Non-invasive Measurement of the Hypotension Prediction Index for the Reduction of Intraoperative Hypotension

Overview

In order to reduce the incidence of IOH, various goal-directed therapy (GDT) protocols have already been introduced with success regarding the reduction of postoperative AKI and MINS. However, these studies used an invasive hemodynamic monitoring which offered a continuous surveillance of the blood pressure. In contrast, standard non-invasive blood pressure monitoring results in a blind gap between two measurements (mostly three or five minutes). In order to address this limitation, different continuous non-invasive blood pressure monitoring devices have been introduced. The next evolutional step of non-invasive cardiac output monitoring was to prevent IOH before their onset by using the Hypotension Prediction Index (HPI). Based on the Edward ́s monitoring platform, HPI is a monitoring tool which aims to predict IOH (defined as MAP\<65 mmHg for at least one minute) up to 15 min before its onset. The underlying machine learning based algorithm uses analyses features from the pressure waveform and was first calculated from a large retrospective data set of surgical patients and subsequently validated in a prospective cohort. In this study HPI showed a sensitivity of 88% and specificity of 87% for predicting IOH 15 min before its onset. Since then, own and studies of other working groups confirmed the effective prevention of IOH by the use of HPI-based GDT. Until today the arterial waveform analysis was dependent on invasive arterial measurement but since Edwards Lifesciences already promoted the start of the HPI on the ClearSight platform a non-invasive measurement will soon be possible. Further, until now it has not yet been proven that the perioperative use of a continuous non-invasive blood pressure monitoring has a beneficial effect on the patient´s outcome. Study objectives The aim of the study is to investigate whether a hemodynamic protocol based on continuous non-invasive cardiac output monitoring (ClearSight system) compared to standard care can reduce the incidence of IOH, postoperative AKI, and MINS in patients undergoing major trauma and orthopedic surgery.

Complications related to intraoperative hypotension (IOH) can be detected in most organ systems of which renal failure poses a relevant complication in the perioperative phase. Acute kidney injury (AKI) affects up to 25 % of patients attending the intensive care unit (ICU). Currently, serum creatinine and urea display the most common parameters used to detect AKI, but it may take a day or more for serum creatinine levels to accumulate in the blood of a patient with an AKI. For this reason, it may not reflect real time kidney damage or loss of function. To overcome this limitation, the cell cycle arrest biomarkers TIMP-2- and IGFBP7-quantification (Nephrocheck) has been successfully evaluated for the detection of AKI. The main advantage of both parameters is the opportunity of early detection of AKI and its point-of-care design, which makes them especially for the use on the ICU valuable. Myocardial injury after non-cardiac surgery (MINS) displays another adverse outcome associated to IOH and endangers particularly patients with an age of 45 years and older and approximately 1% of all patients suffering of MINS die within 30 days after non-cardiac surgery. In order to reduce the incidence of IOH, various goal-directed therapy (GDT) protocols have already been introduced with success regarding the reduction of postoperative AKI and MINS. However, these studies used an invasive hemodynamic monitoring which offered a continuous surveillance of the blood pressure. In contrast, standard non-invasive blood pressure monitoring results in a blind gap between two measurements (mostly three or five minutes). In order to address this limitation, different continuous non-invasive blood pressure monitoring devices have been introduced. The next evolutional step of non-invasive cardiac output monitoring was to prevent IOH before their onset by using the Hypotension Prediction Index (HPI). Based on the Edward ́s monitoring platform, HPI is a monitoring tool which aims to predict IOH (defined as MAP\<65 mmHg for at least one minute) up to 15 min before its onset. Until today the arterial waveform analysis was dependent on invasive arterial measurement but since Edwards Lifesciences already promoted the start of the HPI on the ClearSight platform a non-invasive measurement will soon be possible. Further, until now it has not yet been proven that the perioperative use of a continuous non-invasive blood pressure monitoring has a beneficial effect on the patient´s outcome. Especially, a GDT based on non-invasive blood pressure monitoring might not only be able to reduce the incidence of IOH but also of the occurrence of postoperative renal failure. 2.1 Study hypothesis 2.1.1 Primary study hypothesis The perioperative use of non-invasive HPI-guided GDT reduces the incidence of IOH in patients undergoing major trauma and orthopedic surgery. 2.1.2 Secondary study hypothesis * The perioperative use of non-invasive HPI-guided GDT reduces the occurrence of postoperative renal failure in patients undergoing major trauma and orthopedic surgery. * The perioperative use of non-invasive HPI-guided GDT reduces the occurrence of postoperative MINS in patients undergoing major trauma and orthopedic surgery. 2.2 Study objectives The aim of the study is to investigate whether a hemodynamic protocol based on continuous non-invasive cardiac output monitoring (ClearSight system) compared to standard care can reduce the incidence of IOH, postoperative AKI, and MINS in patients undergoing major trauma and orthopedic surgery. 3 Methodology 3.1 Study design The study is designed as a monocentric randomized prospective interventional trial comparing goal directed hemodynamic management using continuous non-invasive cardiac output monitoring (ClearSight system) to standard care. 3.2 Study centers University Hospital Giessen, Department of Anesthesiology and Intensive Care Medicine 3.3 Study Population 3.3.1 Study groups Major Trauma and Orthopedic Surgery 3.4 Working plan 3.4.1 Preoperative Assessment Patients are recruited before surgery after checking inclusion and exclusion criteria. Informed consent is obtained at this time. Patients will be randomized 1:1 to the two groups after achieving the patient´s informed consent. Further, the following basic characteristics are obtained: * Age, sex, height, weight, ASA score * Pre-existing conditions (hypertension, coronary heart disease with and without history of myocardial infarction, peripheral arterial disease, renal failure, chronic obstructive pulmonary disease, diabetes) * Previous major surgeries * Current prescription of medication Furthermore, the following laboratory results will be gained: * Blood cell count * Global coagulatory function (Internationalized Ratio, thromboplastine time, fibrinogen levels) * Parameters of renal function (Creatinine, urea, Nephrocheck, blood and urinary mitochondrial DNA) * Parameters of cardiac function (Troponin I, Creatinine kinase, Myoglobin, Brain Natriuretic Peptide) * Inflammatory Parameters (C-Reactive Protein, Procalcitonin) * Parameters of endothelial function (Angiopoietins 1 and 2, Syndecan-1 and intercellular adhesion molecule-1 (ICAM-1)), Bio-ADM 3.4.2 Time Points The study time points are defined as followed: prior to surgery as well as immediately, 24, 72, and 168 hours after surgery (depending on the duration of hospital stay). At any time point clinical data, blood and urine will be collected (depending on the duration of hospital stay). 3.4.3 Perioperative Management 3.4.3.1 Induction and Maintenance of anesthesia All patients receive the standard hemodynamic monitoring (electrocardiogram, non-invasive blood pressure, and plethysmography). Non-invasive blood pressure will be measured every three minutes. Independently of the randomized study group, induction of anesthesia will be performed with fentanyl, propofol, and cis-atracurium. Dosages will be chosen according to the patient´s age and body weight as well as pre-existing diseases according to the assessment of the attending physician. After intubation, all patients are ventilated with a tidal volume of 8 ml/kg ideal bodyweight and with regard to the capnography (target end-tidal CO2 of 35-40 mmHg). The control group will be managed according to the investigators´ SOP with the aim of an MAD \&amp;gt; 65mmHg. 3.4.3.2 Management of interventional group patients Prior to the surgery the rest cardiac index and contractility (dp/dt) must be quantified. For this purpose, the cardiac index will be measured in the preoperative night by applicating the HPI ClearSight system through a study team member. A nighttime cardiac index is accepted when more than three reliable measurements were recorded in rest over a time period of 60 minutes. If the rest cardiac index is not available throughout the night because the patient´s sleep is altered by the measurements, the awake cardiac index will be quantified until the monitoring is stopped for the night sleep of the patient. This mean baseline measurements (CI and dp/dt) will then be the target cardiac index throughout the study algorithm (figure 1). In case no sleep measurement was achievable, the awake measurement will be accounted as baseline value. The perioperative study intervention period starts with the beginning of anesthesia and ends at the end of surgery. Intraoperative mean arterial pressure will be maintained at least at 65 mmHg and cardiac index and dp/dt will be individually optimized according to the GDT algorithm. 3.5 Data Processing Data collection is carried out consistently on pre-defined time-points in the investigators´ electronic patient data management system into a separate study database (Microsoft Excel). The collected data is pseudonymised in the database based on a random key method. The chart with the patient data and decrypting keys is kept in the study center for at least 15 years after the end of the study (publication). Data anonymization is intentionally not performed to give patients the option for data insight or deletion of their data in the future. Data management and evaluation is performed by the study team. 3.6 Patient number and Biometrics The aim of the study is to show the impact of non-invasive cardiac output monitoring on the incidence of IOH in a cohort of trauma and orthopedic surgery. Sample size calculation was performed with regard to a recent study by Maheshwari et al. who investigated the effect of HPI on the prevention of hypotension. This study was chosen because the primary endpoint, respectively the definition of hypotension (MAP\<65 mmHg), was identical to the investigators´ study and they investigated also non-cardiac surgical patients. In this study, the mean number of hypotensive periods (given as an area under the curve of MAP ≤65 mmHg) of patients without hemodynamic management accounted to 34.2 \[8,5-112.7\] compared to 32.7 \[6.3-102\] in patients with hemodynamic measurement. Aiming for an alpha of 0.05 and power of 0.95, the sample size calculation resulted in 66 patients per study group (total 132 patients, based on the use of the Wilcoxon test). In order to address potential dropouts (estimated drop-out rate 10-20%) the investigators chose to increase the patients numbers to 75 patients in each study group. Next to the target parameters, data of the hemodynamic and respiratory function will be achieved as well as of the anesthetic and hemodynamic management (please see CRF). Furthermore, general characteristics such as age, gender, body mass index, as well as pre-existing conditions and prescriptions will be assessed. For the target variables, the results will be investigated and analyzed descriptively (e.g., checked for distribution). Metric characteristics (mean and standard deviation) as well as median and interquartile difference and achieved frequencies (with a percentage specification) will be determined. As part of the exploratory analysis, the structural equilibrium (homogeneity) of the treatment groups will also be checked. Depending on the distribution of the observation values, appropriate test methods are used. The outcome of the statistical testing will be controlled for influence of secondary parameters, as there are suspected reasons for hypotension, type and dosage of vasopressors used during the procedure, as well as type and dosage of inotropic medication. All analysis will be done using R-Plus scripting (R Core Team (2016). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/). 3.7 Benefit - Risk assessment 3.7.1 Potential benefit Based on a continuous monitoring by additional monitoring system an early detection of potential life-threatening events and acute kidney injury is possible. This can result in an optimization of the patients' therapy and a better outcome. 3.7.2 Potential Risks The presented study is an interventional study. The potential risks are marginal. The usage of an additional non-invasive cardiac output monitoring is minimal. The time points of blood samples for the study are in line with routine sampling. Based on this, there is no additional risk for the patient. 3.7.3 Benefit/ Risk analysis The benefit for the patients is additional monitoring, based on an additional monitoring device and the supervising study doctor, who can support the treating anesthesiologist with information in potentially critical situations. Thereby, it is possible to treat early goal-directed and possibly improve the patient´s outcome. Considering the potential benefits of the generated information for the patient in comparison to the expected risks, the beneficial effect is overbalanced. The expected gain in knowledge from this study could be used for optimizing perioperative care.