sensors — unlocks the Acumen Hypotension Prediction Index software. The minimally invasive Acumen IQ sensor connects to any existing radial arterial line. The Acumen IQ system automatically updates advanced parameters every 20 seconds, reflecting rapid physiological changes in moderate-to high-risk surgery. Advanced hemodynamic parameters provided by the Acumen IQ sensor offer you continuous insight into your patient’s hemodynamic status.

More on Acumen IQ sensor

Acumen Analytics software

Acumen Analytics software enables you to download and analyze your patient’s data retrospectively to a PC. Monitoring sessions, including demographic data, can be downloaded from the HemoSphere advanced monitoring platform or EV1000 clinical platform into Acumen Analytics software for organization and analysis. Patient identifiers are omitted from the collection of data.

Acumen Analytics software
More on Acumen Analytics software

Available on the platform of the future

HPI secondary screen

HemoSphere advanced monitoring platform offers advanced hemodynamic parameters that can help guide you with proactive decision support in a range of clinical situations and settings so you can maintain optimal patient perfusion.

More on HemoSphere advanced monitoring platform

Edwards clinical education

Hemodynamic education empowering clinical advancement

With a long-term commitment to improving the quality of care for surgical and critical care patients through education, Edwards clinical education meets you no matter where you are in the learning process — with a continuum of resources and tools that continuously support you as you solve the clinical challenges facing you today, and in the future.

For more educational information

TopMedTalk podcasts on Intraoperative Hypotension (IOH)

The TopMedTalk podcast is a leading source of high value medical discussion and education in Anaesthesia, Perioperative Care and Enhanced Recovery. We've featured interviews with key opinion leaders, academics, medical practitioners and policy makers.

Listen below to podcasts on hypotension and Acumen Hypotension Prediction Index (HPI) software. You can also find these episodes of TopMedTalk through wherever you get your podcasts.

Podcast icon

Avoiding the hypotension hazard

The Acumen Hypotension Prediction Index (HPI) software takes 2.6 million features from a single waveform which it then applies to 150 million different wave forms looking for factors which can predict hypotension. It's the first fully approved foray into predictive analytics for the world of anaesthesia.

This conversation looks at how it works and how it is being adopted in detail.

Presented by Desiree Chappell with Monty Mythen and Feras Hatib PhD, Director, Research and Development, Algorithms and Signal Processing at Edwards Lifesciences and Dr Simon Davies, Consultant Anaesthetist at York Teaching Hospital NHS Foundation Trust.

Podcast icon

TopMedTalk with Tim Miller

The Perioperative Quality Initiative (POQI) have featured frequently on TopMedTalk, here we discuss their focus on hypotension and blood pressure.

Presented by Monty Mythen and Desiree Chappell with Joff Lacey, registrar in anaesthesia at St George's Hospital in London, currently undertaking a fellowship in Perioperative Medicine at UCLH as well as being a regular presenter on TopMedTalk and Tim Miller, Associate Professor of Anaesthesiology from Duke University Medical Centre.

Podcast icon

TopMedTalk with Feras Hatib

Streamed live from the 2019 Prehabilitation World Congress and Annual London Peri-Operative Medicine Congress. This is a break out session held at 'Montague on The Gardens', focusing on artificial intelligence and was sponsored by Edwards Lifesciences.

Presented by Desiree Chappell and Monty Mythen with their guest, Feras Hatib, PhD, Director, Research and Development, Algorithms and Signal Processing at Edwards Lifesciences.

Podcast icon

The AI Doctor will see you now...

Streamed live from the 2019 Prehabilitation World Congress and Annual London Peri-Operative Medicine Congress. This is a break out session held at 'Montague on The Gardens', focusing on artificial intelligence and was sponsored by Edwards Lifesciences.

Presented by Monty Mythen and Desiree Chappell, this is a short explanation of the technology in practice.

Contact us

The Edwards Advantage

We are committed to providing your institution, clinicians and staff with a high level of customer service and support to ensure seamless product implementation and ongoing use:


Contact a sales representative

Do you have feedback about our products? Click here

References:
  1. Wijnberge, M., Geerts, B., Hol, L., Lemmers, N., Mulder, M., Berge, P., Schenk, J., Terwindt, L., Hollman, M., Vlaar, A., Veelo, D. (2020) Effect of a Machine Learning-Derived Early Warning System for Intraoperative Hypotension vs Standard Care on Depth and Duration of Intraoperative Hypotension During Elective Noncardiac Surgery: The HYPE Randomized Clinical Trial. JAMA Online, February 17, 2020. doi:10.1001/jama.2020.0592 https://jamanetwork.com/journals/jama/article-abstract/2761469
  2. Schneck, E., Schulte, D., Habig, L., Ruhrmann, S., Edinger, F., Markmann, M., Habicher, M., Rickert, M., Koch, C., Sander, M. (2019) Hypotension Prediction Index based protocolized haemodynamic management reduces the incidence and duration of intraoperative hypotension in primary total hip arthroplasty: a single centre feasibility randomized blinded prospective interventional trial. Journal of Clinical Monitoring and Computing online, November 29, 2019. https://link.springer.com/article/10.1007/s10877-019-00433-6
  3. Davies SJ, Vistisen ST, Jian Z, et al. Ability of an arterial waveform analysis-derived hypotension prediction index to predict future hypotensive events in surgical patients. Anesth Analg 2020;doi: 10.1213/ANE.0000000000004121. https://journals.lww.com/anesthesia-analgesia/Citation/2020/02000/Ability_of_an_Arterial_Waveform_Analysis_Derived.16.aspx
  4. Hatib, F., Zhongping, J., Buddi, S., Lee, C., Settels, J., Sibert, K., Rinehart, J., Cannesson, M. (2018). Machine-learning Algorithm to Predict Hypotension Based on High-fidelity Arterial Pressure Waveform Analysis. Anesthesiology 129, 663-74. https://anesthesiology.pubs.asahq.org/article.aspx?articleid=2685008
  5. Salmasi, V., Maheshwari, K., Yang, G., Mascha, E.J., Singh, A., Sessler, D.I., & Kurz, A. (2017). Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney injury and myocardial injury. Anesthesiology, 126(1), 47-65.
  6. Sun, L.Y., Wijeysundera, D.N., Tait, G.A., & Beattie, W.S. (2015). Association of Intraoperative Hypotension with Acute Kidney Injury after Elective non-cardiac Surgery. Anesthesiology, 123(3), 515-523.
  7. Walsh, M., Devereaux, P.J., Garg, A.X., Kurz, A., Turan, A., Rodseth, R.N., Cywinski, J., Thabane, L., & Sessler, D.I. (2013). Relationship between Intraoperative Mean Arterial Pressure and Clinical Outcomes after non-cardiac Surgery. Anesthesiology, 119(3), 507-515.
  8. Kahn, Alonso-Coello and Devereaux, Myocardial injury after non-cardiac surgery, Curr Opin Cardiol, 2014, 29:307-311
  9. Devereaux and Sessler, Cardiac complications in patients undergoing major non-cardiac surgery, N Engl J Med, 2015, 373(23):2258-2269.
  10. Sellers, D., Srinivas, C., Djaiani, G. (2018). Cardiovascular complications after non-cardiac surgery. Anaesthesia, 73 (Suppl. 1), 34 - 42.
  11. van Waes, J., Nathoe, H., Graaff, J., Kemperman, H., de Borst, G., Peelen, L., van Klei, W. (2013). Myocardial Injury After Noncardiac Surgery and its Association With Short-Term Mortality. Circulation, 127, 2264 - 2271.
  12. Wijnberge et al. HYPE Protocol supplement: Intraoperative implementation of the hypotension probability indicator (HPI) algorithm – A pilot randomized controlled clinical trial, Supplemental to Effect of a machine learning-derived early warning system for intraoperative hypotension vs standard care on depth and duration of intraoperative hypotension during elective noncardiac surgery: The HYPE Randomized Clinical Trial. JAMA 2020.16 Appendix I.
  13. Vining, David & Gladish, Gregory. (1992). Receiver operating characteristic curves: a basic understanding. Radiographics : a review publication of the Radiological Society of North America, Inc. 12. 1147-54. 10.1148/radiographics.12.6.1439017.
  14. Huang, Jin, and C.X. Ling. (2005) Using AUC and Accuracy in Evaluating Learning Algorithms. IEEE Transactions on Knowledge and Data Engineering, vol. 17, no. 3, 2005, pp. 299–310., doi:10.1109/tkde.2005.50.

For professional use

For professional use

For a listing of indications, contraindications, precautions, warnings, and potential adverse events, please refer to the Instructions for Use (consult eifu.edwards.com where applicable).

Edwards Lifesciences devices placed on the European market meeting the essential requirements referred to in Article 3 of the Medical Device Directive 93/42/EEC bear the CE marking of conformity.

Please update your browserClose this window

Please update to a current version of your preferred browser, this site will perform effectively on the following:

Unable to update your browser?

If you are on a computer, that is maintained by an admin and you cannot install a new browser, ask your admin about it. If you can't change your browser because of compatibility issues, think about installing a second browser for browsing and keep this old one for compatibility