Prediction of lactate concentrations after cardiac surgery using machine learning and deep learning approaches.

Background: Although conventional prediction models for surgical patients often ignore intraoperative time-series data, deep learning approaches are well-suited to incorporate time-varying and non-linear data with complex interactions. Blood lactate concentration is one important clinical marker that can reflect the adequacy of systemic perfusion during cardiac surgery. During cardiac surgery and cardiopulmonary bypass, minute-level data is available on key parameters that affect perfusion. The goal of this study was to use machine learning and deep learning approaches to predict maximum blood lactate concentrations after cardiac surgery. We hypothesized that models using minute-level intraoperative data as inputs would have the best predictive performance. Methods: Adults who underwent cardiac surgery with cardiopulmonary bypass were eligible. The primary outcome was maximum lactate concentration within 24 h postoperatively. We considered three classes of predictive models, using the performance metric of mean absolute error across testing folds: (1) static models using baseline preoperative variables, (2) augmentation of the static models with intraoperative statistics, and (3) a dynamic approach that integrates preoperative variables with intraoperative time series data. Results: 2,187 patients were included. For three models that only used baseline characteristics (linear regression, random forest, artificial neural network) to predict maximum postoperative lactate concentration, the prediction error ranged from a median of 2.52 mmol/L (IQR 2.46, 2.56) to 2.58 mmol/L (IQR 2.54, 2.60). The inclusion of intraoperative summary statistics (including intraoperative lactate concentration) improved model performance, with the prediction error ranging from a median of 2.09 mmol/L (IQR 2.04, 2.14) to 2.12 mmol/L (IQR 2.06, 2.16). For two modelling approaches (recurrent neural network, transformer) that can utilize intraoperative time-series data, the lowest prediction error was obtained with a range of median 1.96 mmol/L (IQR 1.87, 2.05) to 1.97 mmol/L (IQR 1.92, 2.05). Intraoperative lactate concentration was the most important predictive feature based on Shapley additive values. Anemia and weight were also important predictors, but there was heterogeneity in the importance of other features. Conclusion: Postoperative lactate concentrations can be predicted using baseline and intraoperative data with moderate accuracy. These results reflect the value of intraoperative data in the prediction of clinically relevant outcomes to guide perioperative management.

Feasibility study of intraoperative pericardial fluid biomarkers and length of stay after cardiac surgery.

Objective: Pericardial fluid biomarkers reflect the physiologic state of the myocardium. Previously, we showed a sustained increase in pericardial fluid biomarkers compared with blood in the 48 hours after cardiac surgery. We assess the feasibility of analyzing 9 common cardiac biomarkers from pericardial fluid collected during cardiac surgery and test a preliminary hypothesis of association between the most common biomarkers, troponin and brain natriuretic peptide, and length of stay after surgery. Methods: We prospectively enrolled 30 patients aged 18 years or more undergoing coronary artery or valvular surgery. Patients with ventricular assist devices, atrial fibrillation surgery, thoracic aorta surgery, redo surgery, concomitant noncardiac surgery, and preoperative inotropic support were excluded. Before pericardial excision during surgery, a 1-cm pericardial incision was made to insert an 18-gauge catheter and collect 10 mL of pericardial fluid. Concentrations of 9 established biomarkers of cardiac injury or inflammation including brain natriuretic peptide and troponin were measured. Zero truncated Poisson regression adjusted for Society of Thoracic Surgery Preoperative Risk of Mortality tested for a preliminary association between pericardial fluid biomarkers and length of stay. Results: Pericardial fluid was collected and pericardial fluid biomarkers resulted for all patients. Adjusted for Society of Thoracic Surgery risk, brain natriuretic peptide, and troponin were associated with increased intensive care unit and overall hospital length of stay. Conclusions: In 30 patients, pericardial fluid was obtained and analyzed for cardiac biomarkers. Adjusting for Society of Thoracic Surgery risk, pericardial fluid troponin and brain natriuretic peptide were preliminarily associated with increased length of stay. Further investigation is needed to validate this finding and to investigate the potential clinical utility of pericardial fluid biomarkers.

Impact of early pericardial fluid chymase activation after cardiac surgery.

Introduction: Chymase is a highly destructive serine protease rapidly neutralized in the circulation by protease inhibitors. Here we test whether pericardial fluid (PCF) chymase activation and other inflammatory biomarkers determine intensive care unit length of stay, and explore mechanisms of chymase delivery by extracellular vesicles to the heart. Methods: PCF was collected from adult patients (17 on-pump; 13 off-pump) 4 h after cardiac surgery. Extracellular vesicles (EVs) containing chymase were injected into Sprague-Dawley rats to test for their ability to deliver chymase to the heart. Results: The mean intensive care unit (ICU) stay and mean total length of stay was 2.17 ± 3.8 days and 6.41 ± 1.3 days respectively. Chymase activity and 32 inflammatory markers did not differ in on-pump vs. off-pump cardiac surgery. Society of Thoracic Surgeons Predicted Risk of Morbidity and Mortality Score (STS-PROM), 4-hour post-surgery PCF chymase activity and C-X-C motif chemokine ligand 6 (CXCL6) were all independent predictors of ICU and total hospital length of stay by univariate analysis. Mass spectrometry of baseline PCF shows the presence of serine protease inhibitors that neutralize chymase activity. The compartmentalization of chymase within and on the surface of PCF EVs was visualized by immunogold labeling and transmission electron microscopy. A chymase inhibitor prevented EV chymase activity (0.28 fmol/mg/min vs. 14.14 fmol/mg/min). Intravenous injection of PCF EVs obtained 24 h after surgery into Sprague Dawley rats shows diffuse human chymase uptake in the heart with extensive cardiomyocyte damage 4 h after injection. Discussion: Early postoperative PCF chymase activation underscores its potential role in cardiac damage soon after on- or off-pump cardiac surgery. In addition, chymase in extracellular vesicles provides a protected delivery mechanism from neutralization by circulating serine protease inhibitors.

Point-of-Care Ultrasound: A Moving Picture Is Worth a Thousand Tests.

The effective utilization of point-of-care ultrasound may decrease the utilization of conventional diagnostic modalities. This review describes the various pathologies that can be effectively and rapidly identified with point-of-care cardiac, lung, abdominal, vascular airway, and ocular ultrasonography.

Domains of Frailty Predict Loss of Independence in Older Adults After Noncardiac Surgery.

IMPORTANCE: Preoperative frailty has been consistently associated with death, severe complications, and loss of independence (LOI) after surgery. LOI is an important patient-centered outcome, but it is unclear which domains of frailty are most strongly associated with LOI. Such information would be important to target individual geriatric domains for optimization. OBJECTIVE: To assess whether impairment in individual domains of the Edmonton Frail Scale (EFS) can predict LOI in older adults after noncardiac surgery. DESIGN: Retrospective Cohort Study. SETTING: One Academic Hospital. PARTICIPANTS: Patients aged 65 or older who were living independently and evaluated with the EFS during a preoperative visit to the Center for Preoperative Optimization at the Johns Hopkins Hospital between June 2018 and January 2020. MAIN OUTCOME: LOI defined as discharge to increased level of care outside of the home with new mobility deficit or functional dependence. New mobility deficit and functional dependence were extracted from chart review of the standardized occupational therapy and physical therapy assessment performed before discharge. RESULTS: A total of 3497 patients were analyzed. Age (mean±SD) was 73.4±6.2 years, and 1579 (45.2%) were female. The median total EFS score was 3 (range 0-16), and 725/3497 (27%) were considered frail (EFS≥6). The frequencies of impairment in each EFS domain were functional performance (33.5% moderately impaired, 11% severely impaired), history of hospital readmission (42%), poor self-described health status (37%), and abnormal cognition (17.1% moderately impaired, 13.8% severely impaired). Overall, 235/3497 (6.7%) patients experienced LOI. Total EFS score was associated with LOI (odds ratio: 1.37, 95% CI, 1.30-1.45, P <0.001) in a model adjusted for age, sex, body mass index, American Society of Anesthesiologists rating, congestive heart failure, valvular heart disease, hypertension diagnosis, chronic lung disease, diabetes, renal failure, liver disease, weight loss, anemia, and depression. Using a nested log likelihood approach, the domains of functional performance, functional dependence, social support, health status, and urinary incontinence improved the base multivariable model. In cross-validation, total EFS improved the prediction of LOI with the final model achieving an area under the curve of 0.840. Functional performance was the single domain that most improved outcome prediction, but together with functional dependence, social support, and urinary incontinence, the model resulted in an area under the curve of 0.838. CONCLUSION AND RELEVANCE: Among domains measured by the EFS before a wide range of noncardiac surgeries in older adults, functional performance, functional dependence, social support, and urinary incontinence were independently associated with and improved the prediction of LOI. Clinical initiatives to mitigate LOI may consider screening with the EFS and targeting abnormalities within these domains.

Preoperative Evaluation and Cardiac Risk Assessment in Vascular Surgery.

We summarize epidemiologic trends, outcomes, and preoperative guidelines for vascular surgery patients from 2010 to 2022. Vascular surgery continues to evolve in technology and engineering to treat a surgical population that suffers from a high prevalence of comorbidities. Preoperative optimization seeks to characterize the burden of disease and to achieve medical control in the timeline available before surgery. Risk assessment, evaluation, optimization, and prediction of major adverse cardiac events is an evolving science where the Vascular Surgery Quality Initiative has made an impact. Ongoing investigation may demonstrate value for preoperative echocardiography, functional capacity, frailty, and mobility assessments.

Comparing three wearable accelerometers to measure early activity after cardiac surgery.

Objective: Wearable activity monitors can provide detailed data on activity after cardiac surgery and discriminate a patient's risk for hospital-based outcomes. However, comparative data for different monitoring approaches, as well as predictive ability over clinical characteristics, are lacking. In addition, data on specific thresholds of activity are needed. The objective of this study was to compare 3 wearable activity monitors and 1 observational mobility scale in discriminating risk for 3 hospital-based outcomes, and to establish clinically relevant step thresholds. Methods: Cardiac surgery patients were enrolled between June 2016 and August 2017 in a cohort study. Postoperative activity was measured by 3 accelerometry monitors (StepWatch Ambulation Monitor, Fitbit Charge HR, and ActiGraph GT9X) and 1 nurse-based observation scale. Monitors represent a spectrum of characteristics, including wear location (ankle/wrist), output (activity counts/steps), consumer accessibility, and cost. Primary outcomes were duration of hospitalization >7 days, discharge to a nonhome location, and 30-day readmission. Results: Data were available from 193 patients (median age 67 years [interquartile range, 58-72]). All postoperative day 2 activity metrics (ie, from StepWatch, Fitbit, ActiGraph, and the observation scale) were independently associated with prolonged hospitalization and discharge to a nonhome location. Only steps as measured by StepWatch was independently associated with 30-day readmission. Overall, StepWatch provided the greatest discrimination (C-statistics 0.71-0.76 for all outcomes). Step thresholds between 250 and 500 steps/day identified between 74% and 96% of patients with any primary outcome. Conclusions: Data from wearable accelerometers provide additive value in early postoperative risk-stratification for hospital-based outcomes. These results both support and provide guidance for activity-monitoring programs after cardiac surgery.

Selected Transesophageal Echocardiographic Parameters of Left Ventricular Diastolic Function Predict Length of Stay Following Coronary Artery Bypass Graft-A Prospective Observational Study.

(1) Importance: Abnormal left ventricular (LV) diastolic function, with or without a diagnosis of heart failure, is a common finding that can be easily diagnosed by intra-operative transesophageal echocardiography (TEE). The association of diastolic function with duration of hospital stay after coronary artery bypass (CAB) is unknown. (2) Objective: To determine if selected TEE parameters of diastolic dysfunction are associated with length of hospital stay after coronary artery bypass surgery (CAB). (3) Design: Prospective observational study. (4) Setting: A single tertiary academic medical center. (5) Participants: Patients with normal systolic function undergoing isolated CAB from September 2017 through June 2018. (6) Exposures: LV function during diastole, as assessed by intra-operative TEE prior to coronary revascularization. (7) Main Outcomes and Measures: The primary outcome was duration of postoperative hospital stay. Secondary intermediate outcomes included common postoperative cardiac, respiratory, and renal complications. (8) Results: The study included 176 participants (mean age 65.2 ± 9.2 years, 73% male); 105 (60.2%) had LV diastolic dysfunction based on selected TEE parameters. Median time to hospital discharge was significantly longer for subjects with selected parameters of diastolic dysfunction (9.1/IQR 6.6−13.5 days) than those with normal LV diastolic function (6.5/IAR 5.3−9.7 days) (p < 0.001). The probability of hospital discharge was 34% lower (HR 0.66/95% CI 0.47−0.93) for subjects with diastolic dysfunction based on selected TEE parameters, independent of potential confounders, including a baseline diagnosis of heart failure. There was a dose−response relation between severity of diastolic dysfunction and probability of discharge. LV diastolic dysfunction based on those selected TEE parameters was also associated with postoperative cardio-respiratory complications; however, these complications did not fully account for the relation between LV diastolic dysfunction and prolonged length of hospital stay. (9) Conclusions and Relevance: In patients with normal systolic function undergoing CAB, diastolic dysfunction based on selected TEE parameters is associated with prolonged duration of postoperative hospital stay. This association cannot be explained by baseline comorbidities or common post-operative complications. The diagnosis of diastolic dysfunction can be made by TEE.

Feasibility Study of a Fully Synchronous Virtual Critical Care Elective Focused on Learner Engagement.

BACKGROUND: The COVID-19 pandemic disrupted clinical education for medical students. With the rise of variants, meaningful in-person clinical experiences remain threatened. This report describes the design, implementation, and evaluation of a fully synchronous virtual critical care elective for medical students focused on learner engagement. METHODS: The two-week elective was offered during June and July 2020 in the COVID-19 extracorporeal membrane oxygenation (ECMO) unit. Medical students remotely participated in multidisciplinary rounds with the attending physician connected from the bedside via a head-mounted camera providing the first-person video view. Other team members connected outside the negative pressure area. Learners electronically completed daily intensive care unit (ICU) goals sheet (GS) for each patient. The daily completion percentage of the GS assessed the learner engagement, and the learners evaluated the experience with a five-point Likert scale survey. RESULTS: Nine medical students participated in two separate cohorts. Cohort A had 53 patient encounters, and Cohort B had 45 patient encounters totaling 301.5 total hours of supervised virtual patient interaction. The mean completion percentage of the daily ICU GS for the combined cohorts was 77.8%, (with a standard deviation of 9.6%), with sustained or increased completion from start to finish for all learners. All medical students agreed that the daily ICU GS was helpful for following rounds, organizing patient assessments and plans, and participating in patient care. The majority (88.9%) agreed that the elective increased their comfort in caring for critically ill patients. CONCLUSIONS: During the COVID-19 pandemic, a fully synchronous virtual critical care elective successfully utilized the first-person view and daily ICU GS to promote and assess learner engagement.

Association Between Left Ventricular Relative Wall Thickness and Acute Kidney Injury After Noncardiac Surgery.

BACKGROUND: Acute kidney injury (AKI) after major noncardiac surgery is commonly attributed to cardiovascular dysfunction. Identifying novel associations between preoperative cardiovascular markers and kidney injury may guide risk stratification and perioperative intervention. Increased left ventricular relative wall thickness (RWT), routinely measured on echocardiography, is associated with myocardial dysfunction and long-term risk of heart failure in patients with preserved left ventricular ejection fraction (LVEF); however, its relationship to postoperative complications has not been studied. We evaluated the association between preoperative RWT and AKI in high-risk noncardiac surgical patients with preserved LVEF. METHODS: Patients ≥18 years of age having major noncardiac surgery (high-risk elective intra-abdominal or noncardiac intrathoracic surgery) between July 1, 2016, and June 30, 2018, who had transthoracic echocardiography in the previous 12 months were eligible. Patients with preoperative creatinine ≥2 mg/dL or reduced LVEF (<50%) were excluded. The association between RWT and AKI, defined as an increase in serum creatinine by 0.3 mg/dL from baseline within 48 hours or by 50% within 7 days after surgery, was assessed using multivariable logistic regression adjusted for preoperative covariates. An additional model adjusted for intraoperative covariates, which are strongly associated with AKI, especially hypotension. RWT was modeled continuously, associating the change in odds of AKI for each 0.1 increase in RWT. RESULTS: The study included 1041 patients (mean ± standard deviation [SD] age 62 ± 15 years; 59% female). A total of 145 subjects (13.9%) developed AKI within 7 days. For RWT quartiles 1 through 4, respectively, 20 of 262 (7.6%), 40 of 259 (15.4%), 39 of 263 (14.8%), and 46 of 257 (17.9%) developed AKI. Log-odds and proportion with AKI increased across the observed RWT values. After adjusting for confounders (demographics, American Society of Anesthesiologists [ASA] physical status, comorbidities, baseline creatinine, antihypertensive medications, and left ventricular mass index), each RWT increase of 0.1 was associated with an estimated 26% increased odds of developing AKI (odds ratio [OR]; 95% confidence interval [CI]) of 1.26 (1.09-1.46; P = .002). After adjusting for intraoperative covariates (length of surgery, presence of an arterial line, intraoperative hypotension, crystalloid administration, transfusion, and urine output), RWT remained independently associated with the odds of AKI (OR; 95% CI) of 1.28 (1.13-1.47; P = .001). Increased RWT was also independently associated with hospital length of stay and adjusted hazard ratio (HR [95% CI]) of 0.94 (0.89-0.99; P = .018). CONCLUSIONS: Left ventricular RWT is a novel cardiovascular factor associated with AKI within 7 days after high-risk noncardiac surgery among patients with preserved LVEF. Application of this commonly available measurement of risk stratification or perioperative intervention warrants further investigation.

Team Physiological Dynamics: A Critical Review.

OBJECTIVE: Review the use of physiological measurement in team settings and propose recommendations to improve the state of the science. BACKGROUND: New sensor and analytical capabilities enable exploration of relationships between team members' physiological dynamics. We conducted a review of physiological measures used in research on teams to understand (1) how these measures are theoretically and operationally related to team constructs and (2) what types of validity evidence exist for physiological measurement in team settings. METHOD: We identified 32 articles that investigated task-performing teams using physiological data. Articles were coded on several dimensions, including team characteristics. Study findings were categorized by relationships tested between team physiological dynamics (TPD) and team inputs, mediators/processes, outputs, or psychometric properties. RESULTS: TPD researchers overwhelmingly measure single physiological systems. Although there is research linking TPD to inputs and outputs, the research on processes is underdeveloped. CONCLUSION: We recommend several theoretical, methodological, and statistical themes to expand the growth of the TPD field. APPLICATION: Physiological measures, once established as reliable indicators of team functioning, might be used to diagnose suboptimal team states and cue interventions to ameliorate these states.

Mentoring Fellows in Adult Cardiothoracic Anesthesiology for Academic Practice in the Contemporary Era-Perspectives From Mentors Around the United States.

This special article presents perspectives on the mentoring of fellows for academic practice in adult cardiothoracic anesthesiology. A comprehensive mentoring model should address the areas of clinical care, educational expertise and exposure to scholarly activity. The additional value of educational exposure to patient safety, quality improvement and critical care medicine in this model is also explored.