Pathophysiological Responses to Bloodstream Infection in Critically Ill Transplant Recipients Compared With Non-Transplant Recipients.

BACKGROUND: Identification of bloodstream infection (BSI) in transplant recipients may be difficult due to immunosuppression. Accordingly, we aimed to compare responses to BSI in critically ill transplant and non-transplant recipients and to modify systemic inflammatory response syndrome (SIRS) criteria for transplant recipients. METHODS: We analyzed univariate risks and developed multivariable models of BSI with 27 clinical variables from adult intensive care unit (ICU) patients at the University of Virginia (UVA) and at the University of Pittsburgh (Pitt). We used Bayesian inference to adjust SIRS criteria for transplant recipients. RESULTS: We analyzed 38.7 million hourly measurements from 41 725 patients at UVA, including 1897 transplant recipients with 193 episodes of BSI and 53 608 patients at Pitt, including 1614 transplant recipients with 768 episodes of BSI. The univariate responses to BSI were comparable in transplant and non-transplant recipients. The area under the receiver operating characteristic curve (AUC) was 0.82 (95% confidence interval [CI], .80-.83) for the model using all UVA patient data and 0.80 (95% CI, .76-.83) when using only transplant recipient data. The UVA all-patient model had an AUC of 0.77 (95% CI, .76-.79) in non-transplant recipients and 0.75 (95% CI, .71-.79) in transplant recipients at Pitt. The relative importance of the 27 predictors was similar in transplant and non-transplant models. An upper temperature of 37.5°C in SIRS criteria improved reclassification performance in transplant recipients. CONCLUSIONS: Critically ill transplant and non-transplant recipients had similar responses to BSI. An upper temperature of 37.5°C in SIRS criteria improved BSI screening in transplant recipients.

Use of artificial intelligence in critical care: opportunities and obstacles.

BACKGROUND: Perhaps nowhere else in the healthcare system than in the intensive care unit environment are the challenges to create useful models with direct time-critical clinical applications more relevant and the obstacles to achieving those goals more massive. Machine learning-based artificial intelligence (AI) techniques to define states and predict future events are commonplace activities of modern life. However, their penetration into acute care medicine has been slow, stuttering and uneven. Major obstacles to widespread effective application of AI approaches to the real-time care of the critically ill patient exist and need to be addressed. MAIN BODY: Clinical decision support systems (CDSSs) in acute and critical care environments support clinicians, not replace them at the bedside. As will be discussed in this review, the reasons are many and include the immaturity of AI-based systems to have situational awareness, the fundamental bias in many large databases that do not reflect the target population of patient being treated making fairness an important issue to address and technical barriers to the timely access to valid data and its display in a fashion useful for clinical workflow. The inherent "black-box" nature of many predictive algorithms and CDSS makes trustworthiness and acceptance by the medical community difficult. Logistically, collating and curating in real-time multidimensional data streams of various sources needed to inform the algorithms and ultimately display relevant clinical decisions support format that adapt to individual patient responses and signatures represent the efferent limb of these systems and is often ignored during initial validation efforts. Similarly, legal and commercial barriers to the access to many existing clinical databases limit studies to address fairness and generalizability of predictive models and management tools. CONCLUSIONS: AI-based CDSS are evolving and are here to stay. It is our obligation to be good shepherds of their use and further development.

Forecasting algorithms in the ICU.

Despite significant advances in modeling methods and access to large datasets, there are very few real-time forecasting systems deployed in highly monitored environment such as the intensive care unit. Forecasting models may be developed as classification, regression or time-to-event tasks; each could be using a variety of machine learning algorithms. An accurate and useful forecasting systems include several components beyond a forecasting model, and its performance is assessed using end-user-centered metrics. Several barriers to implementation and acceptance persist and clinicians will play an active role in the successful deployment of this promising technology.

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.

Continuous ECG monitoring should be the heart of bedside AI-based predictive analytics monitoring for early detection of clinical deterioration.

The idea that we can detect subacute potentially catastrophic illness earlier by using statistical models trained on clinical data is now well-established. We review evidence that supports the role of continuous cardiorespiratory monitoring in these predictive analytics monitoring tools. In particular, we review how continuous ECG monitoring reflects the patient and not the clinician, is less likely to be biased, is unaffected by changes in practice patterns, captures signatures of illnesses that are interpretable by clinicians, and is an underappreciated and underutilized source of detailed information for new mathematical methods to reveal.

A data-driven model of the role of energy in sepsis.

Exposure to pathogens elicits a complex immune response involving multiple interdependent pathways. This response may mitigate detrimental effects and restore health but, if imbalanced, can lead to negative outcomes including sepsis. This complexity and need for balance pose a challenge for clinicians and have attracted attention from modelers seeking to apply computational tools to guide therapeutic approaches. In this work, we address a shortcoming of such past efforts by incorporating the dynamics of energy production and consumption into a computational model of the acute immune response. With this addition, we performed fits of model dynamics to data obtained from non-human primates exposed to Escherichia coli. Our analysis identifies parameters that may be crucial in determining survival outcomes and also highlights energy-related factors that modulate the immune response across baseline and altered glucose conditions.

Artificial Intelligence in Critical Care Medicine.

This article is one of ten reviews selected from the Annual Update in Intensive Care and Emergency Medicine 2022. Other selected articles can be found online at https://www.biomedcentral.com/collections/annualupdate2022 . Further information about the Annual Update in Intensive Care and Emergency Medicine is available from https://link.springer.com/bookseries/8901 .

Association between Net Ultrafiltration Rate and Renal Recovery among Critically Ill Adults with Acute Kidney Injury Receiving Continuous Renal Replacement Therapy: An Observational Cohort Study.

INTRODUCTION: Higher net ultrafiltration (UFNET) rates are associated with mortality among critically ill patients with acute kidney injury (AKI) and treated with continuous renal replacement therapy (CRRT). OBJECTIVE: The aim of the study was to discover whether UFNET rates are associated with renal recovery and independence from renal replacement therapy (RRT). METHODS: Retrospective cohort study using data from the Randomized Evaluation of Normal versus Augmented Level of Renal Replacement Therapy trial that enrolled 1,433 critically ill patients with AKI and treated with CRRT between December 2005 and November 2008 across 35 intensive care units in Australia and New Zealand. We examined the association between UFNET rate and time to independence from RRT by day 90 using competing risk regression after accounting for mortality. The UFNET rate was defined as the volume of fluid removed per hour adjusted for patient body weight. RESULTS AND CONCLUSIONS: Median age was 67.3 (interquartile range [IQR], 57-76.3) years, 64.4% were male, median Acute Physiology and Chronic Health Evaluation-III score was 100 (IQR, 84-118), and 634 (44.2%) died by day 90. Kidney recovery occurred in 755 patients (52.7%). Using tertiles of UFNET rates, 3 groups were defined: high, >1.75; middle, 1.01-1.75; and low, <1.01 mL/kg/h. Proportion of patients alive and independent of RRT among the groups were 47.8 versus 57.2 versus 53.0%; p = 0.01. Using competing risk regression, higher UFNET rate tertile compared with middle (cause-specific hazard ratio [csHR], 0.79, 95% CI, 0.66-0.95; subdistribution hazard ratio [sHR], 0.80, 95% CI, 0.67-0.97) and lower (csHR, 0.69, 95% CI, 0.56-0.85; sHR, 0.78, 95% CI 0.64-0.95) tertiles were associated with a longer time to independence from RRT. Every 1.0 mL/kg/h increase in rate was associated with a lower probability of kidney recovery (csHR, 0.81, 95% CI, 0.74-0.89; and sHR, 0.87, 95% CI, 0.80-0.95). Using the joint model, longitudinal increases in UFNET rates were also associated with a lower renal recovery (ß = -0.29, p < 0.001). UFNET rates >1.75 mL/kg/h compared with rates 1.01-1.75 and <1.01 mL/kg/h were associated with a longer duration of dependence on RRT. Randomized clinical trials are required to confirm this UFNET rate-outcome relationship.

A Research Agenda for Precision Medicine in Sepsis and Acute Respiratory Distress Syndrome: An Official American Thoracic Society Research Statement.

Background: Precision medicine focuses on the identification of therapeutic strategies that are effective for a group of patients based on similar unifying characteristics. The recent success of precision medicine in non-critical care settings has resulted from the confluence of large clinical and biospecimen repositories, innovative bioinformatics, and novel trial designs. Similar advances for precision medicine in sepsis and in the acute respiratory distress syndrome (ARDS) are possible but will require further investigation and significant investment in infrastructure. Methods: This project was funded by the American Thoracic Society Board of Directors. A multidisciplinary and diverse working group reviewed the available literature, established a conceptual framework, and iteratively developed recommendations for the Precision Medicine Research Agenda for Sepsis and ARDS. Results: The following six priority recommendations were developed by the working group: 1) the creation of large richly phenotyped and harmonized knowledge networks of clinical, imaging, and multianalyte molecular data for sepsis and ARDS; 2) the implementation of novel trial designs, including adaptive designs, and embedding trial procedures in the electronic health record; 3) continued innovation in the data science and engineering methods required to identify heterogeneity of treatment effect; 4) further development of the tools necessary for the real-time application of precision medicine approaches; 5) work to ensure that precision medicine strategies are applicable and available to a broad range of patients varying across differing racial, ethnic, socioeconomic, and demographic groups; and 6) the securement and maintenance of adequate and sustainable funding for precision medicine efforts. Conclusions: Precision medicine approaches that incorporate variability in genomic, biologic, and environmental factors may provide a path forward for better individualizing the delivery of therapies and improving care for patients with sepsis and ARDS.

Neurocritical Care Performance Measures Derived from Electronic Health Record Data are Feasible and Reveal Site-Specific Variation: A CHoRUS Pilot Project.

BACKGROUND: We evaluated the feasibility and discriminability of recently proposed Clinical Performance Measures for Neurocritical Care (Neurocritical Care Society) and Quality Indicators for Traumatic Brain Injury (Collaborative European NeuroTrauma Effectiveness Research in TBI; CENTER-TBI) extracted from electronic health record (EHR) flowsheet data. METHODS: At three centers within the Collaborative Hospital Repository Uniting Standards (CHoRUS) for Equitable AI consortium, we examined consecutive neurocritical care admissions exceeding 24 h (03/2015-02/2020) and evaluated the feasibility, discriminability, and site-specific variation of five clinical performance measures and quality indicators: (1) intracranial pressure (ICP) monitoring (ICPM) within 24 h when indicated, (2) ICPM latency when initiated within 24 h, (3) frequency of nurse-documented neurologic assessments, (4) intermittent pneumatic compression device (IPCd) initiation within 24 h, and (5) latency to IPCd application. We additionally explored associations between delayed IPCd initiation and codes for venous thromboembolism documented using the 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) system. Median (interquartile range) statistics are reported. Kruskal-Wallis tests were measured for differences across centers, and Dunn statistics were reported for between-center differences. RESULTS: A total of 14,985 admissions met inclusion criteria. ICPM was documented in 1514 (10.1%), neurologic assessments in 14,635 (91.1%), and IPCd application in 14,175 (88.5%). ICPM began within 24 h for 1267 (83.7%), with site-specific latency differences among sites 1-3, respectively, (0.54 h [2.82], 0.58 h [1.68], and 2.36 h [4.60]; p < 0.001). The frequency of nurse-documented neurologic assessments also varied by site (17.4 per day [5.97], 8.4 per day [3.12], and 15.3 per day [8.34]; p < 0.001) and diurnally (6.90 per day during daytime hours vs. 5.67 per day at night, p < 0.001). IPCds were applied within 24 h for 12,863 (90.7%) patients meeting clinical eligibility (excluding those with EHR documentation of limiting injuries, actively documented as ambulating, or refusing prophylaxis). In-hospital venous thromboembolism varied by site (1.23%, 1.55%, and 5.18%; p < 0.001) and was associated with increased IPCd latency (overall, 1.02 h [10.4] vs. 0.97 h [5.98], p = 0.479; site 1, 2.25 h [10.27] vs. 1.82 h [7.39], p = 0.713; site 2, 1.38 h [5.90] vs. 0.80 h [0.53], p = 0.216; site 3, 0.40 h [16.3] vs. 0.35 h [11.5], p = 0.036). CONCLUSIONS: Electronic health record-derived reporting of neurocritical care performance measures is feasible and demonstrates site-specific variation. Future efforts should examine whether performance or documentation drives these measures, what outcomes are associated with performance, and whether EHR-derived measures of performance measures and quality indicators are modifiable.

Intelligent Clinical Decision Support.

Early recognition of pathologic cardiorespiratory stress and forecasting cardiorespiratory decompensation in the critically ill is difficult even in highly monitored patients in the Intensive Care Unit (ICU). Instability can be intuitively defined as the overt manifestation of the failure of the host to adequately respond to cardiorespiratory stress. The enormous volume of patient data available in ICU environments, both of high-frequency numeric and waveform data accessible from bedside monitors, plus Electronic Health Record (EHR) data, presents a platform ripe for Artificial Intelligence (AI) approaches for the detection and forecasting of instability, and data-driven intelligent clinical decision support (CDS). Building unbiased, reliable, and usable AI-based systems across health care sites is rapidly becoming a high priority, specifically as these systems relate to diagnostics, forecasting, and bedside clinical decision support. The ICU environment is particularly well-positioned to demonstrate the value of AI in saving lives. The goal is to create AI models embedded in a real-time CDS for forecasting and mitigation of critical instability in ICU patients of sufficient readiness to be deployed at the bedside. Such a system must leverage multi-source patient data, machine learning, systems engineering, and human action expertise, the latter being key to successful CDS implementation in the clinical workflow and evaluation of bias. We present one approach to create an operationally relevant AI-based forecasting CDS system.

Accuracy of identifying hospital acquired venous thromboembolism by administrative coding: implications for big data and machine learning research.

Big data analytics research using heterogeneous electronic health record (EHR) data requires accurate identification of disease phenotype cases and controls. Overreliance on ground truth determination based on administrative data can lead to biased and inaccurate findings. Hospital-acquired venous thromboembolism (HA-VTE) is challenging to identify due to its temporal evolution and variable EHR documentation. To establish ground truth for machine learning modeling, we compared accuracy of HA-VTE diagnoses made by administrative coding to manual review of gold standard diagnostic test results. We performed retrospective analysis of EHR data on 3680 adult stepdown unit patients identifying HA-VTE. International Classification of Diseases, Ninth Revision (ICD-9-CM) codes for VTE were identified. 4544 radiology reports associated with VTE diagnostic tests were screened using terminology extraction and then manually reviewed by a clinical expert to confirm diagnosis. Of 415 cases with ICD-9-CM codes for VTE, 219 were identified with acute onset type codes. Test report review identified 158 new-onset HA-VTE cases. Only 40% of ICD-9-CM coded cases (n = 87) were confirmed by a positive diagnostic test report, leaving the majority of administratively coded cases unsubstantiated by confirmatory diagnostic test. Additionally, 45% of diagnostic test confirmed HA-VTE cases lacked corresponding ICD codes. ICD-9-CM coding missed diagnostic test-confirmed HA-VTE cases and inaccurately assigned cases without confirmed VTE, suggesting dependence on administrative coding leads to inaccurate HA-VTE phenotyping. Alternative methods to develop more sensitive and specific VTE phenotype solutions portable across EHR vendor data are needed to support case-finding in big-data analytics.

Effects of 5% Albumin Plus Saline Versus Saline Alone on Outcomes From Large-Volume Resuscitation in Critically Ill Patients.

OBJECTIVES: To compare 5% albumin with 0.9% saline for large-volume resuscitation (> 60 mL/Kg within 24 hr), on mortality and development of acute kidney injury. DESIGN: Retrospective cohort study. SETTING: Patients admitted to ICUs in 13 hospitals across Western Pennsylvania. We analyzed two independent cohorts, the High-Density Intensive Care databases: High-Density Intensive Care-08 (July 2000 to October 2008, H08) and High-Density Intensive Care-15 (October 2008 to December 2014, H15). PATIENTS: Total of 18,629 critically ill patients requiring large-volume resuscitation. INTERVENTIONS: Five percent of albumin in addition to saline versus 0.9% saline. MEASUREMENTS AND MAIN RESULTS: After excluding patients with acute kidney injury prior to large-volume resuscitation, 673 of 2,428 patients (27.7%) and 1,814 of 16,201 patients (11.2%) received 5% albumin in H08 and H15, respectively. Use of 5% albumin was associated with decreased 30-day mortality by multivariate regression in H08 (odds ratio 0.65; 95% CI 0.49-0.85; p = 0.002) and in H15 (0.52; 95% CI 0.44-0.62; p < 0.0001) but was associated with increased acute kidney injury in H08 (odds ratio 1.98; 95% CI 1.56-2.51; p < 0.001) and in H15 (odds ratio 1.75; 95% CI 1.58-1.95; p < 0.001). However, 5% albumin was not associated with persistent acute kidney injury and resulted in decreased major adverse kidney event at 30, 90, and 365 days. Propensity matched analysis confirmed similar associations with mortality and acute kidney injury. CONCLUSIONS: During large-volume resuscitation, 5% albumin was associated with reduced mortality and major adverse kidney event at 30, 90, and 365 days. However, a higher rate of acute kidney injury of any stage was observed that did not translate into persistent renal dysfunction.

Sharing ICU Patient Data Responsibly Under the Society of Critical Care Medicine/European Society of Intensive Care Medicine Joint Data Science Collaboration: The Amsterdam University Medical Centers Database (AmsterdamUMCdb) Example.

OBJECTIVES: Critical care medicine is a natural environment for machine learning approaches to improve outcomes for critically ill patients as admissions to ICUs generate vast amounts of data. However, technical, legal, ethical, and privacy concerns have so far limited the critical care medicine community from making these data readily available. The Society of Critical Care Medicine and the European Society of Intensive Care Medicine have identified ICU patient data sharing as one of the priorities under their Joint Data Science Collaboration. To encourage ICUs worldwide to share their patient data responsibly, we now describe the development and release of Amsterdam University Medical Centers Database (AmsterdamUMCdb), the first freely available critical care database in full compliance with privacy laws from both the United States and Europe, as an example of the feasibility of sharing complex critical care data. SETTING: University hospital ICU. SUBJECTS: Data from ICU patients admitted between 2003 and 2016. INTERVENTIONS: We used a risk-based deidentification strategy to maintain data utility while preserving privacy. In addition, we implemented contractual and governance processes, and a communication strategy. Patient organizations, supporting hospitals, and experts on ethics and privacy audited these processes and the database. MEASUREMENTS AND MAIN RESULTS: AmsterdamUMCdb contains approximately 1 billion clinical data points from 23,106 admissions of 20,109 patients. The privacy audit concluded that reidentification is not reasonably likely, and AmsterdamUMCdb can therefore be considered as anonymous information, both in the context of the U.S. Health Insurance Portability and Accountability Act and the European General Data Protection Regulation. The ethics audit concluded that responsible data sharing imposes minimal burden, whereas the potential benefit is tremendous. CONCLUSIONS: Technical, legal, ethical, and privacy challenges related to responsible data sharing can be addressed using a multidisciplinary approach. A risk-based deidentification strategy, that complies with both U.S. and European privacy regulations, should be the preferred approach to releasing ICU patient data. This supports the shared Society of Critical Care Medicine and European Society of Intensive Care Medicine vision to improve critical care outcomes through scientific inquiry of vast and combined ICU datasets.

Improvements and limitations in developing multivariate models of hemorrhage and transfusion risk for the obstetric population.

BACKGROUND: Maternal hemorrhage protocols involve risk screening. These protocols prepare clinicians for potential hemorrhage and transfusion in individual patients. Patient-specific estimation and stratification of risk may improve maternal outcomes. STUDY DESIGN AND METHODS: Prediction models for hemorrhage and transfusion were trained and tested in a data set of 74 variables from 63 973 deliveries (97.6% of the source population of 65 560 deliveries included in a perinatal database from an academic urban delivery center) with sufficient data at pertinent time points: antepartum, peripartum, and postpartum. Hemorrhage and transfusion were present in 6% and 1.6% of deliveries, respectively. Model performance was evaluated with the receiver operating characteristic (ROC), precision-recall curves, and the Hosmer-Lemeshow calibration statistic. RESULTS: For hemorrhage risk prediction, logistic regression model discrimination showed ROCs of 0.633, 0.643, and 0.661 for the antepartum, peripartum, and postpartum models, respectively. These improve upon the California Maternal Quality Care Collaborative (CMQCC) accuracy of 0.613 for hemorrhage. Predictions of transfusion resulted in ROCs of 0.806, 0.822, and 0.854 for the antepartum, peripartum, and postpartum models, respectively. Previously described and new risk factors were identified. Models were not well calibrated with Hosmer-Lemeshow statistic P values between .001 and .6. CONCLUSIONS: Our models improve on existing risk assessment; however, further enhancement might require the inclusion of more granular, dynamic data. With the goal of increasing translatability, this work was distilled to an online open-source repository, including a form allowing risk factor inputs and outputs of CMQCC risk, alongside our numerical risk estimation and stratification of hemorrhage and transfusion.

Assessment of Dynamic Intracranial Compliance in Children with Severe Traumatic Brain Injury: Proof-of-Concept.

BACKGROUND AND AIMS: Intracranial compliance refers to the relationship between a change in intracranial volume and the resultant change in intracranial pressure (ICP). Measurement of compliance is useful in managing cardiovascular and respiratory failure; however, there are no contemporary means to assess intracranial compliance. Knowledge of intracranial compliance could complement ICP and cerebral perfusion pressure (CPP) monitoring in patients with severe traumatic brain injury (TBI) and may enable a proactive approach to ICP management. In this proof-of-concept study, we aimed to capitalize on the physiologic principles of intracranial compliance and vascular reactivity to CO2, and standard-of-care neurocritical care monitoring, to develop a method to assess dynamic intracranial compliance. METHODS: Continuous ICP and end-tidal CO2 (ETCO2) data from children with severe TBI were collected after obtaining informed consent in this Institutional Review Board-approved study. An intracranial pressure-PCO2 Compliance Index (PCI) was derived by calculating the moment-to-moment correlation between change in ICP and change in ETCO2. As such, "good" compliance may be reflected by a lack of correlation between time-synched changes in ICP in response to changes in ETCO2, and "poor" compliance may be reflected by a positive correlation between changes in ICP in response to changes in ETCO2. RESULTS: A total of 978 h of ICP and ETCO2 data were collected and analyzed from eight patients with severe TBI. Demographic and clinical characteristics included patient age 7.1 ± 5.8 years (mean ± SD); 6/8 male; initial Glasgow Coma Scale score 3 [3-7] (median [IQR]); 6/8 had decompressive surgery; 7.1 ± 1.4 ICP monitor days; ICU length of stay (LOS) 16.1 ± 6.8 days; hospital LOS 25.9 ± 8.4 days; and survival 100%. The mean PCI for all patients throughout the monitoring period was 0.18 ± 0.04, where mean ICP was 13.7 ± 2.1 mmHg. In this cohort, PCI was observed to be consistently above 0.18 by 12 h after monitor placement. Percent time spent with PCI thresholds > 0.1, 0.2, and 0.3 were 62% [24], 38% [14], and 23% [15], respectively. The percentage of time spent with an ICP threshold > 20 mmHg was 5.1% [14.6]. CONCLUSIONS: Indirect assessment of dynamic intracranial compliance in TBI patients using standard-of-care monitoring appears feasible and suggests a prolonged period of derangement out to 5 days post-injury. Further study is ongoing to determine if the PCI-a new physiologic index, complements utility of ICP and/or CPP in guiding management of patients with severe TBI.

Correction: From Inverse Problems in Mathematical Physiology to Quantitative Differential Diagnoses.

[This corrects the article DOI: 10.1371/journal.pcbi.0030204.].

Prediction of hypotension events with physiologic vital sign signatures in the intensive care unit.

BACKGROUND: Even brief hypotension is associated with increased morbidity and mortality. We developed a machine learning model to predict the initial hypotension event among intensive care unit (ICU) patients and designed an alert system for bedside implementation. MATERIALS AND METHODS: From the Medical Information Mart for Intensive Care III (MIMIC-3) dataset, minute-by-minute vital signs were extracted. A hypotension event was defined as at least five measurements within a 10-min period of systolic blood pressure ≤ 90 mmHg and mean arterial pressure ≤ 60 mmHg. Using time series data from 30-min overlapping time windows, a random forest (RF) classifier was used to predict risk of hypotension every minute. Chronologically, the first half of extracted data was used to train the model, and the second half was used to validate the trained model. The model's performance was measured with area under the receiver operating characteristic curve (AUROC) and area under the precision recall curve (AUPRC). Hypotension alerts were generated using risk score time series, a stacked RF model. A lockout time were applied for real-life implementation. RESULTS: We identified 1307 subjects (1580 ICU stays) as the hypotension group and 1619 subjects (2279 ICU stays) as the non-hypotension group. The RF model showed AUROC of 0.93 and 0.88 at 15 and 60 min, respectively, before hypotension, and AUPRC of 0.77 at 60 min before. Risk score trajectories revealed 80% and > 60% of hypotension predicted at 15 and 60 min before the hypotension, respectively. The stacked model with 15-min lockout produced on average 0.79 alerts/subject/hour (sensitivity 92.4%). CONCLUSION: Clinically significant hypotension events in the ICU can be predicted at least 1 h before the initial hypotension episode. With a highly sensitive and reliable practical alert system, a vast majority of future hypotension could be captured, suggesting potential real-life utility.

Parsimony of Hemodynamic Monitoring Data Sufficient for the Detection of Hemorrhage.

BACKGROUND: Individualized hemodynamic monitoring approaches are not well validated. Thus, we evaluated the discriminative performance improvement that might occur when moving from noninvasive monitoring (NIM) to invasive monitoring and with increasing levels of featurization associated with increasing sampling frequency and referencing to a stable baseline to identify bleeding during surgery in a porcine model. METHODS: We collected physiologic waveform (WF) data (250 Hz) from NIM, central venous (CVC), arterial (ART), and pulmonary arterial (PAC) catheters, plus mixed venous O2 saturation and cardiac output from 38 anesthetized Yorkshire pigs bled at 20 mL/min until a mean arterial pressure of 30 mm Hg following a 30-minute baseline period. Prebleed physiologic data defined a personal stable baseline for each subject independently. Nested models were evaluated using simple hemodynamic metrics (SM) averaged over 20-second windows and sampled every minute, beat to beat (B2B), and WF using Random Forest Classification models to identify bleeding with or without normalization to personal stable baseline, using a leave-one-pig-out cross-validation to minimize model overfitting. Model hyperparameters were tuned to detect stable or bleeding states. Bleeding models were compared use both each subject's personal baseline and a grouped-average (universal) baseline. Timeliness of bleed onset detection was evaluated by comparing the tradeoff between a low false-positive rate (FPR) and shortest time to bleed detection. Predictive performance was evaluated using a variant of the receiver operating characteristic focusing on minimizing FPR and false-negative rates (FNR) for true-positive and true-negative rates, respectively. RESULTS: In general, referencing models to a personal baseline resulted in better bleed detection performance for all catheters than using universal baselined data. Increasing granularity from SM to B2B and WF progressively improved bleeding detection. All invasive monitoring outperformed NIM for both time to bleeding detection and low FPR and FNR. In that regard, when referenced to personal baseline with SM analysis, PAC and ART + PAC performed best; for B2B CVC, PAC and ART + PAC performed best; and for WF PAC, CVC, ART + CVC, and ART + PAC performed equally well and better than other monitoring approaches. Without personal baseline, NIM performed poorly at all levels, while all catheters performed similarly for SM, with B2B PAC and ART + PAC performing the best, and for WF PAC, ART, ART + CVC, and ART + PAC performed equally well and better than the other monitoring approaches. CONCLUSIONS: Increasing hemodynamic monitoring featurization by increasing sampling frequency and referencing to personal baseline markedly improves the ability of invasive monitoring to detect bleed.

Leveraging Eye Tracking to Prioritize Relevant Medical Record Data: Comparative Machine Learning Study.

BACKGROUND: Electronic medical record (EMR) systems capture large amounts of data per patient and present that data to physicians with little prioritization. Without prioritization, physicians must mentally identify and collate relevant data, an activity that can lead to cognitive overload. To mitigate cognitive overload, a Learning EMR (LEMR) system prioritizes the display of relevant medical record data. Relevant data are those that are pertinent to a context-defined as the combination of the user, clinical task, and patient case. To determine which data are relevant in a specific context, a LEMR system uses supervised machine learning models of physician information-seeking behavior. Since obtaining information-seeking behavior data via manual annotation is slow and expensive, automatic methods for capturing such data are needed. OBJECTIVE: The goal of the research was to propose and evaluate eye tracking as a high-throughput method to automatically acquire physician information-seeking behavior useful for training models for a LEMR system. METHODS: Critical care medicine physicians reviewed intensive care unit patient cases in an EMR interface developed for the study. Participants manually identified patient data that were relevant in the context of a clinical task: preparing a patient summary to present at morning rounds. We used eye tracking to capture each physician's gaze dwell time on each data item (eg, blood glucose measurements). Manual annotations and gaze dwell times were used to define target variables for developing supervised machine learning models of physician information-seeking behavior. We compared the performance of manual selection and gaze-derived models on an independent set of patient cases. RESULTS: A total of 68 pairs of manual selection and gaze-derived machine learning models were developed from training data and evaluated on an independent evaluation data set. A paired Wilcoxon signed-rank test showed similar performance of manual selection and gaze-derived models on area under the receiver operating characteristic curve (P=.40). CONCLUSIONS: We used eye tracking to automatically capture physician information-seeking behavior and used it to train models for a LEMR system. The models that were trained using eye tracking performed like models that were trained using manual annotations. These results support further development of eye tracking as a high-throughput method for training clinical decision support systems that prioritize the display of relevant medical record data.