Potential Impact of Subphenotyping in Pharmacologic Management of Acute Respiratory Distress Syndrome.

Background: Acute respiratory distress syndrome (ARDS) is an acute inflammatory process in the lungs associated with high morbidity and mortality. Previous research has studied both nonpharmacologic and pharmacologic interventions aimed at targeting this inflammatory process and improving ventilation. Hypothesis: To date, only nonpharmacologic interventions including lung protective ventilation, prone positioning, and high positive end-expiratory pressure ventilation strategies have resulted in significant improvements in patient outcomes. Given the high mortality associated with ARDS despite these advancements, interest in subphenotyping has grown, aiming to improve diagnosis and develop personalized treatment approaches. Data Collection: Previous trials evaluating pharmacologic therapies in heterogeneous populations have primarily demonstrated no positive effect, but hope to show benefit when targeting specific subphenotypes, thus increasing their efficacy, while simultaneously decreasing adverse effects. Results: Although most studies evaluating pharmacologic therapies for ARDS have not demonstrated a mortality benefit, there is limited data evaluating pharmacologic therapies in ARDS subphenotypes, which have found promising results. Neuromuscular blocking agents, corticosteroids, and simvastatin have resulted in a mortality benefit when used in patients with the hyper-inflammatory ARDS subphenotype. Therapeutic Opinion: The use of subphenotyping could revolutionize the way ARDS therapies are applied and therefore improve outcomes while also limiting the adverse effects associated with their ineffective use. Future studies should evaluate ARDS subphenotypes and their response to pharmacologic intervention to advance this area of precision medicine.

MIMIC-IV, a freely accessible electronic health record dataset.

Digital data collection during routine clinical practice is now ubiquitous within hospitals. The data contains valuable information on the care of patients and their response to treatments, offering exciting opportunities for research. Typically, data are stored within archival systems that are not intended to support research. These systems are often inaccessible to researchers and structured for optimal storage, rather than interpretability and analysis. Here we present MIMIC-IV, a publicly available database sourced from the electronic health record of the Beth Israel Deaconess Medical Center. Information available includes patient measurements, orders, diagnoses, procedures, treatments, and deidentified free-text clinical notes. MIMIC-IV is intended to support a wide array of research studies and educational material, helping to reduce barriers to conducting clinical research.

BRAX, Brazilian labeled chest x-ray dataset.

Chest radiographs allow for the meticulous examination of a patient's chest but demands specialized training for proper interpretation. Automated analysis of medical imaging has become increasingly accessible with the advent of machine learning (ML) algorithms. Large labeled datasets are key elements for training and validation of these ML solutions. In this paper we describe the Brazilian labeled chest x-ray dataset, BRAX: an automatically labeled dataset designed to assist researchers in the validation of ML models. The dataset contains 24,959 chest radiography studies from patients presenting to a large general Brazilian hospital. A total of 40,967 images are available in the BRAX dataset. All images have been verified by trained radiologists and de-identified to protect patient privacy. Fourteen labels were derived from free-text radiology reports written in Brazilian Portuguese using Natural Language Processing.

Identification of acute respiratory distress syndrome subphenotypes de novo using routine clinical data: a retrospective analysis of ARDS clinical trials.

OBJECTIVES: The acute respiratory distress syndrome (ARDS) is a heterogeneous condition, and identification of subphenotypes may help in better risk stratification. Our study objective is to identify ARDS subphenotypes using new simpler methodology and readily available clinical variables. SETTING: This is a retrospective Cohort Study of ARDS trials. Data from the US ARDSNet trials and from the international ART trial. PARTICIPANTS: 3763 patients from ARDSNet data sets and 1010 patients from the ART data set. PRIMARY AND SECONDARY OUTCOME MEASURES: The primary outcome was 60-day or 28-day mortality, depending on what was reported in the original trial. K-means cluster analysis was performed to identify subgroups. Sets of candidate variables were tested to assess their ability to produce different probabilities for mortality in each cluster. Clusters were compared with biomarker data, allowing identification of subphenotypes. RESULTS: Data from 4773 patients were analysed. Two subphenotypes (A and B) resulted in optimal separation in the final model, which included nine routinely collected clinical variables, namely heart rate, mean arterial pressure, respiratory rate, bilirubin, bicarbonate, creatinine, PaO2, arterial pH and FiO2. Participants in subphenotype B showed increased levels of proinflammatory markers, had consistently higher mortality, lower number of ventilator-free days at day 28 and longer duration of ventilation compared with patients in the subphenotype A. CONCLUSIONS: Routinely available clinical data can successfully identify two distinct subphenotypes in adult ARDS patients. This work may facilitate implementation of precision therapy in ARDS clinical trials.

Gene Expression Scoring of Immune Activity Levels for Precision Use of Hydrocortisone in Vasodilatory Shock.

PURPOSE: Among patients with vasodilatory shock, gene expression scores may identify different immune states. We aimed to test whether such scores are robust in identifying patients' immune state and predicting response to hydrocortisone treatment in vasodilatory shock. MATERIALS AND METHODS: We selected genes to generate continuous scores to define previously established subclasses of sepsis. We used these scores to identify a patient's immune state. We evaluated the potential for these states to assess the differential effect of hydrocortisone in two randomized clinical trials of hydrocortisone versus placebo in vasodilatory shock. RESULTS: We initially identified genes associated with immune-adaptive, immune-innate, immune-coagulant functions. From these genes, 15 were most relevant to generate expression scores related to each of the functions. These scores were used to identify patients as immune-adaptive prevalent (IA-P) and immune-innate prevalent (IN-P). In IA-P patients, hydrocortisone therapy increased 28-day mortality in both trials (43.3% vs 14.7%, P = 0.028) and (57.1% vs 0.0%, P = 0.99). In IN-P patients, this effect was numerically reversed. CONCLUSIONS: Gene expression scores identified the immune state of vasodilatory shock patients, one of which (IA-P) identified those who may be harmed by hydrocortisone. Gene expression scores may help advance the field of personalized medicine.

Effect of spontaneous breathing on ventilator-free days in critically ill patients-an analysis of patients in a large observational cohort.

BACKGROUND: Mechanical ventilation can injure lung tissue and respiratory muscles. The aim of the present study is to assess the effect of the amount of spontaneous breathing during mechanical ventilation on patient outcomes. METHODS: This is an analysis of the database of the 'Medical Information Mart for Intensive Care (MIMIC)'-III, considering intensive care units (ICUs) of the Beth Israel Deaconess Medical Center (BIDMC), Boston, MA. Adult patients who received invasive ventilation for at least 48 hours were included. Patients were categorized according to the amount of spontaneous breathing, i.e., ≥50% ('high spontaneous breathing') and <50% ('low spontaneous breathing') of time during first 48 hours of ventilation. The primary outcome was the number of ventilator-free days. RESULTS: In total, the analysis included 3,380 patients; 70.2% were classified as 'high spontaneous breathing', and 29.8% as 'low spontaneous breathing'. Patients in the 'high spontaneous breathing' group were older, had more comorbidities, and lower severity scores. In adjusted analysis, the amount of spontaneous breathing was not associated with the number of ventilator-free days [20.0 (0.0-24.2) vs. 19.0 (0.0-23.7) in high vs. low; absolute difference, 0.54 (95% CI, -0.10 to 1.19); P=0.101]. However, 'high spontaneous breathing' was associated with shorter duration of ventilation in survivors [6.5 (3.6 to 12.2) vs. 7.6 (4.1 to 13.9); absolute difference, -0.91 (95% CI, -1.80 to -0.02); P=0.046]. CONCLUSIONS: In patients surviving and receiving ventilation for at least 48 hours, the amount of spontaneous breathing during this period was not associated with an increased number of ventilator-free days.

Impact of sex on use of low tidal volume ventilation in invasively ventilated ICU patients-A mediation analysis using two observational cohorts.

BACKGROUND: Studies in patients receiving invasive ventilation show important differences in use of low tidal volume (VT) ventilation (LTVV) between females and males. The aims of this study were to describe temporal changes in VT and to determine what factors drive the sex difference in use of LTVV. METHODS AND FINDINGS: This is a posthoc analysis of 2 large longitudinal projects in 59 ICUs in the United States, the 'Medical information Mart for Intensive Care III' (MIMIC III) and the 'eICU Collaborative Research DataBase'. The proportion of patients under LTVV (median VT < 8 ml/kg PBW), was the primary outcome. Mediation analysis, a method to dissect total effect into direct and indirect effects, was used to understand which factors drive the sex difference. We included 3614 (44%) females and 4593 (56%) males. Median VT declined over the years, but with a persistent difference between females (from median 10.2 (9.1 to 11.4) to 8.2 (7.5 to 9.1) ml/kg PBW) vs. males (from median 9.2 [IQR 8.2 to 10.1] to 7.3 [IQR 6.6 to 8.0] ml/kg PBW) (P < .001). In females versus males, use of LTVV increased from 5 to 50% versus from 12 to 78% (difference, -27% [-29% to -25%]; P < .001). The sex difference was mainly driven by patients' body height and actual body weight (adjusted average causal mediation effect, -30% [-33% to -27%]; P < .001, and 4 [3% to 4%]; P < .001). CONCLUSIONS: While LTVV is increasingly used in females and males, females continue to receive LTVV less often than males. The sex difference is mainly driven by patients' body height and actual body weight, and not necessarily by sex. Use of LTVV in females could improve by paying more attention to a correct calculation of VT, i.e., using the correct body height.

Prediction on critically ill patients: The role of "big data".

Accurate outcome prediction in Intensive Care Units (ICUs) would allow for better treatment planning, risk adjustment of study populations, and overall improvements in patient care. In the past, prognostic models have focused on mortality using simple ordinal severity of illness scores which could be tabulated manually by a human. With the improvements in computing power and proliferation of electronic medical records, entirely new approaches have become possible. Here we review the latest advances in outcome prediction, paying close attention to methods which are widely applicable and provide a high-level overview of the challenges the field currently faces.

Deidentification of free-text medical records using pre-trained bidirectional transformers.

The ability of caregivers and investigators to share patient data is fundamental to many areas of clinical practice and biomedical research. Prior to sharing, it is often necessary to remove identifiers such as names, contact details, and dates in order to protect patient privacy. Deidentification, the process of removing identifiers, is challenging, however. High-quality annotated data for developing models is scarce; many target identifiers are highly heterogenous (for example, there are uncountable variations of patient names); and in practice anything less than perfect sensitivity may be considered a failure. As a result, patient data is often withheld when sharing would be beneficial, and identifiable patient data is often divulged when a deidentified version would suffice. In recent years, advances in machine learning methods have led to rapid performance improvements in natural language processing tasks, in particular with the advent of large-scale pretrained language models. In this paper we develop and evaluate an approach for deidentification of clinical notes based on a bidirectional transformer model. We propose human interpretable evaluation measures and demonstrate state of the art performance against modern baseline models. Finally, we highlight current challenges in deidentification, including the absence of clear annotation guidelines, lack of portability of models, and paucity of training data. Code to develop our model is open source, allowing for broad reuse.

SEVERITAS: An externally validated mortality prediction for critically ill patients in low and middle-income countries.

OBJECTIVE: Severity of illness scores used in critical care for benchmarking, quality assurance and risk stratification have been mainly created in high-income countries. In low and middle-income countries (LMICs), they cannot be widely utilized due to the demand for large amounts of data that may not be available (e.g. laboratory results). We attempt to create a new severity prognostication model using fewer variables that are easier to collect in an LMIC. SETTING: Two intensive care units, one private and one public, from São Paulo, Brazil PATIENTS: An ICU for the first time. INTERVENTIONS: None. MEASUREMENTS AND MAINS RESULTS: The dataset from the private ICU was used as a training set for model development to predict in-hospital mortality. Three different machine learning models were applied to five different blocks of candidate variables. The resulting 15 models were then validated on a separate dataset from the public ICU, and discrimination and calibration compared to identify the best model. The best performing model used logistic regression on a small set of 10 variables: highest respiratory rate, lowest systolic blood pressure, highest body temperature and Glasgow Coma Scale during the first hour of ICU admission; age; prior functional capacity; type of ICU admission; source of ICU admission; and length of hospital stay prior to ICU admission. On the validation dataset, our new score, named SEVERITAS, had an area under the receiver operating curve of 0.84 (0.82 - 0.86) and standardized mortality ratio of 1.00 (0.91-1.08). Moreover, SEVERITAS had similar discrimination compared to SAPS-3 and better discrimination than the simplified TropICS and R-MPM. CONCLUSIONS: Our study proposes a new ICU mortality prediction model using simple logistic regression on a small set of easily collected variables may be better suited than currently available models for use in low and middle-income countries.

An Evaluation of the Influence of Body Mass Index on Severity Scoring.

OBJECTIVES: Although one third or more of critically ill patients in the United States are obese, obesity is not incorporated as a contributing factor in any of the commonly used severity of illness scores. We hypothesize that selected severity of illness scores would perform differently if body mass index categorization was incorporated and that the performance of these score models would improve after consideration of body mass index as an additional model feature. DESIGN: Retrospective cohort analysis from a multicenter ICU database which contains deidentified data for more than 200,000 ICU admissions from 208 distinct ICUs across the United States between 2014 and 2015. SETTING: First ICU admission of patients with documented height and weight. PATIENTS: One-hundred eight-thousand four-hundred two patients from 189 different ICUs across United States were included in the analyses, of whom 4,661 (4%) were classified as underweight, 32,134 (30%) as normal weight, 32,278 (30%) as overweight, 30,259 (28%) as obese, and 9,070 (8%) as morbidly obese. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: To assess the effect of adding body mass index as a risk adjustment element to the Acute Physiology and Chronic Health Evaluation IV and Oxford Acute Severity of Illness scoring systems, we examined the impact of this addition on both discrimination and calibration. We performed three assessments based upon 1) the original scoring systems, 2) a recalibrated version of the systems, and 3) a recalibrated version incorporating body mass index as a covariate. We also performed a subgroup analysis in groups defined using World Health Organization guidelines for obesity. Incorporating body mass index into the models provided a minor improvement in both discrimination and calibration. In a subgroup analysis, model discrimination was higher in groups with higher body mass index, but calibration worsened. CONCLUSIONS: The performance of ICU prognostic models utilizing body mass index category as a scoring element was inconsistent across body mass index categories. Overall, adding body mass index as a risk adjustment variable led only to a minor improvement in scoring system performance.

Hacking Hackathons: Preparing the next generation for the multidisciplinary world of healthcare technology.

OBJECTIVE: Machine learning in healthcare, and innovative healthcare technology in general, require complex interactions within multidisciplinary teams. Healthcare hackathons are being increasingly used as a model for cross-disciplinary collaboration and learning. The aim of this study is to explore high school student learning experiences during a healthcare hackathon. By optimizing their learning experiences, we hope to prepare a future workforce that can bridge technical and health fields and work seamlessly across disciplines. METHODS: A qualitative exploratory study utilizing focus group interviews was conducted. Eight high school students from the hackathon were invited to participate in this study through convenience sampling Participating students (n = 8) were allocated into three focus groups. Semi structured interviews were completed, and transcripts evaluated using inductive thematic analysis. FINDINGS: Through the structured analysis of focus group transcripts three major themes emerged from the data: (1) Collaboration, (2) Transferable knowledge and skills, and (3) Expectations about hackathons. These themes highlight strengths and potential barriers when bringing this multidisciplinary approach to high school students and the healthcare community. CONCLUSION: This study found that students were empowered by the interdisciplinary experience during a hackathon and felt that the knowledge and skills gained could be applied in real world settings. However, addressing student expectations of hackathons prior to the event is an area for improvement. These findings have implications for future hackathons and can spur further research into using the hackathon model as an educational experience for learners of all ages.

Assessing team effectiveness and affective learning in a datathon.

BACKGROUND: Datathons are increasingly organized in the healthcare field. The goal is to assemble people with different backgrounds to work together as a team and engage in clinically relevant research or develop algorithms using health-related datasets. Criteria to assess the return of investment on such events have traditionally included publications produced, patents for prediction, classification, image recognition and other types of software, and start-up companies around the application of machine learning in healthcare. Previous studies have not evaluated whether a datathon can promote affective learning and effective teamwork. METHODS: Fifty participants of a health datathon event in São Paulo, Brazil at Hospital Israelita Albert Einstein (HIAE) were divided into 8 groups. A survey with 25 questions, using the Affective Learning Scale and Team-Review Questionnaire, was administered to assess team effectiveness and affective learning during the event. Multivariate regression models and Pearson's correlation tests were performed to evaluate the effect of affective learning on teamwork. RESULTS: Majority of the participants were male 76% (37/49); 32% (16/49) were physicians. The mean score for learning (scale from 1 to 10) was 8.38, while that for relevance of the perceived teamwork was 1.20 (scale from 1 to 5; "1" means most relevant). Pearson's correlation between the learning score and perception of teamwork showed moderate association (r = 0.36, p = 0.009). Five learning and 10 teamwork variables were on average positively graded in the event. The final regression model includes all learning and teamwork variables. Effective leadership was strongly correlated with affective learning (ß = -0.27, p < 0.01, R2 = 75%). Effective leadership, team accomplishment, criticism, individual development and creativity were the variables significantly associated with higher levels of affective learning. CONCLUSION: It is feasible to enhance affective knowledge and the skill to work in a team during a datathon. We found that teamwork is associated with higher affective learning from participants' perspectives. Effective leadership is essential for teamwork and is a significant predictor of learning.