Perioperative hypotension 2021: a contrarian view

Intraoperative Hypotension Prediction: Current Methods, Controversies, and Research Outlook

Intraoperative hypotension prediction has been increasingly emphasized due to its potential clinical value in reducing organ injury and the broad availability of large-scale patient datasets and powerful machine learning tools. Hypotension prediction methods can mitigate low blood pressure exposure time. However, they have yet to be convincingly demonstrated to improve objective outcomes; furthermore, they have recently become controversial. This review presents the current state of intraoperative hypotension prediction and makes recommendations on future research. We begin by overviewing the current hypotension prediction methods, which generally rely on the prevailing mean arterial pressure as one of the important input variables and typically show good sensitivity and specificity but low positive predictive value in forecasting near-term acute hypotensive events. We make specific suggestions on improving the definition of acute hypotensive events and evaluating hypotension prediction methods, along with general proposals on extending the methods to predict reduced blood flow and treatment effects. We present a start of a risk-benefit analysis of hypotension prediction methods in clinical practice. We conclude by coalescing this analysis with the current evidence to offer an outlook on prediction methods for intraoperative hypotension. A shift in research toward tailoring hypotension prediction methods to individual patients and pursuing methods to predict appropriate treatment in response to hypotension appear most promising to improve outcomes.

American society of anesthesiologists physical status classification significantly affects the performances of machine learning models in intraoperative hypotension inference

STUDY OBJECTIVE

To explore how American Society of Anesthesiologists (ASA) physical status classification affects different machine learning models in hypotension prediction and whether the prediction uncertainty could be quantified.

DESIGN

Observational Studies SETTING: UofL health hospital PATIENTS: This study involved 562 hysterectomy surgeries performed on patients (≥ 18 years) between June 2020 and July 2021.

INTERVENTIONS

None MEASUREMENTS: Preoperative and intraoperative data is collected. Three parametric machine learning models, including Bayesian generalized linear model (BGLM), Bayesian neural network (BNN), a newly proposed BNN with multivariate mixed responses (BNNMR), and one nonparametric model, Gaussian Process (GP), were explored to predict patients' diastolic and systolic blood pressures (continuous responses) and patients' hypotensive event (binary response) for the next five minutes. Data was separated into American Society of Anesthesiologists (ASA) physical status class 1- 4 before being read in by four machine learning models. Statistical analysis and models' constructions are performed in Python. Sensitivity, specificity, and the confidence/credible intervals were used to evaluate the prediction performance of each model for each ASA physical status class.

MAIN RESULTS

ASA physical status classes require distinct models to accurately predict intraoperative blood pressures and hypotensive events. Overall, high sensitivity (above 0.85) and low uncertainty can be achieved by all models for ASA class 4 patients. In contrast, models trained without controlling ASA classes yielded lower sensitivity (below 0.5) and larger uncertainty. Particularly, in terms of predicting binary hypotensive event, for ASA physical status class 1, BNNMR yields the highest sensitivity of 1. For classes 2 and 3, BNN has the highest sensitivity of 0.429 and 0.415, respectively. For class 4, BNNMR and GP are tied with the highest sensitivity of 0.857. On the other hand, the sensitivity is just 0.031, 0.429, 0.165 and 0.305 for BNNMR, BNN, GBLM and GP models respectively, when training data is not divided by ASA physical status classes. In terms of predicting systolic blood pressure, the GP regression yields the lowest root mean squared errors (RMSE) of 2.072, 7.539, 9.214 and 0.295 for ASA physical status classes 1, 2, 3 and 4, respectively, but a RMSE of 126.894 if model is trained without controlling the ASA physical status class. The RMSEs for other models are far higher. RMSEs are 2.175, 13.861, 17.560 and 22.426 for classes 1, 2, 3 and 4 respectively for the BGLM. In terms of predicting diastolic blood pressure, the GP regression yields the lowest RMSEs of 2.152, 6.573, 5.371 and 0.831 for ASA physical status classes 1, 2, 3 and 4, respectively; RMSE of 8.084 if model is trained without controlling the ASA physical status class. The RMSEs for other models are far higher. Finally, in terms of the width of the 95% confidence interval of the mean prediction for systolic and diastolic blood pressures, GP regression gives narrower confidence interval with much smaller margin of error across all four ASA physical status classes.

CONCLUSIONS

Different ASA physical status classes present different data distributions, and thus calls for distinct machine learning models to improve prediction accuracy and reduce predictive uncertainty. Uncertainty quantification enabled by Bayesian inference provides valuable information for clinicians as an additional metric to evaluate performance of machine learning models for medical decision making.

The safety and efficacy of remimazolam tosylate combined with propofol in upper gastrointestinal endoscopy: A multicenter, randomized clinical trial

INTRODUCTION

Hypotension is the most common adverse event under propofol-mediated sedation and is possible to cause varying degrees of damage to patients. Whereas remimazolam has a poorer sedative effect than propofol.

AIM

The aim of this study was to explore the advantages of the combination of remimazolam tosylate and propofol.

METHODS

304 patients were divided into the remimazolam tosylate group (RT group), the propofol group (P group), and the remimazolam tosylate plus propofol group(R+T group). The primary outcome was the incidence of hypotension. Secondary outcomes included the results of sedation and recovery. The safety results mainly include the incidence of Hypotension, adverse respiratory events, postoperative nausea and vomiting, hiccup, cough, body movement and bradycardia.

RESULTS

The incidence of hypotension was 56.7% in the P group, 12.6% in the RT group, and 31.3% in the R+P group, three groups of pairwise comparisons showed statistical differences, with P< 0.001. The incidence of body movement was significantly higher in the RT group (26.1%) than in the P group (10.3%) and the R+P group (12.5%), P = 0.004. The endoscopist satisfaction was higher in the P (3.87±0.44) and R+P (3.95±0.22)groups than in the RT(3.53±0.84) group. The incidence of adverse events, in descending order, was P group, RT group, and R+P group (93.8%vs.61.3%vs.42.7%).

CONCLUSION

Co-administration had fewer adverse events than propofol monotherapy, also had a better sedative effect and higher endoscopist satisfaction than remimazolam monotherapy.

TRIAL REGISTRATION

Clinical trial registration number: NCT05429086.

Management of Hemorrhagic Shock: Physiology Approach, Timing and Strategies

Hemorrhagic shock (HS) management is based on a timely, rapid, definitive source control of bleeding/s and on blood loss replacement. Stopping the hemorrhage from progressing from any named and visible vessel is the main stem fundamental praxis of efficacy and effectiveness and an essential, obligatory, life-saving step. Blood loss replacement serves the purpose of preventing ischemia/reperfusion toxemia and optimizing tissue oxygenation and microcirculation dynamics. The "physiological classification of HS" dictates the timely management and suits the 'titrated hypotensive resuscitation' tactics and the 'damage control surgery' strategy. In any hypotensive but not yet critical shock, the body's response to a fluid load test determines the cut-off point between compensation and progression between the time for adopting conservative treatment and preparing for surgery or rushing to the theater for rapid bleeding source control. Up to 20% of the total blood volume is given to refill the unstressed venous return volume. In any critical level of shock where, ab initio, the patient manifests signs indicating critical physiology and impending cardiac arrest or cardiovascular accident, the balance between the life-saving reflexes stretched to the maximum and the insufficient distal perfusion (blood, oxygen, and substrates) remains in a liable and delicate equilibrium, susceptible to any minimal change or interfering variable. In a cardiac arrest by exsanguination, the core of the physiological issue remains the rapid restoration of a sufficient venous return, allowing the heart to pump it back into systemic circulation either by open massage via sternotomy or anterolateral thoracotomy or spontaneously after aorta clamping in the chest or in the abdomen at the epigastrium under extracorporeal resuscitation and induced hypothermia. This is the only way to prevent ischemic damage to the brain and the heart. This is accomplishable rapidly and efficiently only by a direct approach, which is a crush laparotomy if the bleeding is coming from an abdominal +/- lower limb site or rapid sternotomy/anterolateral thoracotomy if the bleeding is coming from a chest +/- upper limbs site. Without first stopping the bleeding and refilling the heart, any further exercise is doomed to failure. Direct source control via laparotomy/thoracotomy, with the concomitant or soon following venous refilling, are the two essential, initial life-saving steps.

Hypotension as a marker or mediator of perioperative organ injury: a narrative review

Perioperative hypotension has been repeatedly associated with organ injury and worse outcome, yet many interventions to reduce morbidity by attempting to avoid or reverse hypotension have floundered. In part, this reflects uncertainty as to what threshold of hypotension is relevant in the perioperative setting. Shifting population-based definitions for hypertension, plus uncertainty regarding individualised norms before surgery, both present major challenges in constructing useful clinical guidelines that may help improve clinical outcomes. Aside from these major pragmatic challenges, a wealth of biological mechanisms that underpin the development of higher blood pressure, particularly with increasing age, suggest that hypotension (however defined) or lower blood pressure per se does not account solely for developing organ injury after major surgery. The mosaic theory of hypertension, first proposed more than 60 yr ago, incorporates multiple, complementary mechanistic pathways through which clinical (macrovascular) attempts to minimise perioperative organ injury may unintentionally subvert protective or adaptive pathways that are fundamental in shaping the integrative host response to injury and inflammation. Consideration of the mosaic framework is critical for a more complete understanding of the perioperative response to acute sterile and infectious inflammation. The largely arbitrary treatment of perioperative blood pressure remains rudimentary in the context of multiple complex adaptive hypertensive endotypes, defined by distinct functional or pathobiological mechanisms, including the regulation of reactive oxygen species, autonomic dysfunction, and inflammation. Developing coherent strategies for the management of perioperative hypotension requires smarter, mechanistically solid interventions delivered by RCTs where observer bias is minimised.

Intraoperative prediction of postanaesthesia care unit hypotension

BACKGROUND

Postoperative hypotension is associated with adverse outcomes, but intraoperative prediction of postanaesthesia care unit (PACU) hypotension is not routine in anaesthesiology workflow. Although machine learning models may support clinician prediction of PACU hypotension, clinician acceptance of prediction models is poorly understood.

METHODS

We developed a clinically informed gradient boosting machine learning model using preoperative and intraoperative data from 88 446 surgical patients from 2015 to 2019. Nine anaesthesiologists each made 192 predictions of PACU hypotension using a web-based visualisation tool with and without input from the machine learning model. Questionnaires and interviews were analysed using thematic content analysis for model acceptance by anaesthesiologists.

RESULTS

The model predicted PACU hypotension in 17 029 patients (area under the receiver operating characteristic [AUROC] 0.82 [95% confidence interval {CI}: 0.81-0.83] and average precision 0.40 [95% CI: 0.38-0.42]). On a random representative subset of 192 cases, anaesthesiologist performance improved from AUROC 0.67 (95% CI: 0.60-0.73) to AUROC 0.74 (95% CI: 0.68-0.79) with model predictions and information on risk factors. Anaesthesiologists perceived more value and expressed trust in the prediction model for prospective planning, informing PACU handoffs, and drawing attention to unexpected cases of PACU hypotension, but they doubted the model when predictions and associated features were not aligned with clinical judgement. Anaesthesiologists expressed interest in patient-specific thresholds for defining and treating postoperative hypotension.

CONCLUSIONS

The ability of anaesthesiologists to predict PACU hypotension was improved by exposure to machine learning model predictions. Clinicians acknowledged value and trust in machine learning technology. Increasing familiarity with clinical use of model predictions is needed for effective integration into perioperative workflows.