Beyond Conventional Hemodynamic Monitoring-Monitoring to Improve Our Understanding of Disease Process and Interventions.

Monitoring the hemodynamic state of patients is a hallmark of any intensive care environment. However, no single monitoring strategy can provide all the necessary data to paint the entire picture of the state of a patient; each monitor has strengths and weaknesses, advantages, and limitations. We review the currently available hemodynamic monitors used in pediatric critical care units using a clinical scenario. This provides the reader with a construct to understand the progression from basic to more advanced monitoring modalities and how they serve to inform the practitioner at the bedside.

Making machine learning matter to clinicians: model actionability in medical decision-making.

Machine learning (ML) has the potential to transform patient care and outcomes. However, there are important differences between measuring the performance of ML models in silico and usefulness at the point of care. One lens to use to evaluate models during early development is actionability, which is currently undervalued. We propose a metric for actionability intended to be used before the evaluation of calibration and ultimately decision curve analysis and calculation of net benefit. Our metric should be viewed as part of an overarching effort to increase the number of pragmatic tools that identify a model's possible clinical impacts.

Analyzing Continuous Physiologic Data to Find Hemodynamic Signatures Associated With New Brain Injury After Congenital Heart Surgery.

Continuous data capture technology is becoming more common. Establishing analytic approaches for continuous data could aid in understanding the relationship between physiology and clinical outcomes. OBJECTIVES: Our objective was to design a retrospective analysis for continuous physiologic measurements and their relationship with new brain injury over time after cardiac surgery. DESIGN SETTING AND PARTICIPANTS: Retrospective cohort study in the Cardiac Critical Care Unit at the Hospital for Sick Children in patients after repair of transposition of the great arteries (TGA) or single ventricle (SV) lesions. MAIN OUTCOMES AND MEASURES: Continuously acquired physiologic measurements for up to 72 hours after cardiac surgery were analyzed for association with new brain injury by MRI. Distributions of heart rate (HR), systolic blood pressure (BP), and oxygen saturation (Spo2) for SV and TGA were analyzed graphically and with descriptive statistics over postoperative time for data-driven variable selection. Mixed-effects regression analyses characterized relationships between HR, BP, and Spo2 and new brain injury over time while accounting for variation between patients, measurement heterogeneity, and missingness. RESULTS: Seventy-seven patients (60 TGA; 17 SV) were included. New brain injury was seen in 26 (34%). In SV patients, with and without new brain injury, respectively, in the first 24 hours after cardiac surgery, the median (interquartile range) HR was 172.0 beats/min (bpm) (169.7-176.0 bpm) versus 159.6 bpm (145.0-167.0 bpm); systolic BP 74.8 (67.9-78.5 mm Hg) versus 68.9 mm Hg (61.6-70.9 mm Hg). Higher postoperative HR (parameter estimate, 19.4; 95% CI, 7.8-31; p = 0.003 and BP, 8.6; 1.3-15.8; p = 0.024) were associated with new brain injury in SV patients. The strength of this relationship decreased with time. CONCLUSIONS AND RELEVANCE: Retrospective analysis of continuous physiologic measurements can provide insight into changes in postoperative physiology over time and their relationship with new brain injury. This technique could be applied to assess relationships between physiologic data and many patient interventions or outcomes.

An integration engineering framework for machine learning in healthcare.

Background and Objectives: Machine Learning offers opportunities to improve patient outcomes, team performance, and reduce healthcare costs. Yet only a small fraction of all Machine Learning models for health care have been successfully integrated into the clinical space. There are no current guidelines for clinical model integration, leading to waste, unnecessary costs, patient harm, and decreases in efficiency when improperly implemented. Systems engineering is widely used in industry to achieve an integrated system of systems through an interprofessional collaborative approach to system design, development, and integration. We propose a framework based on systems engineering to guide the development and integration of Machine Learning models in healthcare. Methods: Applied systems engineering, software engineering and health care Machine Learning software development practices were reviewed and critically appraised to establish an understanding of limitations and challenges within these domains. Principles of systems engineering were used to develop solutions to address the identified problems. The framework was then harmonized with the Machine Learning software development process to create a systems engineering-based Machine Learning software development approach in the healthcare domain. Results: We present an integration framework for healthcare Artificial Intelligence that considers the entirety of this system of systems. Our proposed framework utilizes a combined software and integration engineering approach and consists of four phases: (1) Inception, (2) Preparation, (3) Development, and (4) Integration. During each phase, we present specific elements for consideration in each of the three domains of integration: The Human, The Technical System, and The Environment. There are also elements that are considered in the interactions between these domains. Conclusion: Clinical models are technical systems that need to be integrated into the existing system of systems in health care. A systems engineering approach to integration ensures appropriate elements are considered at each stage of model design to facilitate model integration. Our proposed framework is based on principles of systems engineering and can serve as a guide for model development, increasing the likelihood of successful Machine Learning translation and integration.

Care of the critically ill neonate with hypoxemic respiratory failure and acute pulmonary hypertension: framework for practice based on consensus opinion of neonatal hemodynamics working group.

Circulatory transition after birth presents a critical period whereby the pulmonary vascular bed and right ventricle must adapt to rapidly changing loading conditions. Failure of postnatal transition may present as hypoxemic respiratory failure, with disordered pulmonary and systemic blood flow. In this review, we present the biological and clinical contributors to pathophysiology and present a management framework.

Hemodynamic Patterns Before Inhospital Cardiac Arrest in Critically Ill Children: An Exploratory Study.

To characterize prearrest hemodynamic trajectories of children suffering inhospital cardiac arrest. DESIGN: Exploratory retrospective analysis of arterial blood pressure and electrocardiogram waveforms. SETTING: PICU and cardiac critical care unit in a tertiary-care children's hospital. PATIENTS: Twenty-seven children with invasive blood pressure monitoring who suffered a total of 31 inhospital cardiac arrest events between June 2017 and June 2019. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We assessed changes in cardiac output, systemic vascular resistance, stroke volume, and heart rate derived from arterial blood pressure waveforms using three previously described estimation methods. We observed substantial prearrest drops in cardiac output (population median declines of 65-84% depending on estimation method) in all patients in the 10 minutes preceding inhospital cardiac arrest. Most patients' mean arterial blood pressure also decreased, but this was not universal. We identified three hemodynamic patterns preceding inhospital cardiac arrest: subacute pulseless arrest (n = 18), acute pulseless arrest (n = 7), and bradycardic arrest (n = 6). Acute pulseless arrest events decompensated within seconds, whereas bradycardic and subacute pulseless arrest events deteriorated over several minutes. In the subacute and acute pulseless arrest groups, decreases in cardiac output were primarily due to declines in stroke volume, whereas in the bradycardic group, the decreases were primarily due to declines in heart rate. CONCLUSIONS: Critically ill children exhibit distinct physiologic behaviors prior to inhospital cardiac arrest. All events showed substantial declines in cardiac output shortly before inhospital cardiac arrest. We describe three distinct prearrest patterns with varying rates of decline and varying contributions of heart rate and stroke volume changes to the fall in cardiac output. Our findings suggest that monitoring changes in arterial blood pressure waveform-derived heart rate, pulse pressure, cardiac output, and systemic vascular resistance estimates could improve early detection of inhospital cardiac arrest by up to several minutes. Further study is necessary to verify the patterns witnessed in our cohort as a step toward patient rather than provider-centered definitions of inhospital cardiac arrest.

Blood Pressure in Critically Ill Children: Exploratory Analyses of Concurrent Invasive and Noninvasive Measurements.

OBJECTIVES: Differences and biases between directly measured intra-arterial blood pressure and intermittingly measured noninvasive blood pressure using an oscillometric cuff method have been reported in adults and children. At the bedside, clinicians are required to assign a confidence to a specific blood pressure measurement before acting upon it, and this is challenging when there is discordance between measurement techniques. We hypothesized that big data could define and quantify the relationship between noninvasive blood pressure and intra-arterial blood pressure measurements and how they can be influenced by patient characteristics, thereby aiding bedside decision-making. DESIGN: A retrospective analysis of cuff blood pressure readings with associated concurrent invasive arterial blood pressure measurements (452,195 noninvasive blood pressure measurements). SETTING: Critical care unit at The Hospital for Sick Children, Toronto. PATIENTS: Six-thousand two-hundred ninety-seven patients less than or equal to 18 years old, hospitalized in a critical care unit with an indwelling arterial line. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Two-dimensional distributions of intra-arterial blood pressure and noninvasive blood pressure were generated and the conditional distributions of intra-arterial blood pressure examined as a function of the noninvasive systolic, diastolic, or mean blood pressure. Modification of these distributions according to age and gender were examined using a multilevel mixed-effects model. For any given combination of patient age and noninvasive blood pressure, the expected distribution of intra-arterial blood pressure readings exhibited marked variability at the population level and a bias that significantly depended on the noninvasive blood pressure value and age. We developed an online tool that allows exploration of the relationship between noninvasive blood pressure and intra-arterial blood pressure and the conditional probability distributions according to age. CONCLUSIONS: A large physiologic dataset provides clinically applicable insights into the relationship between noninvasive blood pressure and intra-arterial blood pressure measurements that can help guide decision-making at the patient bedside.

A Comparative Analysis of Equations to Estimate Patient Energy Requirements Following Cardiopulmonary Bypass for Correction of Congenital Heart Disease.

BACKGROUND: No consensus exists on the optimal method to estimate resting energy expenditure (REE) in critically ill children following cardiopulmonary bypass (CPB). This study assesses the accuracy of REE estimation equations in children with congenital heart disease following CPB and tests the feasibility of using allometric scaling as an alternative energy prediction equation. METHODS: A retrospective analysis of a pediatric cohort following CPB (n = 107; median age 5.2 months, median weight 5.65 kg) who underwent serial measures (median 5 measurements) of REE using indirect calorimetry for 72 hours following CPB. We estimated REE using common estimation methods (Dietary Reference Intake, Harris Benedict, Schofield, World Health Organization [WHO]) as well as novel allometric equations. We compared estimated with measured REE to determine accuracy of each equation using overall discrepancy, calculated as a time-weighted average of the absolute deviation. RESULTS: All equations incorrectly estimated REE at all time points following CPB, with overestimation error predominating. WHO had the lowest discrepancy at 10.7 ± 8.4 kcal/kg/d. The allometric equation was inferior, with an overall discrepancy of 16.9 ± 10.4. There is a strong nonlinear relationship between body surface area and measured REE in this cohort, which is a key source of estimation error using linear equations. CONCLUSION: In a cohort of pediatric patients with congenital heart disease following CPB, no currently utilized clinical estimation equation reliably estimated REE. Allometric scaling proved inferior in estimating REE in children following CPB. Indirect calorimetry remains the ideal method of determining REE after CPB until nonlinear methods can be derived due to overestimation using linear equations.

Temporal Variability in the Sampling of Vital Sign Data Limits the Accuracy of Patient State Estimation.

OBJECTIVES: Physiologic signals are typically measured continuously in the critical care unit, but only recorded at intermittent time intervals in the patient health record. Low frequency data collection may not accurately reflect the variability and complexity of these signals or the patient's clinical state. We aimed to characterize how increasing the temporal window size of observation from seconds to hours modifies the measured variability and complexity of basic vital signs. DESIGN: Retrospective analysis of signal data acquired between April 1, 2013, and September 30, 2015. SETTING: Critical care unit at The Hospital for Sick Children, Toronto. PATIENTS: Seven hundred forty-seven patients less than or equal to 18 years old (63,814,869 data values), within seven diagnostic/surgical groups. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Measures of variability (SD and the absolute differences) and signal complexity (multiscale sample entropy and detrended fluctuation analysis [expressed as the scaling component α]) were calculated for systolic blood pressure, heart rate, and oxygen saturation. The variability of all vital signs increases as the window size increases from seconds to hours at the patient and diagnostic/surgical group level. Significant differences in the magnitude of variability for all time scales within and between groups was demonstrated (p < 0.0001). Variability correlated negatively with patient age for heart rate and oxygen saturation, but positively with systolic blood pressure. Changes in variability and complexity of heart rate and systolic blood pressure from time of admission to discharge were found. CONCLUSIONS: In critically ill children, the temporal variability of physiologic signals supports higher frequency data capture, and this variability should be accounted for in models of patient state estimation.

Diagnostic errors in paediatric cardiac intensive care.

IntroductionDiagnostic errors cause significant patient harm and increase costs. Data characterising such errors in the paediatric cardiac intensive care population are limited. We sought to understand the perceived frequency and types of diagnostic errors in the paediatric cardiac ICU. METHODS: Paediatric cardiac ICU practitioners including attending and trainee physicians, nurse practitioners, physician assistants, and registered nurses at three North American tertiary cardiac centres were surveyed between October 2014 and January 2015. RESULTS: The response rate was 46% (N=200). Most respondents (81%) perceived that diagnostic errors harm patients more than five times per year. More than half (65%) reported that errors permanently harm patients, and up to 18% perceived that diagnostic errors contributed to death or severe permanent harm more than five times per year. Medication side effects and psychiatric conditions were thought to be most commonly misdiagnosed. Physician groups also ranked pulmonary overcirculation and viral illness to be commonly misdiagnosed as bacterial illness. Inadequate care coordination, data assessment, and high clinician workload were cited as contributory factors. Delayed diagnostic studies and interventions related to the severity of the patient's condition were thought to be the most commonly reported process breakdowns. All surveyed groups ranked improving teamwork and feedback pathways as strategies to explore for preventing future diagnostic errors. CONCLUSIONS: Paediatric cardiac intensive care practitioners perceive that diagnostic errors causing permanent harm are common and associated more with systematic and process breakdowns than with cognitive limitations.

Associations Between Unplanned Cardiac Reinterventions and Outcomes After Pediatric Cardiac Operations.

BACKGROUND: After pediatric heart operations, we sought to determine the incidence of unplanned cardiac reinterventions during the same hospitalization, assess risk factors for these reinterventions, and explore associations between reinterventions and outcomes. We hypothesized that younger patients undergoing more complex operations would be at greater risk for unplanned cardiac reinterventions and that operative mortality and postoperative length of stay (PLOS) would be greater in patients who undergo reintervention than in those who do not. METHODS: Patients aged 18 years or younger in The Society of Thoracic Surgeons Congenital Heart Surgery Database (January 2010 to June 2015) were included. We used multivariable regression to evaluate risk factors for unplanned cardiac reintervention (operation or therapeutic catheterization) and associations of reintervention with operative mortality and PLOS. RESULTS: Of 84,404 patients (117 centers), 21% were neonates and 36% infants. An unplanned cardiac reintervention was performed in 5.4% of patients, including 11.8% of neonates, 5.2% of infants, and 2.8% of children. Independent risk factors for unplanned reintervention included presence of noncardiac anomalies/genetic syndromes, nonwhite race, younger age, lower weight among neonates and infants, prior cardiothoracic operations, preoperative mechanical ventilation, other Society of Thoracic Surgeons preoperative risk factors, and higher Society of Thoracic Surgeons-European Association for Cardiothoracic Surgery Mortality Category (adjusted p < 0.001 for all). Unplanned reintervention was a risk factor for operative mortality (adjusted odds ratio, 5.3; 95% confidence interval, 4.8 to 5.8; p < 0.001) and longer PLOS (adjusted relative risk, 2.3; 95% confidence interval, 2.2 to 2.4; p < 0.001). CONCLUSIONS: Unplanned cardiac reinterventions are not rare, particularly in neonates, and are independently associated with operative mortality and increased PLOS. Patients at greater risk may be identified preoperatively, presenting opportunities for quality improvement.

The role of palliative care in critical congenital heart disease.

Patients with critical congenital heart disease are exposed to significant lifetime morbidity and mortality. Prenatal diagnosis can provide opportunities for anticipatory co-management of patients between palliative subspecialists and the cardiac care team. The benefits of palliative care include support for longitudinal decision-making and avoidance of interventions not consistent with family goals. Effectively counseling families requires an up-to-date understanding of outcomes and knowledge of provider biases. Patient-proxy reported quality of life (QOL) is highly variable in this population and healthcare providers need to be aware of limitations in their own subjective assessment of QOL.

Risk Factors for Cardiac Arrest or Mechanical Circulatory Support in Children with Fulminant Myocarditis.

In children with fulminant myocarditis (FM), we sought to describe presenting characteristics and clinical outcomes, and identify risk factors for cardiac arrest and mechanical circulatory support (MCS). A retrospective review of patients with FM admitted at our institution between January 1, 2004, and June 31, 2015, was performed. We compared characteristics and outcomes of FM patients who received cardiopulmonary resuscitation (CPR) and/or were placed on MCS (CPR/MCS group) to those who did not develop these outcomes (Control group). There were 28 patients who met criteria for FM. Median age was 1.2 years (1 day-17 years). Recovery of myocardial function occurred in 13 patients (46%); 6 (21%) had chronic ventricular dysfunction, 6 (21%) underwent heart transplantation, and 3 (11%) died prior to hospital discharge (including one death following heart transplant). Of the 28 FM patients, 13 (46%) developed cardiac arrest (n = 11) and/or received MCS (n = 8). When compared to controls, patients in the CPR/MCS group had a higher peak b-type natriuretic peptide (BNP) levels (p = 0.03) and peak inotropic scores (p = 0.02). No significant differences were found between groups in demographics; chest radiograph, electrocardiogram, or echocardiogram findings; or initial laboratory values including BNP, troponin, C-reactive protein, lactate, and creatinine (p > 0.05 for all). Children with FM are at high risk of cardiovascular collapse leading to the use of CPR or MCS. Aside from peak BNP levels and inotropic scores, the most presenting characteristics were not helpful for predicting these outcomes. FM patients should ideally receive care in centers that provide emergent MCS.

Shunt Lesions Part I: Patent Ductus Arteriosus, Atrial Septal Defect, Ventricular Septal Defect, and Atrioventricular Septal Defect.

OBJECTIVES: This review summarizes the current understanding of the pathophysiology and perioperative management of patent ductus arteriosus, atrial septal defect, ventricular septal defect, and atrioventricular septal defect. DATA SOURCE: MEDLINE and PubMed. CONCLUSIONS: The four congenital cardiac lesions that are the subject of this review, patent ductus arteriosus, atrial septal defect, ventricular septal defect, and atrioventricular septal defect, are the most commonly found defects causing a left-to-right shunt. These defects frequently warrant transcatheter or surgical intervention. Although the perioperative care is relatively straightforward for many of these patients, there are a number of management strategies and complications associated with each intervention. The treatment outcomes for all of these lesions are very good in the current era.

Pharmacological Manipulation of Peripheral Vascular Resistance in Special Clinical Situations after Pediatric Cardiac Surgery.

Pediatric cardiac surgery patients commonly suffer from alterations in vascular tone in the early post-operative period. Pharmacologic manipulation of systemic vascular resistance (SVR) can be complex in a variety of special patient situations including extremes of age, presence of left sided valvar lesions and the use of mechanical circulatory support. Familiarity with how these special circumstances alter SVR and the response to pharmacologic intervention will allow for tailored therapy and hopefully, optimized outcomes. This article addresses the eighth topic of the special issue entitled "Pharmacologic strategies with afterload reduction in low cardiac output syndrome after pediatric cardiac surgery".

Critical care for paediatric patients with heart failure.

This review offers a critical-care perspective on the pathophysiology, monitoring, and management of acute heart failure syndromes in children. An in-depth understanding of the cardiovascular physiological disturbances in this population of patients is essential to correctly interpret clinical signs, symptoms and monitoring data, and to implement appropriate therapies. In this regard, the myocardial force-velocity relationship, the Frank-Starling mechanism, and pressure-volume loops are discussed. A variety of monitoring modalities are used to provide insight into the haemodynamic state, clinical trajectory, and response to treatment. Critical-care treatment of acute heart failure is based on the fundamental principles of optimising the delivery of oxygen and minimising metabolic demands. The former may be achieved by optimising systemic arterial oxygen content and the variables that determine cardiac output: heart rate and rhythm, preload, afterload, and contractility. Metabolic demands may be decreased by a number of ways including positive pressure ventilation, temperature control, and sedation. Mechanical circulatory support should be considered for refractory cases. In the near future, monitoring modalities may be improved by the capture and analysis of complex clinical data such as pressure waveforms and heart rate variability. Using predictive modelling and streaming analytics, these data may then be used to develop automated, real-time clinical decision support tools. Given the barriers to conducting multi-centre trials in this population of patients, the thoughtful analysis of data from multi-centre clinical registries and administrative databases will also likely have an impact on clinical practice.

Unplanned reinterventions are associated with postoperative mortality in neonates with critical congenital heart disease.

OBJECTIVE: Neonates with critical congenital heart disease remain at risk of adverse outcomes after cardiac surgery. Residual or undiagnosed anatomic lesions might be contributory. The present study aimed to describe the incidence and type of cardiac lesions that lead to early, unplanned cardiac reintervention, identify the risk factors for unplanned reintervention, and explore the associations between unplanned reinterventions and hospital mortality. METHODS: The present single-center retrospective cohort study included 943 consecutive neonates with critical congenital heart disease who underwent cardiac surgery from 2002 to 2008. An unplanned cardiac reintervention was defined as a cardiac reoperation or interventional cardiac catheterization performed during the same hospitalization as the initial operation. Multivariate logistic regression analyses were used to identify the risk factors for unplanned cardiac reintervention and hospital mortality. RESULTS: Of the 943 neonates, 104 (11%) underwent an unplanned cardiac reintervention. The independent predictors of unplanned reintervention included prenatal diagnosis, lower birth weight, need for mechanical ventilation before the initial cardiac operation, lower attending surgeon experience, and greater Risk Adjustment in Congenital Heart Surgery, version 1, category. Those who underwent reintervention had increased hospital mortality (n = 33/104, 32%) relative to those who did not (n = 31/839, 4%; adjusted odds ratio, 8.6; 95% confidence interval, 4.7 to 15.6; P < .001). The mortality rates among patients undergoing surgical reintervention (23/66, 35%) or transcatheter reintervention (4/16, 25%), or both (6/22, 27%) were similar (P = .66). CONCLUSIONS: The need for unplanned cardiac reintervention in neonates with critical congenital heart disease is strongly associated with increased mortality. Early unplanned reinterventions might be an important covariate in outcomes studies and useful as a quality improvement measure.

Hypoplastic left heart syndrome is heritable.

OBJECTIVES: This study sought to determine the size of the genetic effect (heritability) in families identified by a hypoplastic left heart syndrome (HLHS) proband. BACKGROUND: Hypoplastic left heart syndrome is a severe form of cardiovascular malformation (CVM), and it remains a leading cause of infant mortality and childhood morbidity. Familial clustering of HLHS and bicuspid aortic valve (BAV) has been observed, and pedigree analysis has suggested recessive inheritance. The genetic significance of these observations is unknown. METHODS: In 38 probands with HLHS, a 3-generation family history was obtained; using a sequential sampling strategy, echocardiograms on family members were performed. A total of 235 participants were recruited. Heritability (h2) of HLHS and associated CVM was estimated using maximum-likelihood-based variance decomposition. RESULTS: All HLHS probands had aortic valve hypoplasia and dysplasia; dysplasia of the mitral (94%), tricuspid (56%), and pulmonary (11%) valves was also noted. Overall, 21 of 38 (55%) families had more than 1 affected individual, and 36% of participants had CVM, including 11% with BAV. The heritability of HLHS alone and with associated CVM were 99% and 74% (p < 0.00001), respectively. The sibling recurrence risk for HLHS was 8%, and for CVM was 22%. CONCLUSIONS: The high heritability of HLHS suggests that it is determined largely by genetic factors. The frequent occurrence of left- and right-sided valve dysplasia in HLHS probands and the increased prevalence of BAV in family members suggests that HLHS is a severe form of valve malformation.