Prediction of imminent, severe deterioration of children with parallel circulations using real-time processing of physiologic data

A retrospective evaluation of SwePEWS use in paediatric patients with COVID-19 and RSV infection

AIM

As early detection of deterioration is a challenge in children, the Swedish Paediatric Early Warning Score (SwePEWS) is used to systematically assess paediatric patients' clinical state. Here, we aimed to evaluate the use and predictive ability of SwePEWS.

METHODS

Electronic health records of paediatric patients admitted due to respiratory syncytial virus infection or COVID-19 were reviewed retrospectively. Registered vital signs were compared to the assigned SwePEWS score and monitored vital sign values to identify discrepancies. Additionally, SwePEWS's ability to predict transfer to the paediatric intensive care unit (PICU) was assessed.

RESULTS

Among 1374 SwePEWS assessments, one-third were either incomplete or contained errors. Incomplete SwePEWS assessments were more frequent during night-time. Single measurements of oxygen saturation presented lower values compared to average saturation from continuous monitoring. SwePEWS's ability to predict PICU transfer was low.

CONCLUSION

There was a surprisingly high occurrence of underestimated SwePEWS scores. This study provides new insights into pitfalls when developing and implementing paediatric early warning scores for systematic re-evaluations in paediatric patients.

Can perioperative electroencephalogram and adverse hemodynamic events predict neurodevelopmental outcomes in infants with congenital heart disease?

OBJECTIVE

The study objective was to characterize preoperative and postoperative continuous electroencephalogram metrics and hemodynamic adverse events as predictors of neurodevelopment in congenital heart disease infants undergoing cardiac surgery.

METHODS

From 2010 to 2021, 320 infants underwent congenital heart disease surgery at our institution, of whom 217 had perioperative continuous electroencephalogram monitoring and were included in our study. Neurodevelopment was assessed in 76 patients by the Bayley Scales of Infant and Toddler Development, 3rd edition, consisting of cognitive, communication, and motor scaled scores. Patient and procedural factors, including hemodynamic adverse events, were included by means of the likelihood of covariate selection in our predictive model. Median (25th, 75th percentile) follow-up was 1.03 (0.09, 3.44) years with 3 (1, 6) Bayley Scales of Infant and Toddler Development, 3rd Edition evaluations per patient.

RESULTS

Median age at index surgery was 7 (4, 23) days, and 81 (37%) were female. Epileptiform discharges, encephalopathy, and abnormality (lethargy and coma) were more prevalent on postoperative continuous electroencephalograms, compared with preoperative continuous electroencephalograms (P < .005). In 76 patients with Bayley Scales of Infant and Toddler Development, 3rd edition evaluations, patients with diffuse abnormality (P = .009), waveform discontinuity (P = .007), and lack of continuity (P = .037) on preoperative continuous electroencephalogram had lower cognitive scores. Patients with synchrony (P < .005) on preoperative and waveform continuity (P = .009) on postoperative continuous electroencephalogram had higher fine motor scores. Patients with postoperative adverse events had lower cognitive (P < .005) and gross motor scores (P < .005).

CONCLUSIONS

Phenotypic patterns of perioperative continuous electroencephalogram metrics are associated with late-term neurologic injury in infants with congenital heart disease requiring surgery. Continuous electroencephalogram metrics can be integrated with hemodynamic adverse events in a predictive algorithm for neurologic impairment.

Precision Anesthesia in 2050

Over the last few decades, the field of anesthesia has advanced far beyond its humble beginnings. Today's anesthetics are better and safer than ever, thanks to innovations in drugs, monitors, equipment, and patient safety.1-4 At the same time, we remain limited by our herd approach to medicine. Each of our patients is unique, but health care today is based on a one-size-fits-all approach, while our patients grow older and more medically complex every year. By 2050, we believe that precision medicine will play a central role across all medical specialties, including anesthesia. In addition, we expect that health care and consumer technology will continually evolve to improve and simplify the interactions between patients, providers, and the health care system. As demonstrated by 2 hypothetical patient experiences, these advancements will enable more efficient and safe care, earlier and more accurate diagnoses, and truly personalized treatment plans.

Development and Temporal Validation of a Machine Learning Model to Predict Clinical Deterioration

OBJECTIVES

Early warning scores detecting clinical deterioration in pediatric inpatients have wide-ranging performance and use a limited number of clinical features. This study developed a machine learning model leveraging multiple static and dynamic clinical features from the electronic health record to predict the composite outcome of unplanned transfer to the ICU within 24 hours and inpatient mortality within 48 hours in hospitalized children.

METHODS

Using a retrospective development cohort of 17 630 encounters across 10 388 patients, 2 machine learning models (light gradient boosting machine [LGBM] and random forest) were trained on 542 features and compared with our institutional Pediatric Early Warning Score (I-PEWS).

RESULTS

The LGBM model significantly outperformed I-PEWS based on receiver operating characteristic curve (AUROC) for the composite outcome of ICU transfer or mortality for both internal validation and temporal validation cohorts (AUROC 0.785 95% confidence interval [0.780-0.791] vs 0.708 [0.701-0.715] for temporal validation) as well as lead-time before deterioration events (median 11 hours vs 3 hours; P = .004). However, LGBM performance as evaluated by precision recall curve was lesser in the temporal validation cohort with associated decreased positive predictive value (6% vs 29%) and increased number needed to evaluate (17 vs 3) compared with I-PEWS.

CONCLUSIONS

Our electronic health record based machine learning model demonstrated improved AUROC and lead-time in predicting clinical deterioration in pediatric inpatients 24 to 48 hours in advance compared with I-PEWS. Further work is needed to optimize model positive predictive value to allow for integration into clinical practice.

Forecasting imminent atrial fibrillation in long-term electrocardiogram recordings

BACKGROUND

Despite the morbidity associated with acute atrial fibrillation (AF), no models currently exist to forecast its imminent onset. We sought to evaluate the ability of deep learning to forecast the imminent onset of AF with sufficient lead time, which has important implications for inpatient care.

METHODS

We utilized the Physiobank Long-Term AF Database, which contains 24-h, labeled ECG recordings from patients with a history of AF. AF episodes were defined as ≥5 min of sustained AF. Three deep learning models incorporating convolutional and transformer layers were created for forecasting, with two models focusing on the predictive nature of sinus rhythm segments and AF epochs separately preceding an AF episode, and one model utilizing all preceding waveform as input. Cross-validated performance was evaluated using area under time-dependent receiver operating characteristic curves (AUC(t)) at 7.5-, 15-, 30-, and 60-min lead times, precision-recall curves, and imminent AF risk trajectories.

RESULTS

There were 367 AF episodes from 84 ECG recordings. All models showed average risk trajectory divergence of those with an AF episode from those without ∼15 min before the episode. Highest AUC was associated with the sinus rhythm model [AUC = 0.74; 7.5-min lead time], though the model using all preceding waveform data had similar performance and higher AUCs at longer lead times.

CONCLUSIONS

In this proof-of-concept study, we demonstrated the potential utility of neural networks to forecast the onset of AF in long-term ECG recordings with a clinically relevant lead time. External validation in larger cohorts is required before deploying these models clinically.

Markov model for detection of ECG instability prior to cardiac arrest in single-ventricle patients

OBJECTIVES

Assess the degree of instability in the electrocardiogram (ECG) waveform in patients with single-ventricle physiology before a cardiac arrest and compare them with similar patients who did not experience a cardiac arrest.

METHODS

Retrospective control study in patients with single-ventricle physiology who underwent Norwood, Blalock-Taussig shunt, pulmonary artery band, and aortic arch repair from 2013 to 2018. Electronic medical records were obtained for all included patients. For each subject, 6 h of ECG data were analyzed. In the arrest group, the end of the sixth hour coincides with the cardiac arrest. In the control group, the 6-h windows were randomly selected. We used a Markov chain framework and the likelihood ratio test to measure the degree of ECG instability and to classify the arrest and control groups.

RESULTS

The study dataset consists of 38 cardiac arrest events and 67 control events. Our Markov model was able to classify the arrest and control groups based on the ECG instability with an ROC AUC of 82% at the hour preceding the cardiac arrests.

CONCLUSION

We designed a method using the Markov chain framework to measure the level of instability in the beat-to-beat ECG morphology. Furthermore, we were able to show that the Markov model performed well to distinguish patients in the arrest group compared to the control group.

Automated prediction of cardiorespiratory deterioration in patients with single-ventricle parallel circulation: A multicenter validation study

Objectives

Patients with single-ventricle physiology have a significant risk of cardiorespiratory deterioration between their first- and second-stage palliation surgeries. Detection of deterioration episodes may allow for early intervention and improved outcomes.

Methods

A prospective study was executed at Nationwide Children's Hospital, Children's Hospital of Philadelphia, and Children's Hospital Colorado to collect physiologic data of subjects with single ventricle physiology during all hospitalizations between neonatal palliation and II surgeries using the Sickbay software platform (Medical Informatics Corp). Timing of cardiorespiratory deterioration events was captured via chart review. The predictive algorithm previously developed and validated at Texas Children's Hospital was applied to these data without retraining. Standard metrics such as receiver operating curve area, positive and negative likelihood ratio, and alert rates were calculated to establish clinical performance of the predictive algorithm.

Results

Our cohort consisted of 58 subjects admitted to the cardiac intensive care unit and stepdown units of participating centers over 14 months. Approximately 28,991 hours of high-resolution physiologic waveform and vital sign data were collected using the Sickbay. A total of 30 cardiorespiratory deterioration events were observed. the risk index metric generated by our algorithm was found to be both sensitive and specific for detecting impending events one to two hours in advance of overt extremis (receiver operating curve = 0.927).

Conclusions

Our algorithm can provide a 1- to 2-hour advanced warning for 53.6% of all cardiorespiratory deterioration events in children with single ventricle physiology during their initial postop course as well as interstage hospitalizations after stage I palliation with only 2.5 alarms being generated per patient per day.

Continuous Data-Driven Monitoring in Critical Congenital Heart Disease: Clinical Deterioration Model Development

BACKGROUND

Critical congenital heart disease (cCHD)-requiring cardiac intervention in the first year of life for survival-occurs globally in 2-3 of every 1000 live births. In the critical perioperative period, intensive multimodal monitoring at a pediatric intensive care unit (PICU) is warranted, as their organs-especially the brain-may be severely injured due to hemodynamic and respiratory events. These 24/7 clinical data streams yield large quantities of high-frequency data, which are challenging in terms of interpretation due to the varying and dynamic physiology innate to cCHD. Through advanced data science algorithms, these dynamic data can be condensed into comprehensible information, reducing the cognitive load on the medical team and providing data-driven monitoring support through automated detection of clinical deterioration, which may facilitate timely intervention.

OBJECTIVE

This study aimed to develop a clinical deterioration detection algorithm for PICU patients with cCHD.

METHODS

Retrospectively, synchronous per-second data of cerebral regional oxygen saturation (rSO₂) and 4 vital parameters (respiratory rate, heart rate, oxygen saturation, and invasive mean blood pressure) in neonates with cCHD admitted to the University Medical Center Utrecht, the Netherlands, between 2002 and 2018 were extracted. Patients were stratified based on mean oxygen saturation during admission to account for physiological differences between acyanotic and cyanotic cCHD. Each subset was used to train our algorithm in classifying data as either stable, unstable, or sensor dysfunction. The algorithm was designed to detect combinations of parameters abnormal to the stratified subpopulation and significant deviations from the patient's unique baseline, which were further analyzed to distinguish clinical improvement from deterioration. Novel data were used for testing, visualized in detail, and internally validated by pediatric intensivists.

RESULTS

A retrospective query yielded 4600 hours and 209 hours of per-second data in 78 and 10 neonates for, respectively, training and testing purposes. During testing, stable episodes occurred 153 times, of which 134 (88%) were correctly detected. Unstable episodes were correctly noted in 46 of 57 (81%) observed episodes. Twelve expert-confirmed unstable episodes were missed in testing. Time-percentual accuracy was 93% and 77% for, respectively, stable and unstable episodes. A total of 138 sensorial dysfunctions were detected, of which 130 (94%) were correct.

CONCLUSIONS

In this proof-of-concept study, a clinical deterioration detection algorithm was developed and retrospectively evaluated to classify clinical stability and instability, achieving reasonable performance considering the heterogeneous population of neonates with cCHD. Combined analysis of baseline (ie, patient-specific) deviations and simultaneous parameter-shifting (ie, population-specific) proofs would be promising with respect to enhancing applicability to heterogeneous critically ill pediatric populations. After prospective validation, the current-and comparable-models may, in the future, be used in the automated detection of clinical deterioration and eventually provide data-driven monitoring support to the medical team, allowing for timely intervention.

Predicting Cardiac Arrest in Children with Heart Disease: A Novel Machine Learning Algorithm

BACKGROUND

Children with congenital and acquired heart disease are at a higher risk of cardiac arrest compared to those without heart disease. Although the monitoring of cardiopulmonary resuscitation quality and extracorporeal resuscitation technologies have advanced, survival after cardiac arrest in this population has not improved. Cardiac arrest prevention, using predictive algorithms with machine learning, has the potential to reduce cardiac arrest rates. However, few studies have evaluated the use of these algorithms in predicting cardiac arrest in children with heart disease.

METHODS

We collected demographic, laboratory, and vital sign information from the electronic health records (EHR) of all the patients that were admitted to a single-center pediatric cardiac intensive care unit (CICU), between 2010 and 2019, who had a cardiac arrest during their CICU admission, as well as a comparator group of randomly selected non-cardiac-arrest controls. We compared traditional logistic regression modeling against a novel adaptation of a machine learning algorithm (functional gradient boosting), using time series data to predict the risk of cardiac arrest.

RESULTS

A total of 160 unique cardiac arrest events were matched to non-cardiac-arrest time periods. Using 11 different variables (vital signs and laboratory values) from the EHR, our algorithm's peak performance for the prediction of cardiac arrest was at one hour prior to the cardiac arrest (AUROC of 0.85 [0.79,0.90]), a performance that was similar to our previously published multivariable logistic regression model.

CONCLUSIONS

Our novel machine learning predictive algorithm, which was developed using retrospective data that were collected from the EHR and predicted cardiac arrest in the children that were admitted to a single-center pediatric cardiac intensive care unit, demonstrated a performance that was similar to that of a traditional logistic regression model. While these results are encouraging, future research, including prospective validations with multicenter data, is warranted prior to the implementation of this algorithm as a real-time clinical decision support tool.

Novel approaches to capturing and using continuous cardiorespiratory physiological data in hospitalized children

Continuous cardiorespiratory physiological monitoring is a cornerstone of care in hospitalized children. The data generated by monitoring devices coupled with machine learning could transform the way we provide care. This scoping review summarizes existing evidence on novel approaches to continuous cardiorespiratory monitoring in hospitalized children. We aimed to identify opportunities for the development of monitoring technology and the use of machine learning to analyze continuous physiological data to improve the outcomes of hospitalized children. We included original research articles published on or after January 1, 2001, involving novel approaches to collect and use continuous cardiorespiratory physiological data in hospitalized children. OVID Medline, PubMed, and Embase databases were searched. We screened 2909 articles and performed full-text extraction of 105 articles. We identified 58 articles describing novel devices or approaches, which were generally small and single-center. In addition, we identified 47 articles that described the use of continuous physiological data in prediction models, but only 7 integrated multidimensional data (e.g., demographics, laboratory results). We identified three areas for development: (1) further validation of promising novel devices; (2) more studies of models integrating multidimensional data with continuous cardiorespiratory data; and (3) further dissemination, implementation, and validation of prediction models using continuous cardiorespiratory data. IMPACT: We performed a comprehensive scoping review of novel approaches to capture and use continuous cardiorespiratory physiological data for monitoring, diagnosis, providing care, and predicting events in hospitalized infants and children, from novel devices to machine learning-based prediction models. We identified three key areas for future development: (1) further validation of promising novel devices; (2) more studies of models integrating multidimensional data with continuous cardiorespiratory data; and (3) further dissemination, implementation, and validation of prediction models using cardiorespiratory data.

The use of machine learning and artificial intelligence within pediatric critical care

The field of pediatric critical care has been hampered in the era of precision medicine by our inability to accurately define and subclassify disease phenotypes. This has been caused by heterogeneity across age groups that further challenges the ability to perform randomized controlled trials in pediatrics. One approach to overcome these inherent challenges include the use of machine learning algorithms that can assist in generating more meaningful interpretations from clinical data. This review summarizes machine learning and artificial intelligence techniques that are currently in use for clinical data modeling with relevance to pediatric critical care. Focus has been placed on the differences between techniques and the role of each in the clinical arena. The various forms of clinical decision support that utilize machine learning are also described. We review the applications and limitations of machine learning techniques to empower clinicians to make informed decisions at the bedside. IMPACT: Critical care units generate large amounts of under-utilized data that can be processed through artificial intelligence. This review summarizes the machine learning and artificial intelligence techniques currently being used to process clinical data. The review highlights the applications and limitations of these techniques within a clinical context to aid providers in making more informed decisions at the bedside.

Machine Learning-Based Systems for the Anticipation of Adverse Events After Pediatric Cardiac Surgery

Pediatric congenital heart disease (CHD) patients are at higher risk of postoperative complications and clinical deterioration either due to their underlying pathology or due to the cardiac surgery, contributing significantly to mortality, morbidity, hospital and family costs, and poor quality of life. In current clinical practice, clinical deterioration is detected, in most of the cases, when it has already occurred. Several early warning scores (EWS) have been proposed to assess children at risk of clinical deterioration using vital signs and risk indicators, in order to intervene in a timely manner to reduce the impact of deterioration and risk of death among children. However, EWS are based on measurements performed at a single time point without incorporating trends nor providing information about patient's risk trajectory. Moreover, some of these measurements rely on subjective assessment making them susceptible to different interpretations. All these limitations could explain why the implementation of EWS in high-resource settings failed to show a significant decrease in hospital mortality. By means of machine learning (ML) based algorithms we could integrate heterogeneous and complex data to predict patient's risk of deterioration. In this perspective article, we provide a brief overview of the potential of ML technologies to improve the identification of pediatric CHD patients at high-risk for clinical deterioration after cardiac surgery, and present the CORTEX traffic light, a ML-based predictive system that Sant Joan de Déu Barcelona Children's Hospital is implementing, as an illustration of the application of an ML-based risk stratification system in a relevant hospital setting.

Clinical Informatics and Quality Improvement in the Pediatric Intensive Care Unit

Clinical informatics can support quality improvement and patient safety in the pediatric intensive care unit (PICU) in several ways including data extraction, analysis, and decision support enabled by electronic health records (EHRs), and databases and registries. Clinical decision support (CDS), embedded in EHRs, now an integral part of the workflow in the PICU, includes several tools and is increasingly leveraging artificial intelligence (AI). Understanding the opportunities and challenges can improve the engagement of clinicians with the design, validation, and implementation of CDS, improve satisfaction with CDS, and improve patient safety, care quality, and value.

A Novel Situational Awareness Scoring System in Pediatric Cardiac Intensive Care Unit Patients

The aim of this study was to describe a novel Situational Awareness Scoring System (SASS)'s performance in discriminating between patients who had cardiac arrest (CA) and those who did not, in a pediatric cardiac intensive care unit (PCICU). Retrospective, observational-cohort study in a quaternary-care PCICU. Patients who had CA in the PCICU between January 2014 and December 2018, and patients admitted to the PCICU in 2018 who did not have CA were included. Patients with do not resuscitate or do not intubate orders, extracorporeal membrane oxygenation, ventricular assist device, and PCICU stay < 2 hours were excluded. SASS score statistics were calculated within 2, 4-, 6-, and 8-hour time intervals counting backward from the time of CA, or end of PCICU stay in patients who did not have CA. Cross-validated discrete time logistic regression models were used to calculate area under the receiver operating characteristic (AUC) curves. Odds ratios (ORs) for CA were calculated per unit increase of the SASS score. Twenty-eight CA events were analyzed in 462 PCICU admissions from 267 patients. Maximum SASS score within 4-hour time interval before CA achieved the highest AUC of 0.91 (95% confidence interval [CI]: 0.86–0.96) compared with maximum SASS score within 2-, 6-, and 8-hour time intervals before CA of 0.88 (0.79–96), 0.90 (0.85–0.95), and 0.89 (0.83–0.95), respectively. A cutoff value of 60 for maximum SASS score within 4-hour time interval before CA resulted in 82.1 and 83.2% of sensitivity and specificity, respectively. OR for CA was 1.32 (95% CI: 1.26–1.39) for every 10 units increase in the maximum SASS score within each 4-hour time interval before CA. The maximum SASS score within various time intervals before CA achieved promising performance in discriminating patients regarding occurrence of CA.

High-resolution perioperative cerebral blood flow autoregulation measurement: a practical and feasible approach for widespread clinical monitoring

A growing body of evidence demonstrates that excursions of BP below or above the limits of cerebral blood flow autoregulation are associated with complications in patients with neurological injury or for those undergoing cardiac surgery. Moreover, recent evidence suggests that maintaining MAP above the lower limit of cerebral autoregulation during cardiopulmonary bypass reduces the frequency of postoperative delirium and is associated with improved memory 1 month after surgery. Continuous measurement of BP in relation to cerebral autoregulation limits using a virtual patient monitoring platform processing near-infrared spectroscopy digital signals offers the hope of bringing this application to the bedside.

Early prediction of clinical deterioration using data-driven machine-learning modeling of electronic health records

OBJECTIVES

To develop and evaluate a high-dimensional, data-driven model to identify patients at high risk of clinical deterioration from routinely collected electronic health record (EHR) data.

MATERIALS AND METHODS

In this single-center, retrospective cohort study, 488 patients with single-ventricle and shunt-dependent congenital heart disease <6 months old were admitted to the cardiac intensive care unit before stage 2 palliation between 2014 and 2019. Using machine-learning techniques, we developed the Intensive care Warning Index (I-WIN), which systematically assessed 1028 regularly collected EHR variables (vital signs, medications, laboratory tests, and diagnoses) to identify patients in the cardiac intensive care unit at elevated risk of clinical deterioration. An ensemble of 5 extreme gradient boosting models was developed and validated on 203 cases (130 emergent endotracheal intubations, 34 cardiac arrests requiring cardiopulmonary resuscitation, 10 extracorporeal membrane oxygenation cannulations, and 29 cardiac arrests requiring cardiopulmonary resuscitation onto extracorporeal membrane oxygenation) and 378 control periods from 446 patients.

RESULTS

At 4 hours before deterioration, the model achieved an area under the receiver operating characteristic curve of 0.92 (95% confidence interval, 0.84-0.98), 0.881 sensitivity, 0.776 positive predictive value, 0.862 specificity, and 0.571 Brier skill score. Performance remained high at 8 hours before deterioration with 0.815 (0.688-0.921) area under the receiver operating characteristic curve.

CONCLUSIONS

I-WIN accurately predicted deterioration events in critically-ill infants with high-risk congenital heart disease up to 8 hours before deterioration, potentially allowing clinicians to target interventions. We propose a paradigm shift from conventional expert consensus-based selection of risk factors to a data-driven, machine-learning methodology for risk prediction. With the increased availability of data capture in EHRs, I-WIN can be extended to broader applications in data-rich environments in critical care.

Use of a Risk Analytic Algorithm to Inform Weaning From Vasoactive Medication in Patients Following Pediatric Cardiac Surgery

OBJECTIVES

Advanced clinical decision support tools, such as real-time risk analytic algorithms, show promise in assisting clinicians in making more efficient and precise decisions. These algorithms, which calculate the likelihood of a given underlying physiology or future event, have predominantly been used to identify the risk of impending clinical decompensation. There may be broader clinical applications of these models. Using the inadequate delivery of oxygen index, a U.S. Food and Drug Administration-approved risk analytic algorithm predicting the likelihood of low cardiac output state, the primary objective was to evaluate the association of inadequate delivery of oxygen index with success or failure of weaning vasoactive support in postoperative cardiac surgery patients.

DESIGN

Multicenter retrospective cohort study.

SETTING

Three pediatric cardiac ICUs at tertiary academic children's hospitals.

PATIENTS

Infants and children greater than 2 kg and less than 12 years following cardiac surgery, who required vasoactive infusions for greater than 6 hours in the postoperative period.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Postoperative patients were identified who successfully weaned off initial vasoactive infusions (n = 2,645) versus those who failed vasoactive wean (required reinitiation of vasoactive, required mechanical circulatory support, renal replacement therapy, suffered cardiac arrest, or died) (n = 516). Inadequate delivery of oxygen index for final 6 hours of vasoactive wean was captured. Inadequate delivery of oxygen index was significantly elevated in patients with failed versus successful weans (inadequate delivery of oxygen index 11.6 [sd 19.0] vs 6.4 [sd 12.6]; p < 0.001). Mean 6-hour inadequate delivery of oxygen index greater than 50 had strongest association with failed vasoactive wean (adjusted odds ratio, 4.0; 95% CI, 2.5-6.6). In patients who failed wean, reinitiation of vasoactive support was associated with concomitant fall in inadequate delivery of oxygen index (11.1 [sd 18] vs 8.9 [sd 16]; p = 0.007).

CONCLUSIONS

During the de-escalation phase of postoperative cardiac ICU management, elevation of the real-time risk analytic model, inadequate delivery of oxygen index, was associated with failure to wean off vasoactive infusions. Future studies should prospectively evaluate utility of risk analytic models as clinical decision support tools in de-escalation practices in critically ill patients.

Automated Prediction of Cardiorespiratory Deterioration in Patients With Single Ventricle

BACKGROUND

Patients with single-ventricle physiology have a significant risk of cardiorespiratory deterioration between their first and second stage palliation surgeries.

OBJECTIVES

The objective of this study is to develop and validate a real-time computer algorithm that can automatically recognize physiological precursors of cardiorespiratory deterioration in children with single-ventricle physiology during their interstage hospitalization.

METHODS

A retrospective study was conducted from prospectively collected physiological data of subjects with single-ventricle physiology. Deterioration events were defined as a cardiac arrest requiring cardiopulmonary resuscitation or an unplanned intubation. Physiological metrics were derived from the electrocardiogram (heart rate, heart rate variability, ST-segment elevation, and ST-segment variability) and the photoplethysmogram (peripheral oxygen saturation and pleth variability index). A logistic regression model was trained to separate the physiological dynamics of the pre-deterioration phase from all other data generated by study subjects. Data were split 50/50 into model training and validation sets to enable independent model validation.

RESULTS

Our cohort consisted of 238 subjects admitted to the cardiac intensive care unit and stepdown units of Texas Children's Hospital over a period of 6 years. Approximately 300,000 h of high-resolution physiological waveform and vital sign data were collected using the Sickbay software platform (Medical Informatics Corp., Houston, Texas). A total of 112 cardiorespiratory deterioration events were observed. Seventy-two of the subjects experienced at least 1 deterioration event. The risk index metric generated by our optimized algorithm was found to be both sensitive and specific for detecting impending events 1 to 2 h in advance of overt extremis (receiver-operating characteristic curve area: 0.958; 95% confidence interval: 0.950 to 0.965).

CONCLUSIONS

Our algorithm can provide 1 to 2 h of advanced warning for 62% of all cardiorespiratory deterioration events in children with single-ventricle physiology during their interstage period, with only 1 alarm being generated at the bedside per patient per day.

RBC Transfusion Induced ST Segment Variability Following the Norwood Procedure

Supplemental Digital Content is available in the text.

Objectives:

The transfusion of stored RBCs decreases nitric oxide bioavailability, which may have an adverse effect on vascular function. We assessed the effects of RBC transfusion on coronary vascular function by evaluating the relationship between myocardial oxygen delivery and demand as evidenced by ST segment variability.

Design:

Retrospective case-control study.

Setting:

Nine-hundred seventy-three–bed pediatric hospital with a 54-bed cardiovascular ICU.

Patients:

Seventy-three neonates with hypoplastic left heart syndrome following the Norwood procedure, 38 with a Blalock-Taussig shunt and 35 with a right ventricle to pulmonary artery shunt.

Interventions:

RBC transfusion.

Materials and Main Results:

High-frequency physiologic data were captured 30 minutes prior to the initiation of (baseline) and during the 120 minutes of the transfusion. A rate pressure product was calculated for each subject and used as an indicator of myocardial oxygen demand. Electrocardiogram leads (aVL, V1, II) were used to construct a 3D ST segment vector to assess ST segment variability and functioned as a surrogate indicator of myocardial ischemia. One-hundred thirty-eight transfusions occurred in the Blalock-Taussig shunt group and 139 in the right ventricle to pulmonary artery shunt group. There was no significant change in the rate pressure product for either group; however, ST segment variability progressively increased for the entire cohort during the transfusion, becoming statistically significant by the end of the transfusion. Upon subgroup analysis, this finding was noted with statistical significance in the Blalock-Taussig shunt group and trending toward significance in the right ventricle to pulmonary artery shunt group.

Conclusions:

We found a significant increase in the ST segment variability and evidence of myocardial ischemia temporally associated with RBC transfusions in neonates following the Norwood procedure, specifically among those in the Blalock-Taussig shunt group, which may impact immediate and long-term outcomes.

Lessons Learned From the First Pilot Study of the Compensatory Reserve Index After Congenital Heart Surgery Requiring Cardiopulmonary Bypass

BACKGROUND

Early warning systems that utilize dense physiologic data and machine learning may aid prediction of decompensation after congenital heart surgery (CHS). The Compensatory Reserve Index (CRI) analyzes changing features of the pulse waveform to predict hemodynamic decompensation in adults, but it has never been studied after CHS. This study sought to understand the feasibility, safety, and potential utility of CRI monitoring after CHS with cardiopulmonary bypass (CPB).

METHODS

A single-center prospective pilot cohort of patients undergoing pulmonary valve replacement was studied. Compensatory Reserve Index was continuously measured from preoperative baseline through the first 24 postoperative hours. Average CRI values during selected procedural phases were compared between patients with an intensive care unit (ICU) length of stay (LOS) <3 days versus LOS ≥3 days.

RESULTS

Twenty-three patients were enrolled. On average, 17,445 (±3,152) CRI data points were collected and 0.33% (±0.40) of data were missing per patient. There were no adverse events related to monitoring. Five (21.7%) patients had an ICU LOS ≥3 days. Compared to the ICU LOS <3 days group, the ICU LOS ≥3 days group had a greater decrease in CRI from baseline to immediately after CPB (-0.3 ± 0.1 vs -0.1 ± 0.2, P = .003) and were less likely to recover to baseline CRI during the monitoring period (20% vs 83%, P = .017).

CONCLUSIONS

Compensatory Reserve Index monitoring after CHS with CPB seems feasible and safe. Early changes in CRI may precede meaningful clinical outcomes, but this requires further study.

Comparing regression and neural network techniques for personalized predictive analytics to promote lung protective ventilation in Intensive Care Units

Mechanical ventilation is a lifesaving tool and provides organ support for patients with respiratory failure. However, injurious ventilation due to inappropriate delivery of high tidal volume can initiate or potentiate lung injury. This could lead to acute respiratory distress syndrome, longer duration of mechanical ventilation, ventilator associated conditions and finally increased mortality. In this study, we explore the viability and compare machine learning methods to generate personalized predictive alerts indicating violation of the safe tidal volume per ideal body weight (IBW) threshold that is accepted as the upper limit for lung protective ventilation (LPV), prior to application to patients. We process streams of patient respiratory data recorded per minute from ventilators in an intensive care unit and apply several state-of-the-art time series prediction methods to forecast the behavior of the tidal volume metric per patient, 1 hour ahead. Our results show that boosted regression delivers better predictive accuracy than other methods that we investigated and requires relatively short execution times. Long short-term memory neural networks can deliver similar levels of accuracy but only after much longer periods of data acquisition, further extended by several hours computing time to train the algorithm. Utilizing Artificial Intelligence, we have developed a personalized clinical decision support tool that can predict tidal volume behavior within 10% accuracy and compare alerts recorded from a real world system to highlight that our models would have predicted violations 1 hour ahead and can therefore conclude that the algorithms can provide clinical decision support.

Machine Learning Applications in the Neuro ICU: A Solution to Big Data Mayhem?

The neurological ICU (neuro ICU) often suffers from significant limitations due to scarce resource availability for their neurocritical care patients. Neuro ICU patients require frequent neurological evaluations, continuous monitoring of various physiological parameters, frequent imaging, and routine lab testing. This amasses large amounts of data specific to each patient. Neuro ICU teams are often overburdened by the resulting complexity of data for each patient. Machine Learning algorithms (ML), are uniquely capable of interpreting high-dimensional datasets that are too difficult for humans to comprehend. Therefore, the application of ML in the neuro ICU could alleviate the burden of analyzing big datasets for each patient. This review serves to (1) briefly summarize ML and compare the different types of MLs, (2) review recent ML applications to improve neuro ICU management and (3) describe the future implications of ML to neuro ICU management.

The Compensatory Reserve Index Responds to Acute Hemodynamic Changes in Patients with Congenital Heart Disease: A Proof of Concept Study

Patients with congenital heart disease (CHD) who undergo cardiac procedures may become hemodynamically unstable. Predictive algorithms that utilize dense physiologic data may be useful. The compensatory reserve index (CRI) trends beat-to-beat progression from normovolemia (CRI = 1) to decompensation (CRI = 0) in hemorrhagic shock by continuously analyzing unique sets of features in the changing pulse photoplethysmogram (PPG) waveform. We sought to understand if the CRI accurately reflects changing hemodynamics during and after a cardiac procedure for patients with CHD. A transcatheter pulmonary valve replacement (TcPVR) model was used because left ventricular stroke volume decreases upon sizing balloon occlusion of the right ventricular outflow tract (RVOT) and increases after successful valve placement. A single-center, prospective cohort study was performed. The CRI was continuously measured to determine the change in CRI before and after RVOT occlusion and successful TcPVR. Twenty-six subjects were enrolled with a median age of 19 (interquartile range (IQR) 13-29) years. The mean (± standard deviation) CRI decreased from 0.66 ± 0.15 1-min before balloon inflation to 0.53 ± 0.16 (p = 0.03) 1-min after balloon deflation. The mean CRI increased from a pre-valve mean CRI of 0.63 [95% confidence interval (CI) 0.56-0.70] to 0.77 (95% CI 0.71-0.83) after successful TcPVR. In this study, the CRI accurately reflected acute hemodynamic changes associated with TcPVR. Further research is justified to determine if the CRI can be useful as an early warning tool in patients with CHD at risk for decompensation during and after cardiac procedures.

Use of Continuous Physiological Monitor Data to Evaluate Doxapram Therapy in Preterm Infants

INTRODUCTION

Evaluation of pharmacotherapy during intensive care treatment is commonly based on subjective, intermittent interpretations of physiological parameters. Real-time visualization and analysis may improve drug effect evaluation. We aimed to evaluate the effects of the respiratory stimulant doxapram objectively in preterm infants using continuous physiological parameters.

METHODS

In this longitudinal observational study, preterm infants who received doxapram therapy were eligible for inclusion. Physiological data (1 Hz) were used to assess respiration and to evaluate therapy effects. The oxygen saturation (SpO2)/fraction of inspired oxygen (FiO2) ratio and the area under the 89% SpO2 curve (duration × saturation depth below target) were calculated as measures of hypoxemia. Regression analyses were performed in 1-h timeframes to discriminate therapy failure (intubation or death) from success (no intubation).

RESULTS

Monitor data of 61 patients with a median postmenstrual age (PMA) at doxapram initiation of 28.7 (IQR 27.6-30.0) weeks were available. The success rate of doxapram therapy was 56%. Doxapram pharmacodynamics were reflected in an increased SpO2 and SpO2/FiO2 ratio as well as a decrease in episodes with saturations below target (SpO2 <89%). The SpO2/FiO2 ratio, corrected for PMA and mechanical ventilation before therapy start, discriminated best between therapy failure and success (highest AUC ROC of 0.83).

CONCLUSION

The use of continuous physiological monitor data enables objective and detailed interpretation of doxapram in preterm infants. The SpO2/FiO2 ratio is the best predictive parameter for therapy failure or success. Further implementation of real-time data analysis and treatment algorithms would provide new opportunities to treat newborns.

Predictive analytics in the pediatric intensive care unit for early identification of sepsis: capturing the context of age

BACKGROUND

Early recognition of patients at risk for sepsis is paramount to improve clinical outcomes. We hypothesized that subtle signatures of illness are present in physiological and biochemical time series of pediatric-intensive care unit (PICU) patients in the early stages of sepsis.

METHODS

We developed multivariate models in a retrospective observational cohort to predict the clinical diagnosis of sepsis in children. We focused on age as a predictor and asked whether random forest models, with their potential for multiple cut points, had better performance than logistic regression.

RESULTS

One thousand seven hundred and eleven admissions for 1425 patients admitted to a mixed cardiac and medical/surgical PICU were included. We identified, through individual chart review, 187 sepsis diagnoses that were not within 14 days of a prior sepsis diagnosis. Multivariate models predicted sepsis in the next 24 h: cross-validated C-statistic for logistic regression and random forest were 0.74 (95% confidence interval (CI): 0.71-0.77) and 0.76 (95% CI: 0.73-0.79), respectively.

CONCLUSIONS

Statistical models based on physiological and biochemical data already available in the PICU identify high-risk patients up to 24 h prior to the clinical diagnosis of sepsis. The random forest model was superior to logistic regression in capturing the context of age.

Comparing the predictive ability of a commercial artificial intelligence early warning system with physician judgement for clinical deterioration in hospitalised general internal medicine patients: a prospective observational study

OBJECTIVE

Our study compares physician judgement with an automated early warning system (EWS) for predicting clinical deterioration of hospitalised general internal medicine patients.

DESIGN

Prospective observational study of clinical predictions made at the end of the daytime work-shift for an academic general internal medicine floor team compared with the risk assessment from an automated EWS collected at the same time.

SETTING

Internal medicine teaching wards at a single tertiary care academic medical centre in the USA.

PARTICIPANTS

Intern physicians working on the internal medicine wards and an automated EWS (Rothman Index by PeraHealth).

OUTCOME

Clinical deterioration within 24 hours including cardiac or pulmonary arrest, rapid response team activation or unscheduled intensive care unit transfer.

RESULTS

We collected predictions for 1874 patient days and saw 35 clinical deteriorations (1.9%). The area under the receiver operating curve (AUROC) for the EWS was 0.73 vs 0.70 for physicians (p=0.571). A linear regression model combining physician and EWS predictions had an AUROC of 0.75, outperforming physicians (p=0.016) and the EWS (p=0.05).

CONCLUSIONS

There is no significant difference in the performance of the EWS and physicians in predicting clinical deterioration at 24 hours on an inpatient general medicine ward. A combined model outperformed either alone. The EWS and physicians identify partially overlapping sets of at-risk patients suggesting they rely on different cues or decision rules for their predictions.

TRIAL REGISTRATION NUMBER

NCT02648828.

A deep learning model for real-time mortality prediction in critically ill children

Background

The rapid development in big data analytics and the data-rich environment of intensive care units together provide unprecedented opportunities for medical breakthroughs in the field of critical care. We developed and validated a machine learning-based model, the Pediatric Risk of Mortality Prediction Tool (PROMPT), for real-time prediction of all-cause mortality in pediatric intensive care units.

Methods

Utilizing two separate retrospective observational cohorts, we conducted model development and validation using a machine learning algorithm with a convolutional neural network. The development cohort comprised 1445 pediatric patients with 1977 medical encounters admitted to intensive care units from January 2011 to December 2017 at Severance Hospital (Seoul, Korea). The validation cohort included 278 patients with 364 medical encounters admitted to the pediatric intensive care unit from January 2016 to November 2017 at Samsung Medical Center.

Results

Using seven vital signs, along with patient age and body weight on intensive care unit admission, PROMPT achieved an area under the receiver operating characteristic curve in the range of 0.89–0.97 for mortality prediction 6 to 60 h prior to death. Our results demonstrated that PROMPT provided high sensitivity with specificity and outperformed the conventional severity scoring system, the Pediatric Index of Mortality, in predictive ability. Model performance was indistinguishable between the development and validation cohorts.

Conclusions

PROMPT is a deep model-based, data-driven early warning score tool that can predict mortality in critically ill children and may be useful for the timely identification of deteriorating patients.

Electronic supplementary material

The online version of this article (10.1186/s13054-019-2561-z) contains supplementary material, which is available to authorized users.

The Inadequate Oxygen Delivery Index and Low Cardiac Output Syndrome Score As Predictors of Adverse Events Associated With Low Cardiac Output Syndrome Early After Cardiac Bypass

OBJECTIVES

To evaluate the effectiveness of two scoring systems, the inadequate oxygen delivery index, a risk analytics algorithm (Etiometry, Boston, MA) and the Low Cardiac Output Syndrome Score, in predicting adverse events recognized as indicative of low cardiac output syndrome within 72 hours of surgery.

DESIGN

A retrospective observational pair-matched study.

SETTING

Tertiary pediatric cardiac ICU.

PATIENTS

Children undergoing cardiac bypass for congenital heart defects. Cases experienced an adverse event linked to low cardiac output syndrome in the 72 hours following surgery (extracorporeal membrane oxygenation, renal replacement therapy, cardiopulmonary resuscitation, and necrotizing enterocolitis) and were matched with a control patient on criteria of procedure, diagnosis, and age who experienced no such event.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Of a total 536 bypass operations in the study period, 38 patients experienced one of the defined events. Twenty-eight cases were included in the study after removing patients who suffered an event after 72 hours or who had insufficient data. Clinical and laboratory data were collected to derive scores for the first 12 hours after surgery. The inadequate oxygen delivery index was calculated by Etiometry using vital signs and laboratory data. A modified Low Cardiac Output Syndrome Score was calculated from clinical and therapeutic markers. The mean inadequate oxygen delivery and modified Low Cardiac Output Syndrome Score were compared within each matched pair using the Wilcoxon signed-rank test. Inadequate oxygen delivery correctly differentiated adverse events in 13 of 28 matched pairs, with no evidence of inadequate oxygen delivery being higher in cases (p = 0.71). Modified Low Cardiac Output Syndrome Score correctly differentiated adverse events in 23 of 28 matched pairs, with strong evidence of a raised score in low cardiac output syndrome cases (p < 0.01).

CONCLUSIONS

Although inadequate oxygen delivery is an Food and Drug Administration approved indicator of risk for low mixed venous oxygen saturation, early postoperative average values were not linked with medium-term adverse events. The indicators included in the modified Low Cardiac Output Syndrome Score had a much stronger association with the specified adverse events.

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.

Hand-held echocardiography in children with hypoplastic left heart syndrome

BACKGROUND

When performed by cardiologists, hand-held echocardiography (HHE) can assess ventricular systolic function and valve disease in adults, but its accuracy and utility in congenital heart disease is unknown. In hypoplastic left heart syndrome (HLHS), the echocardiographic detection of depressed right ventricular (RV) systolic function and higher grade tricuspid regurgitation (TR) can identify patients who are at increased risk of morbidity and mortality and who may benefit from additional imaging or medical therapies.

METHODS

Children with HLHS after Stage I or II surgical palliation (Norwood or Glenn procedures) were prospectively enrolled. Subjects underwent HHE by a pediatric cardiologist on the same day as standard echocardiography (SE). Using 4-point scales, bedside HHE assessment of RV systolic function and TR were compared with blinded assessment of offline SE images. Concordance correlation coefficient (CCC) was used to evaluate agreement.

RESULTS

Thirty-two HHEs were performed on 15 subjects (Stage I: n = 17 and Stage II: n = 15). Median subject age was 3.4 months (14 days-4.2 years). Median weight was 5.9 kg (2.6-15.4 kg). Bedside HHE assessment of RV systolic function and TR severity had substantial agreement with SE (CCC = 0.80, CCC = 0.74, respectively; P < .001). HHE sensitivity and specificity for any grade of depressed RV systolic function were 100% and 92%, respectively, and were 94% and 88% for moderate or greater TR, respectively. Average HHE scan time was 238 seconds.

CONCLUSIONS

HHE offers a rapid, bedside tool for pediatric cardiologists to detect RV systolic dysfunction and hemodynamically significant TR in HLHS.

Early prediction of critical events for infants with single-ventricle physiology in critical care using routinely collected data

OBJECTIVE

Critical events are common and difficult to predict among infants with congenital heart disease and are associated with mortality and long-term sequelae. We aimed to achieve early prediction of critical events, that is, cardiopulmonary resuscitation, emergency endotracheal intubation, and extracorporeal membrane oxygenation in infants with single-ventricle physiology before second-stage surgery. We hypothesized that naïve Bayesian models learned from expert knowledge and clinical data can predict critical events early and accurately.

METHODS

We collected 93 patients with single-ventricle physiology admitted to intensive care units in a single tertiary pediatric hospital between 2014 and 2017. Using knowledge elicited from experienced cardiac-intensive-care-unit providers and machine-learning techniques, we developed and evaluated the Cardiac-intensive-care Warning INdex (C-WIN) system, consisting of a set of naïve Bayesian models that leverage routinely collected data. We evaluated predictive performance using the area under the receiver operating characteristic curve, sensitivity, and specificity. We performed the evaluation at 5 different prediction horizons: 1, 2, 4, 6, and 8 hours before the onset of critical events.

RESULTS

The area under the receiver operating characteristic curves of the C-WIN models ranged between 0.73 and 0.88 at different prediction horizons. At 1 hour before critical events, C-WIN was able to detect events with an area under the receiver operating characteristic curve of 0.88 (95% confidence interval, 0.84-0.92) and a sensitivity of 84% at the 81% specificity level.

CONCLUSIONS

Predictive models may enhance clinicians' ability to identify infants with single-ventricle physiology at high risk of critical events. Early prediction of critical events may indicate the need to perform timely interventions, potentially reducing morbidity, mortality, and health care costs.

The Impact of Big Data on Chronic Disease Management

Population health management and specifically chronic disease management depend on the ability of providers to prevent development of high-cost and high-risk conditions such as diabetes, heart failure, and chronic respiratory diseases and to control them. The advent of big data analytics has potential to empower health care providers to make timely and truly evidence-based informed decisions to provide more effective and personalized treatment while reducing the costs of this care to patients. The goal of this study was to identify real-world health care applications of big data analytics to determine its effectiveness in both patient outcomes and the relief of financial burdens. The methodology for this study was a literature review utilizing 49 articles. Evidence of big data analytics being largely beneficial in the areas of risk prediction, diagnostic accuracy and patient outcome improvement, hospital readmission reduction, treatment guidance, and cost reduction was noted. Initial applications of big data analytics have proved useful in various phases of chronic disease management and could help reduce the chronic disease burden.

Impact of predictive analytics based on continuous cardiorespiratory monitoring in a surgical and trauma intensive care unit

Predictive analytics monitoring, the use of patient data to provide continuous risk estimation of deterioration, is a promising new application of big data analytical techniques to the care of individual patients. We tested the hypothesis that continuous display of novel electronic risk visualization of respiratory and cardiovascular events would impact intensive care unit (ICU) patient outcomes. In an adult tertiary care surgical trauma ICU, we displayed risk estimation visualizations on a large monitor, but in the medical ICU in the same institution we did not. The risk estimates were based solely on analysis of continuous cardiorespiratory monitoring. We examined 4275 individual patient records within a 7 month time period preceding and following data display. We determined cases of septic shock, emergency intubation, hemorrhage, and death to compare rates per patient care pre-and post-implementation. Following implementation, the incidence of septic shock fell by half (p < 0.01 in a multivariate model that included age and APACHE) in the surgical trauma ICU, where the data were continuously on display, but by only 10% (p = NS) in the control Medical ICU. There were no significant changes in the other outcomes. Display of a predictive analytics monitor based on continuous cardiorespiratory monitoring was followed by a reduction in the rate of septic shock, even when controlling for age and APACHE score.

Current monitoring and innovative predictive modeling to improve care in the pediatric cardiac intensive care unit

The objectives of this review are (I) to describe the challenges associated with monitoring patients in the pediatric cardiac intensive care unit (PCICU) and (II) to discuss the use of innovative statistical and artificial intelligence (AI) software programs to attempt to predict significant clinical events. Patients cared for in the PCICU are clinically fragile and at risk for fatal decompensation. Current monitoring modalities are often ineffective, sometimes inaccurate, and fail to detect a deteriorating clinical status in a timely manner. Predictive models created by AI and machine learning may lead to earlier detection of patients at risk for clinical decompensation and thereby improve care for critically ill pediatric cardiac patients.

Finding representative electrocardiogram beat morphologies with CUR

In this paper, we use the CUR matrix factorization as a means of dimension reduction to identify important subsequences in electrocardiogram (ECG) time series. As opposed to other factorizations typically used in dimension reduction that characterize data in terms of abstract representatives (for example, an orthogonal basis), the CUR factorization describes the data in terms of actual instances within the original data set. Therefore, the CUR characterization can be directly related back to the clinical setting. We apply CUR to a synthetic ECG data set as well as to data from the MIT-BIH Arrhythmia, MGH-MF, and Incart databases using the discrete empirical interpolation method (DEIM) and an incremental QR factorization. In doing so, we demonstrate that CUR is able to identify beat morphologies that are representative of the data set, including rare-occurring beat events, providing a robust summarization of the ECG data. We also see that using CUR-selected beats to label the remaining unselected beats via 1-nearest neighbor classification produces results comparable to those presented in other works. While the electrocardiogram is of particular interest here, this work demonstrates the utility of CUR in detecting representative subsequences in quasiperiodic physiological time series.

Landmark lecture on cardiac intensive care and anaesthesia: continuum and conundrums

Cardiac anesthesia and critical care provide an important continuum of care for patients with congenital heart disease. Clinicians in both areas work in complex environments in which the interactions between humans and technology is critical. Understanding our contributions to outcomes (modifiable risk) and our ability to perceive and predict an evolving clinical state (low failure-to-predict rate) are important performance metrics. Improved methods for capturing continuous physiologic signals will allow for new and interactive approaches to data visualization, and for sophisticated and iterative data modeling that will help define a patient's phenotype and response to treatment (precision physiology).