Ability of an Arterial Waveform Analysis-Derived Hypotension Prediction Index to Predict Future Hypotensive Events in Surgical Patients

Individualized Hemodynamic Management in Sepsis

Hemodynamic optimization remains the cornerstone of resuscitation in the treatment of sepsis and septic shock. Delay or inadequate management will inevitably lead to hypoperfusion, tissue hypoxia or edema, and fluid overload, leading eventually to multiple organ failure, seriously affecting outcomes. According to a large international survey (FENICE study), physicians frequently use inadequate indices to guide fluid management in intensive care units. Goal-directed and "restrictive" infusion strategies have been recommended by guidelines over "liberal" approaches for several years. Unfortunately, these "fixed regimen" treatment protocols neglect the patient's individual needs, and what is shown to be beneficial for a given population may not be so for the individual patient. However, applying multimodal, contextualized, and personalized management could potentially overcome this problem. The aim of this review was to give an insight into the pathophysiological rationale and clinical application of this relatively new approach in the hemodynamic management of septic patients.

Physiology of Human Hemorrhage and Compensation

Hemorrhage is a leading cause of death following traumatic injuries in the United States. Much of the previous work in assessing the physiology and pathophysiology underlying blood loss has focused on descriptive measures of hemodynamic responses such as blood pressure, cardiac output, stroke volume, heart rate, and vascular resistance as indicators of changes in organ perfusion. More recent work has shifted the focus toward understanding mechanisms of compensation for reduced systemic delivery and cellular utilization of oxygen as a more comprehensive approach to understanding the complex physiologic changes that occur following and during blood loss. In this article, we begin with applying dimensional analysis for comparison of animal models, and progress to descriptions of various physiological consequences of hemorrhage. We then introduce the complementary side of compensation by detailing the complexity and integration of various compensatory mechanisms that are activated from the initiation of hemorrhage and serve to maintain adequate vital organ perfusion and hemodynamic stability in the scenario of reduced systemic delivery of oxygen until the onset of hemodynamic decompensation. New data are introduced that challenge legacy concepts related to mechanisms that underlie baroreflex functions and provide novel insights into the measurement of the integrated response of compensation to central hypovolemia known as the compensatory reserve. The impact of demographic and environmental factors on tolerance to hemorrhage is also reviewed. Finally, we describe how understanding the physiology of compensation can be translated to applications for early assessment of the clinical status and accurate triage of hypovolemic and hypotensive patients. © 2021 American Physiological Society. Compr Physiol 11:1531-1574, 2021.

Ensemble machine learning prediction and variable importance analysis of 5-year mortality after cardiac valve and CABG operations

Despite having a similar post-operative complication profile, cardiac valve operations are associated with a higher mortality rate compared to coronary artery bypass grafting (CABG) operations. For long-term mortality, few predictors are known. In this study, we applied an ensemble machine learning (ML) algorithm to 88 routinely collected peri-operative variables to predict 5-year mortality after different types of cardiac operations. The Super Learner algorithm was trained using prospectively collected peri-operative data from 8241 patients who underwent cardiac valve, CABG and combined operations. Model performance and calibration were determined for all models, and variable importance analysis was conducted for all peri-operative parameters. Results showed that the predictive accuracy was the highest for solitary mitral (0.846 [95% CI 0.812–0.880]) and solitary aortic (0.838 [0.813–0.864]) valve operations, confirming that ensemble ML using routine data collected perioperatively can predict 5-year mortality after cardiac operations with high accuracy. Additionally, post-operative urea was identified as a novel and strong predictor of mortality for several types of operation, having a seemingly additive effect to better known risk factors such as age and postoperative creatinine.

Deep learning models for the prediction of intraoperative hypotension

BACKGROUND

Intraoperative hypotension is associated with a risk of postoperative organ dysfunction. In this study, we aimed to present deep learning algorithms for real-time predictions 5, 10, and 15 min before a hypotensive event.

METHODS

In this retrospective observational study, deep learning algorithms were developed and validated using biosignal waveforms acquired from patient monitoring of noncardiac surgery. The classification model was a binary classifier of a hypotensive event (MAP <65 mm Hg) or a non-hypotensive event by analysing biosignal waveforms. The regression model was developed to directly estimate the MAP. The primary outcome was area under the receiver operating characteristic (AUROC) curve and the mean absolute error (MAE).

RESULTS

In total, 3301 patients were included. For invasive models, the multichannel model with an arterial pressure waveform, electrocardiography, photoplethysmography, and capnography showed greater AUROC than the arterial-pressure-only models (AUROC_15-min, 0.897 [95% confidence interval {CI}: 0.894-0.900] vs 0.891 [95% CI: 0.888-0.894]) and lesser MAE (MAE_15-min, 7.76 mm Hg [95% CI: 7.64-7.87 mm Hg] vs 8.12 mm Hg [95% CI: 8.02-8.21 mm Hg]). For the noninvasive models, the multichannel model showed greater AUROCs than that of the photoplethysmography-only models (AUROC_15-min, 0.762 [95% CI: 0.756-0.767] vs 0.694 [95% CI: 0.686-0.702]) and lesser MAEs (MAE_15-min, 11.68 mm Hg [95% CI: 11.57-11.80 mm Hg] vs 12.67 [95% CI: 12.56-12.79 mm Hg]).

CONCLUSIONS

Deep learning models can predict hypotensive events based on biosignals acquired using invasive and noninvasive patient monitoring. In addition, the model shows better performance when using combined rather than single signals.

Use of the Hypotension Prediction Index During Cardiac Surgery

OBJECTIVE

The hypotension prediction index (HPI) is a novel parameter developed by Edwards Lifesciences (Irvine, CA) that is obtained through an algorithm based on arterial pressure waveform characteristics. Past studies have demonstrated its accuracy in predicting hypotensive events in noncardiac surgeries. The authors aimed to evaluate the use of the HPI in cardiac surgeries requiring cardiopulmonary bypass (CPB).

DESIGN

Prospective cohort feasibility study.

SETTING

Single university medical center.

PARTICIPANTS

Sequential adult patients undergoing elective cardiac surgeries requiring CPB between October 1, 2018, and December 31, 2018.

INTERVENTIONS

HPI monitor was connected to the patient's arterial pressure transducer. Anesthesiologists and surgeons were blinded to the monitor output.

MEASUREMENTS AND MAIN RESULTS

HPI values and hypotensive events were recorded before and after CPB. The primary outcomes were the area under the curve (AUC) of the receiver operating characteristic curve, sensitivity, and specificity of HPI predicting hypotension. The AUC, sensitivity, and specificity for HPI lead time to hypotension five minutes before the event were 0.90 (95% confidence interval [CI]: 0.853-0.949), 84% (95% CI: 77.7-90.5), and 84% (95% CI: 70.9-96.8), respectively. Ten minutes before the event AUC, sensitivity, and specificity for HPI lead time to hypotension were 0.83 (95% CI: 0.750-0.905), 79% (95% CI: 69.8-88.1), and 74% (95% CI: 58.8-89.6), respectively. Fifteen minutes before the hypotensive event AUC, sensitivity, and specificity for HPI lead time to hypotension were 0.83 (95% CI: 0.746-0.911), 79% (95% CI: 68.4-89.0), and 74% (95% CI: 58.8-89.6), respectively.

CONCLUSION

HPI predicted hypotensive episodes during cardiac surgeries with a high degree of sensitivity and specificity.

One of the first validations of an artificial intelligence algorithm for clinical use: The impact on intraoperative hypotension prediction and clinical decision-making

This review describes the steps and conclusions from the development and validation of an artificial intelligence algorithm (the Hypotension Prediction Index), one of the first machine learning predictive algorithms used in the operating room environment. The algorithm has been demonstrated to reduce intraoperative hypotension in two randomized controlled trials via real-time prediction of upcoming hypotensive events prompting anesthesiologists to act earlier, more often, and differently in managing impending hypotension. However, the algorithm entails no dynamic learning process that evolves from use in clinical patient care, meaning the algorithm is fixed, and furthermore provides no insight into the decisional process that leads to an early warning for intraoperative hypotension, which makes the algorithm a "black box." Many other artificial intelligence machine learning algorithms have these same disadvantages. Clinical validation of such algorithms is relatively new and requires more standardization, as guidelines are lacking or only now start to be drafted. Before adaptation in clinical practice, impact of artificial intelligence algorithms on clinical behavior, outcomes and economic advantages should be studied too.

The role of artificial intelligence in surgical patient perioperative management

Perioperative medicine is a patient-centered, multidisciplinary and integrated clinical practice that starts from the moment of contemplation of surgery until full recovery. Every perioperative phase (preoperative, intraoperative and postoperative) must be studied and planned in order to optimize the entire patient management. Perioperative optimization does not only concern a short-term outcome improvement, but it has also a strong impact on long term survival. Clinical cases variability leads to the collection and analysis of a huge amount of different data, coming from multiple sources, making perioperative management standardization very difficult. Artificial Intelligence (AI) can play a primary role in this challenge, helping human mind in perioperative practice planning and decision-making process. AI refers to the ability of a computer system to perform functions and reasoning typical of the human mind; Machine Learning (ML) could play a fundamental role in presurgical planning, during intraoperative phase and postoperative management. Perioperative medicine is the cornerstone of surgical patient management and the tools deriving from the application of AI seem very promising as a support in optimizing the management of each individual patient. Despite the increasing help that will derive from the use of AI tools, the uniqueness of the patient and the particularity of each individual clinical case will always keep the role of the human mind central in clinical and perioperative management. The role of the physician, who must analyze the outputs provided by AI by following his own experience and knowledge, remains and will always be essential.

The value of hypotensive prediction index and dP/dtmax to predict and treat hypotension in a patient with a dilated cardiomyopathy

The Hypotension Prediction Index (HPi) is a new parameter, recently developed to predict the risk of a patient developing a hypotensive event, defined as a fall in mean arterial pressure below 65 mmHg. The calculated HPi value is displayed on a monitor as a number ranging from 1 to 100; where the first warning for the appearance of such event occurs when HPi values exceed 85. A secondary screen shows the stroke volume variation value; the dP/dt max; and the dynamic arterial elastance. We described a patient with a mild to moderately dilated cardiomyopathy that presented several episodes of hypotension after induction of anaesthesia and how by using HPi technology, these were successfully solved. We recommend the use of a HPi value >85 as a warning of intervention, and to use the secondary screen to determine the cause and the treatment. We consider that HPi technology may be a valid alternative for the anaesthetic management of patients with a dilated cardiomyopathy.

Prediction of hypotension events with physiologic vital sign signatures in the intensive care unit

BACKGROUND

Even brief hypotension is associated with increased morbidity and mortality. We developed a machine learning model to predict the initial hypotension event among intensive care unit (ICU) patients and designed an alert system for bedside implementation.

MATERIALS AND METHODS

From the Medical Information Mart for Intensive Care III (MIMIC-3) dataset, minute-by-minute vital signs were extracted. A hypotension event was defined as at least five measurements within a 10-min period of systolic blood pressure ≤ 90 mmHg and mean arterial pressure ≤ 60 mmHg. Using time series data from 30-min overlapping time windows, a random forest (RF) classifier was used to predict risk of hypotension every minute. Chronologically, the first half of extracted data was used to train the model, and the second half was used to validate the trained model. The model's performance was measured with area under the receiver operating characteristic curve (AUROC) and area under the precision recall curve (AUPRC). Hypotension alerts were generated using risk score time series, a stacked RF model. A lockout time were applied for real-life implementation.

RESULTS

We identified 1307 subjects (1580 ICU stays) as the hypotension group and 1619 subjects (2279 ICU stays) as the non-hypotension group. The RF model showed AUROC of 0.93 and 0.88 at 15 and 60 min, respectively, before hypotension, and AUPRC of 0.77 at 60 min before. Risk score trajectories revealed 80% and > 60% of hypotension predicted at 15 and 60 min before the hypotension, respectively. The stacked model with 15-min lockout produced on average 0.79 alerts/subject/hour (sensitivity 92.4%).

CONCLUSION

Clinically significant hypotension events in the ICU can be predicted at least 1 h before the initial hypotension episode. With a highly sensitive and reliable practical alert system, a vast majority of future hypotension could be captured, suggesting potential real-life utility.

Wearable Sensors Incorporating Compensatory Reserve Measurement for Advancing Physiological Monitoring in Critically Injured Trauma Patients

Vital signs historically served as the primary method to triage patients and resources for trauma and emergency care, but have failed to provide clinically-meaningful predictive information about patient clinical status. In this review, a framework is presented that focuses on potential wearable sensor technologies that can harness necessary electronic physiological signal integration with a current state-of-the-art predictive machine-learning algorithm that provides early clinical assessment of hypovolemia status to impact patient outcome. The ability to study the physiology of hemorrhage using a human model of progressive central hypovolemia led to the development of a novel machine-learning algorithm known as the compensatory reserve measurement (CRM). Greater sensitivity, specificity, and diagnostic accuracy to detect hemorrhage and onset of decompensated shock has been demonstrated by the CRM when compared to all standard vital signs and hemodynamic variables. The development of CRM revealed that continuous measurements of changes in arterial waveform features represented the most integrated signal of physiological compensation for conditions of reduced systemic oxygen delivery. In this review, detailed analysis of sensor technologies that include photoplethysmography, tonometry, ultrasound-based blood pressure, and cardiogenic vibration are identified as potential candidates for harnessing arterial waveform analog features required for real-time calculation of CRM. The integration of wearable sensors with the CRM algorithm provides a potentially powerful medical monitoring advancement to save civilian and military lives in emergency medical settings.

Automated Blood Pressure Control

Arterial pressure management is a crucial task in the operating room and intensive care unit. In high-risk surgical and in critically ill patients, sustained hypotension is managed with continuous infusion of vasopressor agents, which most commonly have direct α agonist activity like phenylephrine or norepinephrine. The current standard of care to guide vasopressor infusion is manual titration to an arterial pressure target range. This approach may be improved by using automated systems that titrate vasopressor infusions to maintain a target pressure. In this article, we review the evidence behind blood pressure management in the operating room and intensive care unit and discuss current and potential future applications of automated blood pressure control.

The application of a neural network to predict hypotension and vasopressor requirements non-invasively in obstetric patients having spinal anesthesia for elective cesarean section (C/S)

Background

Neural networks are increasingly used to assess physiological processes or pathologies, as well as to predict the increased likelihood of an impending medical crisis, such as hypotension.

Method

We compared the capabilities of a single hidden layer neural network of 12 nodes to those of a discrete-feature discrimination approach with the goal being to predict the likelihood of a given patient developing significant hypotension under spinal anesthesia when undergoing a Cesarean section (C/S). Physiological input information was derived from a non-invasive blood pressure device (Caretaker [CT]) that utilizes a finger cuff to measure blood pressure and other hemodynamic parameters via pulse contour analysis. Receiver-operator-curve/area-under-curve analyses were used to compare performance.

Results

The results presented here suggest that a neural network approach (Area Under Curve [AUC] = 0.89 [p < 0.001]), at least at the implementation level of a clinically relevant prediction algorithm, may be superior to a discrete feature quantification approach (AUC = 0.87 [p < 0.001]), providing implicit access to a plurality of features and combinations thereof. In addition, the expansion of the approach to include the submission of other physiological data signals, such as heart rate variability, to the network can be readily envisioned.

Conclusion

This pilot study has demonstrated that increased coherence in Arterial Stiffness (AS) variability obtained from the pulse wave analysis of a continuous non-invasive blood pressure device appears to be an effective predictor of hypotension after spinal anesthesia in the obstetrics population undergoing C/S.

This allowed us to predict specific dosing thresholds of phenylephrine required to maintain systolic blood pressure above 90 mmHg.

Dynamic Arterial Elastance as a Ventriculo-Arterial Coupling Index: An Experimental Animal Study

Dynamic arterial elastance (Ea_dyn), the ratio between arterial pulse pressure and stroke volume changes during respiration, has been postulated as an index of the coupling between the left ventricle (LV) and the arterial system. We aimed to confirm this hypothesis using the gold-standard for defining LV contractility, afterload, and evaluating ventricular-arterial (VA) coupling and LV efficiency during different loading and contractile experimental conditions. Twelve Yorkshire healthy female pigs submitted to three consecutive stages with two opposite interventions each: changes in afterload (phenylephrine/nitroprusside), preload (bleeding/fluid bolus), and contractility (esmolol/dobutamine). LV pressure-volume data was obtained with a conductance catheter, and arterial pressures were measured via a fluid-filled catheter in the proximal aorta and the radial artery. End-systolic elastance (Ees), a load-independent index of myocardial contractility, was calculated during an inferior vena cava occlusion. Effective arterial elastance (Ea, an index of LV afterload) was calculated as LV end-systolic pressure/stroke volume. VA coupling was defined as the ratio Ea/Ees. LV efficiency (LV_eff) was defined as the ratio between stroke work and the LV pressure-volume area. Ea_dyn was calculated as the ratio between the aortic pulse pressure variation (PPV) and conductance-derived stroke volume variation (SVV). A linear mixed model was used for evaluating the relationship between Ees, Ea, VA coupling, LV_eff with Ea_dyn. Ea_dyn was inversely related to VA coupling and directly to LV_eff. The higher the Ea_dyn, the higher the LV_eff and the lower the VA coupling. Thus, Ea_dyn, an easily measured parameter at the bedside, may be of clinical relevance for hemodynamic assessment of the unstable patient.

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

IMPORTANCE

Intraoperative hypotension is associated with increased morbidity and mortality. A machine learning-derived early warning system to predict hypotension shortly before it occurs has been developed and validated.

OBJECTIVE

To test whether the clinical application of the early warning system in combination with a hemodynamic diagnostic guidance and treatment protocol reduces intraoperative hypotension.

DESIGN, SETTING, AND PARTICIPANTS

Preliminary unblinded randomized clinical trial performed in a tertiary center in Amsterdam, the Netherlands, among adult patients scheduled for elective noncardiac surgery under general anesthesia and an indication for continuous invasive blood pressure monitoring, who were enrolled between May 2018 and March 2019. Hypotension was defined as a mean arterial pressure (MAP) below 65 mm Hg for at least 1 minute.

INTERVENTIONS

Patients were randomly assigned to receive either the early warning system (n = 34) or standard care (n = 34), with a goal MAP of at least 65 mm Hg in both groups.

MAIN OUTCOMES AND MEASURES

The primary outcome was time-weighted average of hypotension during surgery, with a unit of measure of millimeters of mercury. This was calculated as the depth of hypotension below a MAP of 65 mm Hg (in millimeters of mercury) × time spent below a MAP of 65 mm Hg (in minutes) divided by total duration of operation (in minutes).

RESULTS

Among 68 randomized patients, 60 (88%) completed the trial (median age, 64 [interquartile range {IQR}, 57-70] years; 26 [43%] women). The median length of surgery was 256 minutes (IQR, 213-430 minutes). The median time-weighted average of hypotension was 0.10 mm Hg (IQR, 0.01-0.43 mm Hg) in the intervention group vs 0.44 mm Hg (IQR, 0.23-0.72 mm Hg) in the control group, for a median difference of 0.38 mm Hg (95% CI, 0.14-0.43 mm Hg; P = .001). The median time of hypotension per patient was 8.0 minutes (IQR, 1.33-26.00 minutes) in the intervention group vs 32.7 minutes (IQR, 11.5-59.7 minutes) in the control group, for a median difference of 16.7 minutes (95% CI, 7.7-31.0 minutes; P < .001). In the intervention group, 0 serious adverse events resulting in death occurred vs 2 (7%) in the control group.

CONCLUSIONS AND RELEVANCE

In this single-center preliminary study of patients undergoing elective noncardiac surgery, the use of a machine learning-derived early warning system compared with standard care resulted in less intraoperative hypotension. Further research with larger study populations in diverse settings is needed to understand the effect on additional patient outcomes and to fully assess safety and generalizability.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT03376347.

Performance of the Hypotension Prediction Index with non-invasive arterial pressure waveforms in non-cardiac surgical patients

An algorithm derived from machine learning uses the arterial waveform to predict intraoperative hypotension some minutes before episodes, possibly giving clinician’s time to intervene and prevent hypotension. Whether the Hypotension Prediction Index works well with noninvasive arterial pressure waveforms remains unknown. We therefore evaluated sensitivity, specificity, and positive predictive value of the Index based on non-invasive arterial waveform estimates. We used continuous hemodynamic data measured from ClearSight (formerly Nexfin) noninvasive finger blood pressure monitors in surgical patients. We re-evaluated data from a trial that included 320 adults ≥ 45 years old designated ASA physical status 3 or 4 who had moderate-to-high-risk non-cardiac surgery with general anesthesia. We calculated sensitivity and specificity for predicting hypotension, defined as mean arterial pressure ≤ 65 mmHg for at least 1 min, and characterized the relationship with receiver operating characteristics curves. We also evaluated the number of hypotensive events at various ranges of the Hypotension Prediction Index. And finally, we calculated the positive predictive value for hypotension episodes when the Prediction Index threshold was 85. The algorithm predicted hypotension 5 min in advance, with a sensitivity of 0.86 [95% confidence interval 0.82, 0.89] and specificity 0.86 [0.82, 0.89]. At 10 min, the sensitivity was 0.83 [0.79, 0.86] and the specificity was 0.83 [0.79, 0.86]. And at 15 min, the sensitivity was 0.75 [0.71, 0.80] and the specificity was 0.75 [0.71, 0.80]. The positive predictive value of the algorithm prediction at an Index threshold of 85 was 0.83 [0.79, 0.87]. A Hypotension Prediction Index of 80–89 provided a median of 6.0 [95% confidence interval 5.3, 6.7] minutes warning before mean arterial pressure decreased to < 65 mmHg. The Hypotension Prediction Index, which was developed and validated with invasive arterial waveforms, predicts intraoperative hypotension reasonably well from non-invasive estimates of the arterial waveform. Hypotension prediction, along with appropriate management, can potentially reduce intraoperative hypotension. Being able to use the non-invasive pressure waveform will widen the range of patients who might benefit.

Clinical Trial Number: ClinicalTrials.gov NCT02872896.

Intraoperative hypotension and its prediction

Intraoperative hypotension (IOH) very commonly accompanies general anaesthesia in patients undergoing major surgical procedures. The development of IOH is unwanted, since it is associated with adverse outcomes such as acute kidney injury and myocardial injury, stroke and mortality. Although the definition of IOH is variable, harm starts to occur below a mean arterial pressure (MAP) threshold of 65 mmHg. The odds of adverse outcome increase for increasing duration and/or magnitude of IOH below this threshold, and even short periods of IOH seem to be associated with adverse outcomes. Therefore, reducing the hypotensive burden by predicting and preventing IOH through proactive appropriate treatment may potentially improve patient outcome. In this review article, we summarise the current state of the prediction of IOH by the use of so-called machine-learning algorithms. Machine-learning algorithms that use high-fidelity data from the arterial pressure waveform, may be used to reveal 'traits' that are unseen by the human eye and are associated with the later development of IOH. These algorithms can use large datasets for 'training', and can subsequently be used by clinicians for haemodynamic monitoring and guiding therapy. A first clinically available application, the hypotension prediction index (HPI), is aimed to predict an impending hypotensive event, and additionally, to guide appropriate treatment by calculated secondary variables to asses preload (dynamic preload variables), contractility (dP/dtmax), and afterload (dynamic arterial elastance, Eadyn). In this narrative review, we summarise the current state of the prediction of hypotension using such novel, automated algorithms and we will highlight HPI and the secondary variables provided to identify the probable origin of the (impending) hypotensive event.

A Narrative Review of Analytics in Pediatric Cardiac Anesthesia and Critical Care Medicine

Congenital heart disease (CHD) is one of the most common birth anomalies, and the care of children with CHD has improved over the past 4 decades. However, children with CHD who undergo general anesthesia remain at increased risk for morbidity and mortality. The proliferation of electronic health record systems and sophisticated patient monitors affords the opportunity to capture and analyze large amounts of CHD patient data, and the application of novel, effective analytics methods to these data can enable clinicians to enhance their care of pediatric CHD patients. This narrative review covers recent efforts to leverage analytics in pediatric cardiac anesthesia and critical care to improve the care of children with CHD.