Timely and accurate prediction of short-term mortality is critical in intensive care units (ICUs), where patients' conditions change rapidly. Traditional scoring systems, such as the Simplified Acute Physiology Score and Acute Physiology and Chronic Health Evaluation, rely on static variables collected within the first 24 hours of admission and do not account for continuously evolving clinical states. These systems lack real-time adaptability, interpretability, and generalizability. With the increasing availability of high-frequency electronic medical record (EMR) data, machine learning (ML) approaches have emerged as powerful tools to model complex temporal patterns and support dynamic clinical decision-making. However, existing models are often limited by their inability to handle irregular sampling and missing values, and many lack rigorous external validation across institutions.

We aimed to develop a real-time, interpretable risk prediction model that continuously assesses ICU patient mortality using irregular, longitudinal EMR data, with improved performance and generalizability over traditional static scoring systems.

A time-aware bidirectional attention-based long short-term memory (TBAL) model was developed using EMR data from the MIMIC-IV (Medical Information Mart for Intensive Care) and eICU Collaborative Research Database (eICU-CRD) databases, comprising 176,344 ICU stays. The model incorporated dynamic variables, including vital signs, laboratory results, and medication data, updated hourly, to perform static and continuous mortality risk assessments. External cross-validation and subgroup sensitivity analyses were conducted to evaluate robustness and fairness. Model performance was assessed using the area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), accuracy, and F1-score. Interpretability was enhanced using integrated gradients to identify key predictors.

For the static 12-hour to 1-day mortality prediction task, the TBAL model achieved AUROCs of 95.9 (95% CI 94.2-97.5) and 93.3 (95% CI 91.5-95.3) and AUPRCs of 48.5 and 21.6 in MIMIC-IV and eICU-CRD, respectively. Accuracy and F1-scores reached 94.1 and 46.7 in MIMIC-IV and 92.2 and 28.1 in eICU-CRD. In dynamic prediction tasks, AUROCs reached 93.6 (95% CI 93.2-93.9) and 91.9 (95% CI 91.6-92.1), with AUPRCs of 41.3 and 50, respectively. The model maintained high recall for positive cases (82.6% and 79.1% in MIMIC-IV and eICU-CRD). Cross-database validation yielded AUROCs of 81.3 and 76.1, confirming generalizability. Subgroup analysis showed stable performance across age, sex, and severity strata, with top predictors including lactate, vasopressor use, and Glasgow Coma Scale score.

The TBAL model offers a robust, interpretable, and generalizable solution for dynamic real-time mortality risk prediction in ICU patients. Its ability to adapt to irregular temporal patterns and to provide hourly updated predictions positions it as a promising decision-support tool. Future work should validate its utility in prospective clinical trials and investigate its integration into real-world ICU workflows to enhance patient outcomes.

The number of elderly patients (≥90 years) admitted to intensive care units (ICUs) is continuously increasing. Survival estimations that differentiate into population and excess mortality are lacking but might be warranted given the higher baseline mortality in this group.

We evaluated excess mortality and risk factors in 1076 ICU patients ≥90 treated at the University Medical Center Hamburg-Eppendorf, Germany, from 2008 to 2019 by relative survival analysis. The Human Mortality Database of the German population served as reference.

Population mortality - vulgo baseline mortality - accounted for 22.2 % of the observed 1-year mortality rate of 57.8 %. Within the tenth and eleventh decade of life, excess mortality was not significantly influenced by age - neither in ICU patients (HR 0.98 (0.94-1.03), p = 0.46) nor in ICU survivors (HR 0.96 (0.91-1.02), p = 0·15, each per year). Yet, SAPS-II, SOFA scores, dementia, aortic valve stenosis, myocardial infarction and other factors significantly influenced survival. The initial excess mortality rate was high after ICU admission (excess mortality risk λ=0.059 for all patients), but lower after ICU survival (λ=0.022). After around 100 days, the excess mortality risk had disappeared.

In the largest dataset analyzed to date, we were able show that age did not significantly affect excess mortality of ICU patients aged 90 years or older. With respect to the severity of their illness, excess mortality of patients aged 90 years and older in ICUs is within an adequate range. Briefly after patients have left the ICU alive, mortality equals that of the general population.

BackgroundTo develop and validate a mortality prediction model for patients with sepsis-associated Acute Respiratory Distress Syndrome (ARDS).MethodsThis retrospective cohort study included 2466 patients diagnosed with sepsis and ARDS within 24 h of ICU admission. Demographic, clinical, and laboratory parameters were extracted from Medical Information Mart for Intensive Care III (MIMIC-III) database. Feature selection was performed using the Boruta algorithm, followed by the construction of seven ML models: logistic regression, Naive Bayes, k-nearest neighbor, support vector machine, decision tree, Random Forest, and extreme gradient boosting. Model performance was evaluated using the area under the receiver operating characteristic curve, accuracy, sensitivity, specificity, positive predictive value, and negative predictive value.ResultsThe study identified 24 variables significantly associated with mortality. The optimal ML model, a Random Forest model, demonstrated an AUC of 0.8015 in the test set, with high accuracy and specificity. The model highlighted the importance of blood urea nitrogen, age, urine output, Simplified Acute Physiology Score II, and albumin levels in predicting mortality.ConclusionsThe model's superior predictive performance underscores the potential for integrating advanced analytics into clinical decision-making processes, potentially improving patient outcomes and resource allocation in critical care settings.

Several artificial intelligence (AI)-driven tools have emerged for the hemodynamic evaluation of critically ill and surgical patients. This article provides an overview of current developments and potential clinical applications of machine learning (ML) for blood pressure measurements, hypotension prediction, hemodynamic profiling, and echocardiography. ML algorithms have shown promise in enabling continuous, non-invasive blood pressure monitoring by analyzing pulse oximetry waveforms, though these methods require periodic calibration with traditional oscillometric brachial cuffs. Additionally, a variety of ML models have been trained to forecast impending hypotension. However, clinical research indicates that these algorithms often primarily rely on mean arterial pressure, leading to questions about their added predictive value. The issue of false-positive alerts is also significant and can result in unwarranted clinical interventions. In terms of hemodynamic profiling, ML algorithms have been proposed to automatically classify patients into specific hemodynamic endotypes. However, current evidence suggests these models tend to replicate conventional hemodynamic profiles found in medical textbooks or depicted on advanced hemodynamic monitors. This raises questions about their practical clinical utility, especially given occasional discrepancies that could impact treatment decisions. Point-of-care ultrasound (POCUS) has gained traction for evaluating cardiac function in patients experiencing circulatory shock. ML algorithms now embedded in some POCUS systems can assist by recognizing ultrasound images, guiding users for optimal imaging, automating and reducing the variability of key echocardiographic measurements. These capabilities are especially beneficial for novice operators, potentially enhancing accuracy and confidence in clinical decision-making. In conclusion, while several AI-based technologies show promise for refining hemodynamic assessment in both critically ill and surgical patients, their clinical value varies. Comprehensive validation studies and real-world testing are essential to identify which innovations will genuinely contribute to improving the quality of care.

Machine learning (ML) is increasingly used to predict clinical deterioration in intensive care unit (ICU) patients through scoring systems. Although promising, such algorithms often overfit their training cohort and perform worse at new hospitals. Thus, external validation is a critical - but frequently overlooked - step to establish the reliability of predicted risk scores to translate them into clinical practice. We systematically reviewed how regularly external validation of ML-based risk scores is performed and how their performance changed in external data.

We searched MEDLINE, Web of Science, and arXiv for studies using ML to predict deterioration of ICU patients from routine data. We included primary research published in English before December 2023. We summarised how many studies were externally validated, assessing differences over time, by outcome, and by data source. For validated studies, we evaluated the change in area under the receiver operating characteristic (AUROC) attributable to external validation using linear mixed-effects models.

We included 572 studies, of which 84 (14.7%) were externally validated, increasing to 23.9% by 2023. Validated studies made disproportionate use of open-source data, with two well-known US datasets (MIMIC and eICU) accounting for 83.3% of studies. On average, AUROC was reduced by -0.037 (95% CI -0.052 to -0.027) in external data, with more than 0.05 reduction in 49.5% of studies.

External validation, although increasing, remains uncommon. Performance was generally lower in external data, questioning the reliability of some recently proposed ML-based scores. Interpretation of the results was challenged by an overreliance on the same few datasets, implicit differences in case mix, and exclusive use of AUROC.

Intensive care unit (ICU)-acquired weakness (ICU-AW) is a critical complication that significantly worsens patient prognosis. It is widely thought that risk prediction models can be harnessed to guide preventive interventions. While the number of ICU-AW risk prediction models is increasing, the quality and applicability of these models in clinical practice remain unclear.

The objective of this study was to systematically review published studies on risk prediction models for ICU-AW.

We searched electronic databases (PubMed, Web of Science, The Cochrane Library, Embase, Cumulative Index to Nursing and Allied Health Literature (CINAHL), China National Knowledge Infrastructure (CNKI), China Science and Technology Periodical Database (VIP), and Wanfang Database) from inception to October 2023 for studies on ICU-AW risk prediction models. Two independent researchers screened the literature, extracted data, and assessed the risk of bias and applicability of the included studies.

A total of 2709 articles were identified. After screening, 25 articles were selected, encompassing 25 risk prediction models. The area under the curve for these models ranged from 0.681 to 0.926. Evaluation of bias risk indicated that all included models exhibited a high risk of bias, with three models demonstrating poor applicability. The top five predictors among these models were mechanical ventilation duration, age, Acute Physiology and Chronic Health Evaluation II score, blood lactate levels, and the length of ICU stay. The combined area under the curve of the ten validation models was 0.83 (95% confidence interval: 0.77-0.88), indicating a strong discriminative ability.

Overall, ICU-AW risk prediction models demonstrate promising discriminative ability. However, further optimisation is needed to address limitations, including data source heterogeneity, potential biases in study design, and the need for robust statistical validation. Future efforts should prioritise external validation of existing models or the development of high-quality predictive models with superior performance.

The protocol for this study is registered with the International Prospective Register of Systematic Reviews (registration number: CRD42023453187).

Artificial intelligence is experiencing rapid growth, with continual innovation and advancements in the health care field.

This study aims to evaluate the application of artificial intelligence technologies across various domains of respiratory care.

We conducted a narrative review to examine the latest advancements in the use of artificial intelligence in the field of respiratory care. The search was independently conducted by respiratory care experts, each focusing on their respective scope of practice and area of interest.

This review illuminates the diverse applications of artificial intelligence, highlighting its use in areas associated with respiratory care. Artificial intelligence is harnessed across various areas in this field, including pulmonary diagnostics, respiratory care research, critical care or mechanical ventilation, pulmonary rehabilitation, telehealth, public health or health promotion, sleep clinics, home care, smoking or vaping behavior, and neonates and pediatrics. With its multifaceted utility, artificial intelligence can enhance the field of respiratory care, potentially leading to superior health outcomes for individuals under this extensive umbrella.

As artificial intelligence advances, elevating academic standards in the respiratory care profession becomes imperative, allowing practitioners to contribute to research and understand artificial intelligence's impact on respiratory care. The permanent integration of artificial intelligence into respiratory care creates the need for respiratory therapists to positively influence its progression. By participating in artificial intelligence development, respiratory therapists can augment their clinical capabilities, knowledge, and patient outcomes.

The purpose of this study was to develop and validate a mortality risk algorithm for pediatric surgery patients treated at KidsOR sites in 14 low- and middle-income countries.

A SuperLearner machine learning algorithm was trained to predict post-operative mortality by hospital discharge using the retrospectively and prospectively collected KidsOR database including patients treated at 20 KidsOR sites from June 2018 to June 2023. Algorithm performance was evaluated by internal-external cross-validated AUC and calibration.

Of 23,905 eligible patients, 21,703 with discharge status recorded were included in the analysis, representing a post-operative mortality rate of 3.1% (671 mortality events). The candidate algorithm with the best cross-validated performance was an extreme gradient boosting model. The cross-validated AUC was 0.945 (95% CI 0.936 to 0.954) and cross-validated calibration slope and intercept were 1.01 (95% CI 0.96 to 1.06) and 0.05 (95% CI -0.10 to 0.21). For Super Learner models trained on all but one site and evaluated in the holdout site for sites with at least 25 mortality events, overall external validation AUC was 0.864 (95% CI 0.846 to 0.882) with calibration slope and intercept of 1.03 (95% CI 0.97 to 1.09) and 1.18 (95% CI 0.98 to 1.39).

The KidsOR post-operative mortality risk algorithm had outstanding cross-validated discrimination and strong cross-validated calibration. Across all external validation sites, discrimination of Super Learner models trained on the remaining sites was excellent, though re-calibration may be necessary prior to use at new sites. This model has the potential to inform clinical practice and guide resource allocation at KidsOR sites world-wide.

Observational Study, Level III.

A fundamental goal of evaluating the performance of a clinical model is to ensure it performs well across a diverse intended patient population. A primary challenge is that the data used in model development and testing often consist of many overlapping, heterogeneous patient subgroups that may not be explicitly defined or labeled. While a model's average performance on a dataset may be high, the model can have significantly lower performance for certain subgroups, which may be hard to detect. We describe an algorithmic framework for identifying subgroups with potential performance disparities (AFISP), which produces a set of interpretable phenotypes corresponding to subgroups for which the model's performance may be relatively lower. This could allow model evaluators, including developers and users, to identify possible failure modes prior to wide-scale deployment. We illustrate the application of AFISP by applying it to a patient deterioration model to detect significant subgroup performance disparities, and show that AFISP is significantly more scalable than existing algorithmic approaches.

To evaluate the transferability of deep learning (DL) models for the early detection of adverse events to previously unseen hospitals.

Retrospective observational cohort study utilizing harmonized intensive care data from four public datasets.

ICUs across Europe and the United States.

Adult patients admitted to the ICU for at least 6 hours who had good data quality.

None.

Using carefully harmonized data from a total of 334,812 ICU stays, we systematically assessed the transferability of DL models for three common adverse events: death, acute kidney injury (AKI), and sepsis. We tested whether using more than one data source and/or algorithmically optimizing for generalizability during training improves model performance at new hospitals. We found that models achieved high area under the receiver operating characteristic (AUROC) for mortality (0.838-0.869), AKI (0.823-0.866), and sepsis (0.749-0.824) at the training hospital. As expected, AUROC dropped when models were applied at other hospitals, sometimes by as much as -0.200. Using more than one dataset for training mitigated the performance drop, with multicenter models performing roughly on par with the best single-center model. Dedicated methods promoting generalizability did not noticeably improve performance in our experiments.

Our results emphasize the importance of diverse training data for DL-based risk prediction. They suggest that as data from more hospitals become available for training, models may become increasingly generalizable. Even so, good performance at a new hospital still depended on the inclusion of compatible hospitals during training.

There has been a growing body of research focusing on patients with Congestive Heart Failure (CHF) and chronic obstructive pulmonary disease (COPD) admitted to the intensive care unit (ICU). However, the optimal blood pressure (BP) level for such patients remains insufficiently explored. This study aimed to investigate the associations between systolic blood pressure (SBP) and in-hospital mortality among ICU patients with both CHF and COPD.

This retrospective cohort study enrolled 6309 patients from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. SBP was examined as both a continuous and categorical variable, with the primary outcome being in-hospital mortality. The investigation involved multivariable logistic regression, restricted cubic spline regression, and subgroup analysis to determine the relationship between SBP and mortality.

The cohort consisted of 6309 patients with concurrent CHF and COPD (3246 females and 3063 males), with an average age of 73.0 ± 12.5 years. The multivariate analysis revealed an inverse association between SBP and in-hospital mortality, both as a continuous variable (odds ratio = 0.99 [95% CI, 0.99~1]) and as a categorical variable (divided into quintiles). Restricted cubic spline analysis demonstrated an L-shaped relationship between SBP and mortality risk (P nonlinearity < 0.001), with an inflection point at 99.479 mmHg. Stratified analyses further supported the robustness of this correlation.

The relationship between SBP and in-hospital mortality in patients with both CHF and COPD follows an L-shaped pattern, with an inflection point at approximately 99.479 mmHg.

Clinical prediction modeling plays a pivotal part in modern clinical care, particularly in predicting the risk of in-hospital mortality. Recent modeling efforts have focused on leveraging intraoperative data sources to improve model performance. However, the individual and collective benefit of pre- and intraoperative data for clinical decision-making remains unknown. We hypothesized that pre- and intraoperative predictors contribute equally to the net benefit in a decision curve analysis (DCA) of in-hospital mortality prediction models that include pre- and intraoperative predictors.

Data from the VitalDB database featuring a subcohort of 6043 patients were used. A total of 141 predictors for in-hospital mortality were grouped into preoperative (demographics, intervention characteristics, and laboratory measurements) and intraoperative (laboratory and monitor data, drugs, and fluids) data. Prediction models using either preoperative, intraoperative, or all data were developed with multiple methods (logistic regression, neural network, random forest, gradient boosting machine, and a stacked learner). Predictive performance was evaluated by the area under the receiver-operating characteristic curve (AUROC) and under the precision-recall curve (AUPRC). Clinical utility was examined with a DCA in the predefined risk preference range (denoted by so-called treatment threshold probabilities) between 0% and 20%.

AUROC performance of the prediction models ranged from 0.53 to 0.78. AUPRC values ranged from 0.02 to 0.25 (compared to the incidence of 0.09 in our dataset) and high AUPRC values resulted from prediction models based on preoperative laboratory values. A DCA of pre- and intraoperative prediction models highlighted that preoperative data provide the largest overall benefit for decision-making, whereas intraoperative values provide only limited benefit for decision-making compared to preoperative data. While preoperative demographics, comorbidities, and surgery-related data provide the largest benefit for low treatment thresholds up to 5% to 10%, preoperative laboratory measurements become the dominant source for decision support for higher thresholds.

When it comes to predicting in-hospital mortality and subsequent decision-making, preoperative demographics, comorbidities, and surgery-related data provide the largest benefit for clinicians with risk-averse preferences, whereas preoperative laboratory values provide the largest benefit for decision-makers with more moderate risk preferences. Our decision-analytic investigation of different predictor categories moves beyond the question of whether certain predictors provide a benefit in traditional performance metrics (eg, AUROC). It offers a nuanced perspective on for whom these predictors might be beneficial in clinical decision-making. Follow-up studies requiring larger datasets and dedicated deep-learning models to handle continuous intraoperative data are essential to examine the robustness of our results.

Healthcare professionals may be able to anticipate more accurately a patient's timing of death and assess their possibility of recovery by implementing a real-time clinical decision support system. Using such a tool, the healthcare system can better understand a patient's condition and make more informed judgements about distributing limited resources. This scoping review aimed to analyze various death prediction AI (Artificial Intelligence) algorithms that have been used in ICU (Intensive Care Unit) patient populations.

The search strategy of this study involved keyword combinations of outcome and patient setting such as mortality, survival, ICU, terminal care. These terms were used to perform database searches in MEDLINE, Embase, and PubMed up to July 2022. The variables, characteristics, and performance of the identified predictive models were summarized. The accuracy of the models was compared using their Area Under the Curve (AUC) values.

Databases search yielded an initial pool of 8271 articles. A two-step screening process was then applied: first, titles and abstracts were reviewed for relevance, reducing the pool to 429 articles. Next, a full-text review was conducted, further narrowing down the selection to 400 key studies. Out of 400 studies on different tools or models for prediction of mortality in ICUs, 16 papers focused on AI-based models which were ultimately included in this study that have deployed different AI-based and machine learning models to make a prediction about negative patient outcome. The accuracy and performance of the different models varied depending on the patient populations and medical conditions. It was found that AI models compared with traditional tools like SAP3 or APACHE IV score were more accurate in death prediction, with some models achieving an AUC of up to 92.9%. The overall mortality rate ranged from 5% to more than 60% in different studies.

We found that AI-based models exhibit varying performance across different patient populations. To enhance the accuracy of mortality prediction, we recommend customizing models for specific patient groups and medical contexts. By doing so, healthcare professionals may more effectively assess mortality risk and tailor treatments accordingly. Additionally, incorporating additional variables-such as genetic information-into new models can further improve their accuracy.

Surgery in geriatric patients often poses risk of major postoperative complications. Acute kidney injury (AKI) is a common complication following noncardiac surgery and is associated with increased mortality. Early identification of geriatric patients at high risk of AKI could facilitate preventive measures and improve patient prognosis. This study used machine learning methods to identify important features and predict AKI following noncardiac surgery in geriatric patients.

The data for this study were obtained from a prospective cohort. Patients aged ≥ 65 years who received noncardiac surgery from June 2019 to December 2021 were enrolled. Data were split into training set (from June 2019 to March 2021) and internal validation set (from April 2021 to December 2021) by time. The least absolute shrinkage and selection operator (LASSO) regularization algorithm and the random forest recursive feature elimination algorithm (RF-RFE) were used to screen important predictors. Models were trained through extreme gradient boosting (XGBoost), random forest, and LASSO. The SHapley Additive exPlanations (SHAP) package was used to interpret the machine learning model.

The training set included 6753 geriatric patients. Of these, 250 (3.70%) patients developed AKI. The XGBoost model with RF-RFE selected features outperformed other models with an area under the precision-recall curve (AUPRC) of 0.505 (95% confidence interval [CI]: 0.369-0.626) and an area under the receiver operating characteristic curve (AUROC) of 0.806 (95%CI: 0.733-0.875). The model incorporated ten predictors, including operation site and hypertension. The internal validation set included 3808 geriatric patients, and 96 (2.52%) patients developed AKI. The model maintained good predictive performance with an AUPRC of 0.431 (95%CI: 0.331-0.524) and an AUROC of 0.845 (95%CI: 0.796-0.888) in the internal validation.

This study developed a simple machine learning model and a web calculator for predicting AKI following noncardiac surgery in geriatric patients. This model may be a valuable tool for guiding preventive measures and improving patient prognosis.

The protocol of this study was approved by the Committee of Ethics from West China Hospital of Sichuan University (2019-473) with a waiver of informed consent and registered at www.chictr.org.cn (ChiCTR1900025160, 15/08/2019).

A major focus of academia, industry, and global governmental agencies is to develop and apply artificial intelligence and other advanced analytical tools to transform health care delivery. The American Heart Association supports the creation of tools and services that would further the science and practice of precision medicine by enabling more precise approaches to cardiovascular and stroke research, prevention, and care of individuals and populations. Nevertheless, several challenges exist, and few artificial intelligence tools have been shown to improve cardiovascular and stroke care sufficiently to be widely adopted. This scientific statement outlines the current state of the art on the use of artificial intelligence algorithms and data science in the diagnosis, classification, and treatment of cardiovascular disease. It also sets out to advance this mission, focusing on how digital tools and, in particular, artificial intelligence may provide clinical and mechanistic insights, address bias in clinical studies, and facilitate education and implementation science to improve cardiovascular and stroke outcomes. Last, a key objective of this scientific statement is to further the field by identifying best practices, gaps, and challenges for interested stakeholders.

The lack of transparency is a prevalent issue among the current machine-learning (ML) algorithms utilized for predicting mortality risk. Herein, we aimed to improve transparency by utilizing the latest ML explicable technology, SHapley Additive exPlanation (SHAP), to develop a predictive model for critically ill patients.

We extracted data from the Medical Information Mart for Intensive Care IV database, encompassing all intensive care unit admissions. We employed nine different methods to develop the models. The most accurate model, with the highest area under the receiver operating characteristic curve, was selected as the optimal model. Additionally, we used SHAP to explain the workings of the ML model.

The study included 21 395 critically ill patients, with a median age of 68 years (interquartile range, 56-79 years), and most patients were male (56.9%). The cohort was randomly split into a training set (N = 16 046) and a validation set (N = 5349). Among the nine models developed, the Random Forest model had the highest accuracy (87.62%) and the best area under the receiver operating characteristic curve value (0.89). The SHAP summary analysis showed that Glasgow Coma Scale, urine output, and blood urea nitrogen were the top three risk factors for outcome prediction. Furthermore, SHAP dependency analysis and SHAP force analysis were used to interpret the Random Forest model at the factor level and individual level, respectively.

A transparent ML model for predicting outcomes in critically ill patients using SHAP methodology is feasible and effective. SHAP values significantly improve the explainability of ML models.

A real-time model for predicting short-term mortality in critically ill patients is needed to identify patients at imminent risk. However, the performance of the model needs to be validated in various clinical settings and ethnicities before its clinical application. In this study, we aim to develop an ensemble machine learning model using routinely measured clinical variables at a single academic institution in South Korea.

We developed an ensemble model using deep learning and light gradient boosting machine models. Internal validation was performed using the last two years of the internal cohort dataset, collected from a single academic hospital in South Korea between 2007 and 2021. External validation was performed using the full Medical Information Mart for Intensive Care (MIMIC), eICU-Collaborative Research Database (eICU-CRD), and Amsterdam University Medical Center database (AmsterdamUMCdb) data. The area under the receiver operating characteristic curve (AUROC) was calculated and compared to that for the National Early Warning Score (NEWS).

The developed model (iMORS) demonstrated high predictive performance with an internal AUROC of 0.964 (95% confidence interval [CI] 0.963-0.965) and external AUROCs of 0.890 (95% CI 0.889-0.891) for MIMIC, 0.886 (95% CI 0.885-0.887) for eICU-CRD, and 0.870 (95% CI 0.868-0.873) for AmsterdamUMCdb. The model outperformed the NEWS with higher AUROCs in the internal and external validation (0.866 for the internal, 0.746 for MIMIC, 0.798 for eICU-CRD, and 0.819 for AmsterdamUMCdb; p < 0.001).

Our real-time machine learning model to predict short-term mortality in critically ill patients showed excellent performance in both internal and external validations. This model could be a useful decision-support tool in the intensive care units to assist clinicians.

Critical care medicine in the 21st century has witnessed remarkable advancements that have significantly improved patient outcomes in intensive care units (ICUs). This abstract provides a concise summary of the latest developments in critical care, highlighting key areas of innovation. Recent advancements in critical care include Precision Medicine: Tailoring treatments based on individual patient characteristics, genomics, and biomarkers to enhance the effectiveness of therapies. The objective is to describe the recent advancements in Critical Care Medicine. Telemedicine: The integration of telehealth technologies for remote patient monitoring and consultation, facilitating timely interventions. Artificial intelligence (AI): AI-driven tools for early disease detection, predictive analytics, and treatment optimization, enhancing clinical decision-making. Organ Support: Advanced life support systems, such as Extracorporeal Membrane Oxygenation and Continuous Renal Replacement Therapy provide better organ support. Infection Control: Innovative infection control measures to combat emerging pathogens and reduce healthcare-associated infections. Ventilation Strategies: Precision ventilation modes and lung-protective strategies to minimize ventilator-induced lung injury. Sepsis Management: Early recognition and aggressive management of sepsis with tailored interventions. Patient-Centered Care: A shift towards patient-centered care focusing on psychological and emotional well-being in addition to medical needs. We conducted a thorough literature search on PubMed, EMBASE, and Scopus using our tailored strategy, incorporating keywords such as critical care, telemedicine, and sepsis management. A total of 125 articles meeting our criteria were included for qualitative synthesis. To ensure reliability, we focused only on articles published in the English language within the last two decades, excluding animal studies, in vitro/molecular studies, and non-original data like editorials, letters, protocols, and conference abstracts. These advancements reflect a dynamic landscape in critical care medicine, where technology, research, and patient-centered approaches converge to improve the quality of care and save lives in ICUs. The future of critical care promises even more innovative solutions to meet the evolving challenges of modern medicine.

Systemic inflammatory response syndrome (SIRS) and sepsis are the most common causes of in-hospital death. However, the characteristics associated with the improvement in the patient conditions during the ICU stay were not fully elucidated for each population as well as the possible differences between the two.

The aim of this study is to highlight the differences between the prognostic clinical features for the survival of patients diagnosed with SIRS and those of patients diagnosed with sepsis by using a multi-variable predictive modeling approach with a reduced set of easily available measurements collected at the admission to the intensive care unit (ICU).

Data were collected from 1,257 patients (816 non-sepsis SIRS and 441 sepsis) admitted to the ICU. We compared the performance of five machine learning models in predicting patient survival. Matthews correlation coefficient (MCC) was used to evaluate model performances and feature importance, and by applying Monte Carlo stratified Cross-Validation.

Extreme Gradient Boosting (MCC = 0.489) and Logistic Regression (MCC = 0.533) achieved the highest results for SIRS and sepsis cohorts, respectively. In order of importance, APACHE II, mean platelet volume (MPV), eosinophil counts (EoC), and C-reactive protein (CRP) showed higher importance for predicting sepsis patient survival, whereas, SOFA, APACHE II, platelet counts (PLTC), and CRP obtained higher importance in the SIRS cohort.

By using complete blood count parameters as predictors of ICU patient survival, machine learning models can accurately predict the survival of SIRS and sepsis ICU patients. Interestingly, feature importance highlights the role of CRP and APACHE II in both SIRS and sepsis populations. In addition, MPV and EoC are shown to be important features for the sepsis population only, whereas SOFA and PLTC have higher importance for SIRS patients.

The limitations of current early warning scores have prompted the development of deep learning-based systems, such as deep learning-based cardiac arrest risk management systems (DeepCARS). Unfortunately, in South Korea, only two institutions operate 24-hour Rapid Response System (RRS), whereas most hospitals have part-time or no RRS coverage at all. This study validated the predictive performance of DeepCARS during RRS operation and nonoperation periods and explored its potential beyond RRS operating hours.

Retrospective cohort study.

In this 1-year retrospective study conducted at Yonsei University Health System Severance Hospital in South Korea, DeepCARS was compared with conventional early warning systems for predicting in-hospital cardiac arrest (IHCA). The study focused on adult patients admitted to the general ward, with the primary outcome being IHCA-prediction performance within 24 hours of the alarm.

We analyzed the data records of adult patients admitted to a general ward from September 1, 2019, to August 31, 2020.

None.

Performance evaluation was conducted separately for the operational and nonoperational periods of the RRS, using the area under the receiver operating characteristic curve (AUROC) as the metric. DeepCARS demonstrated a superior AUROC as compared with the Modified Early Warning Score (MEWS) and the National Early Warning Score (NEWS), both during RRS operating and nonoperating hours. Although the MEWS and NEWS exhibited varying performance across the two periods, DeepCARS showed consistent performance.

The accuracy and efficiency for predicting IHCA of DeepCARS were superior to that of conventional methods, regardless of whether the RRS was in operation. These findings emphasize that DeepCARS is an effective screening tool suitable for hospitals with full-time RRS, part-time RRS, and even those without any RRS.

End-of-life care (EOLC) exemplifies the joint mission of intensive and palliative care (PC) in their human-centeredness. The explosion of technological advances in medicine must be balanced with the culture of holistic care. Inevitably, it brings together the science and the art of medicine in their full expression. High-quality EOLC in the ICU is grounded in evidence, ethical principles, and professionalism within the framework of the Law. Expert professional statements over the last two decades in India were developed while the law was evolving. Recent landmark Supreme Court judgments have necessitated a review of the clinical pathway for EOLC outlined in the previous statements. Much empirical and interventional evidence has accumulated since the position statement in 2014. This iteration of the joint Indian Society of Critical Care Medicine-Indian Association of Palliative Care (ISCCM-IAPC) Position Statement for EOLC combines contemporary evidence, ethics, and law for decision support by the bedside in Indian ICUs.

Mani RK, Bhatnagar S, Butola S, Gursahani R, Mehta D, Simha S, et al. Indian Society of Critical Care Medicine and Indian Association of Palliative Care Expert Consensus and Position Statements for End-of-life and Palliative Care in the Intensive Care Unit. Indian J Crit Care Med 2024;28(3):200-250.

This study aims to develop a predictive model using artificial intelligence to estimate the ICU length of stay (LOS) for Congenital Heart Defects (CHD) patients after surgery, improving care planning and resource management.

We analyze clinical data from 2240 CHD surgery patients to create and validate the predictive model. Twenty AI models are developed and evaluated for accuracy and reliability.

The study is conducted in a Brazilian hospital's Cardiovascular Surgery Department, focusing on transplants and cardiopulmonary surgeries.

Retrospective analysis is conducted on data from 2240 consecutive CHD patients undergoing surgery.

Ninety-three pre and intraoperative variables are used as ICU LOS predictors.

Utilizing regression and clustering methodologies for ICU LOS (ICU Length of Stay) estimation, the Light Gradient Boosting Machine, using regression, achieved a Mean Squared Error (MSE) of 15.4, 11.8, and 15.2 days for training, testing, and unseen data. Key predictors included metrics such as "Mechanical Ventilation Duration", "Weight on Surgery Date", and "Vasoactive-Inotropic Score". Meanwhile, the clustering model, Cat Boost Classifier, attained an accuracy of 0.6917 and AUC of 0.8559 with similar key predictors.

Patients with higher ventilation times, vasoactive-inotropic scores, anoxia time, cardiopulmonary bypass time, and lower weight, height, BMI, age, hematocrit, and presurgical oxygen saturation have longer ICU stays, aligning with existing literature.

Goal: The purpose of this study is to evaluate the importance of cardiorespiratory variables within a Reinforcement Learning (RL) recommendation system aimed at establishing optimal strategies for drug treatment of septic patients in the intensive care unit (ICU). Methods: We developed a RL model in order to establish drug administration strategies for septic patients using only a set of cardiorespiratory variables. We then compared this model with other RL models trained with a different set of features. We selected patients meeting the Sepsis-3 criteria from the Multi-parameter Intelligent Monitoring in Intensive Care (MIMIC III) database, resulting in a total of 20,496 ICU admissions. A Markov Decision Process (MDP) was built on the extracted discrete time-series. A policy iteration algorithm was used to obtain the optimal AI policy for the MDP. The policy performance was then evaluated using the WIS estimator. The process was repeated for each set of variables and compared to a set of baseline benchmark policies. Results: The model trained with cardiorespiratory variables outperformed all other models considered, resulting in a 95% confidence lower bound score of 97.48. This finding highlights the importance of cardiovascular variables in the clinical RL recommendation system. Conclusions: We established an efficient RL model for sepsis treatment in the ICU and demonstrated that cardiorespiratory variables provides critical information in devising optimal policies. Given the potentially continuous availability of cardiorespiratory features extracted from bedside physiological waveform monitoring, the proposed framework paves the way for a real time recommendation system for sepsis treatment.

In intensive care units (ICUs), accurate mortality prediction is crucial for effective patient management and resource allocation. The Simplified Acute Physiology Score II (SAPS-2), though commonly used, relies heavily on comprehensive clinical data and blood samples. This study sought to develop an artificial intelligence (AI) model utilizing key hemodynamic parameters to predict ICU mortality within the first 24 h and assess its performance relative to SAPS-2.

We conducted an analysis of select hemodynamic parameters and the structure of heart rate curves to identify potential predictors of ICU mortality. A machine-learning model was subsequently trained and validated on distinct patient cohorts. The AI algorithm's performance was then compared to the SAPS-2, focusing on classification accuracy, calibration, and generalizability.

The study included 1298 ICU admissions from March 27th, 2015, to March 27th, 2017. An additional cohort from 2022 to 2023 comprised 590 patients, resulting in a total dataset of 1888 patients. The observed mortality rate stood at 24.0%. Key determinants of mortality were the Glasgow Coma Scale score, heart rate complexity, patient age, duration of diastolic blood pressure below 50 mmHg, heart rate variability, and specific mean and systolic blood pressure thresholds. The AI model, informed by these determinants, exhibited a performance profile in predicting mortality that was comparable, if not superior, to the SAPS-2.

The AI model, which integrates heart rate and blood pressure curve analyses with basic clinical parameters, provides a methodological approach to predict in-hospital mortality in ICU patients. This model offers an alternative to existing tools that depend on extensive clinical data and laboratory inputs. Its potential integration into ICU monitoring systems may facilitate more streamlined mortality prediction processes.

COVID-19 has a range of complications, from no symptoms to severe pneumonia. It can also affect multiple organs including the nervous system. COVID-19 affects the brain, leading to neurological symptoms such as delirium. Delirium, a sudden change in consciousness, can increase the risk of death and prolong the hospital stay. However, research on delirium prediction in patients with COVID-19 is insufficient. This study aimed to identify new risk factors that could predict the onset of delirium in patients with COVID-19 using machine learning (ML) applied to nursing records. This retrospective cohort study used natural language processing and ML to develop a model for classifying the nursing records of patients with delirium. We extracted the features of each word from the model and grouped similar words. To evaluate the usefulness of word groups in predicting the occurrence of delirium in patients with COVID-19, we analyzed the temporal changes in the frequency of occurrence of these word groups before and after the onset of delirium. Moreover, the sensitivity, specificity, and odds ratios were calculated. We identified (1) elimination-related behaviors and conditions and (2) abnormal patient behavior and conditions as risk factors for delirium. Group 1 had the highest sensitivity (0.603), whereas group 2 had the highest specificity and odds ratio (0.938 and 6.903, respectively). These results suggest that these parameters may be useful in predicting delirium in these patients. The risk factors for COVID-19-associated delirium identified in this study were more specific but less sensitive than the ICDSC (Intensive Care Delirium Screening Checklist) and CAM-ICU (Confusion Assessment Method for the Intensive Care Unit). However, they are superior to the ICDSC and CAM-ICU because they can predict delirium without medical staff and at no cost.

The slowing of agricultural productivity growth globally over the past two decades has brought a new urgency to detect its drivers and potential solutions. We show that air pollution, particularly surface ozone (O3), is strongly associated with declining agricultural total factor productivity (TFP) in China. We employ machine learning algorithms to generate estimates of high-resolution surface O3 concentrations from 2002 to 2019. Results indicate that China's O3 pollution has intensified over this 18-year period. We coupled these O3 estimates with a statistical model to show that rising O3 pollution during nonwinter seasons has reduced agricultural TFP by 18% over the 2002-2015 period. Agricultural TFP is projected to increase by 60% if surface O3 concentrations were reduced to meet the WHO air quality standards. This productivity gain has the potential to counter expected productivity losses from 2°C warming.

Sepsis is accompanied by a considerably high risk of mortality in the short term, despite the availability of recommended mortality risk assessment tools. However, these risk assessment tools seem to have limited predictive value. With the gradual integration of machine learning into clinical practice, some researchers have attempted to employ machine learning for early mortality risk prediction in sepsis patients. Nevertheless, there is a lack of comprehensive understanding regarding the construction of predictive variables using machine learning and the value of various machine learning methods. Thus, we carried out this systematic review and meta-analysis to explore the predictive value of machine learning for sepsis-related death at different time points.

PubMed, Embase, Cochrane, and Web of Science databases were searched until August 9th, 2022. The risk of bias in predictive models was assessed using the Prediction model Risk of Bias Assessment Tool (PROBAST). We also performed subgroup analysis according to time of death and type of model and summarized current predictive variables used to construct models for sepsis death prediction.

Fifty original studies were included, covering 104 models. The combined Concordance index (C-index), sensitivity, and specificity of machine learning models were 0.799, 0.81, and 0.80 in the training set, and 0.774, 0.71, and 0.68 in the validation set, respectively. Machine learning outperformed conventional clinical scoring tools and showed excellent C-index, sensitivity, and specificity in different subgroups. Random Forest (RF) and eXtreme Gradient Boosting (XGBoost) are the preferred machine learning models because they showed more favorable accuracy with similar modeling variables. This study found that lactate was the most frequent predictor but was seriously ignored by current clinical scoring tools.

Machine learning methods demonstrate relatively favorable accuracy in predicting the mortality risk in sepsis patients. Given the limitations in accuracy and applicability of existing prediction scoring systems, there is an opportunity to explore updates based on existing machine learning approaches. Specifically, it is essential to develop or update more suitable mortality risk assessment tools based on the specific contexts of use, such as emergency departments, general wards, and intensive care units.

To assess the value of machine learning approaches in the development of a multivariable model for early prediction of ICU death in patients with acute respiratory distress syndrome (ARDS).

A development, testing, and external validation study using clinical data from four prospective, multicenter, observational cohorts.

A network of multidisciplinary ICUs.

A total of 1,303 patients with moderate-to-severe ARDS managed with lung-protective ventilation.

None.

We developed and tested prediction models in 1,000 ARDS patients. We performed logistic regression analysis following variable selection by a genetic algorithm, random forest and extreme gradient boosting machine learning techniques. Potential predictors included demographics, comorbidities, ventilatory and oxygenation descriptors, and extrapulmonary organ failures. Risk modeling identified some major prognostic factors for ICU mortality, including age, cancer, immunosuppression, Pa o2 /F io2 , inspiratory plateau pressure, and number of extrapulmonary organ failures. Together, these characteristics contained most of the prognostic information in the first 24 hours to predict ICU mortality. Performance with machine learning methods was similar to logistic regression (area under the receiver operating characteristic curve [AUC], 0.87; 95% CI, 0.82-0.91). External validation in an independent cohort of 303 ARDS patients confirmed that the performance of the model was similar to a logistic regression model (AUC, 0.91; 95% CI, 0.87-0.94).

Both machine learning and traditional methods lead to promising models to predict ICU death in moderate/severe ARDS patients. More research is needed to identify markers for severity beyond clinical determinants, such as demographics, comorbidities, lung mechanics, oxygenation, and extrapulmonary organ failure to guide patient management.

Artificial intelligence (AI) holds great promise for transforming the healthcare industry. However, despite its potential, AI is yet to see widespread deployment in clinical settings in significant part due to the lack of publicly available clinical data and the lack of transparency in the published AI algorithms. There are few clinical data repositories publicly accessible to researchers to train and test AI algorithms, and even fewer that contain specialized data from the perioperative setting. To address this gap, we present and release the Medical Informatics Operating Room Vitals and Events Repository (MOVER).

This first release of MOVER includes adult patients who underwent surgery at the University of California, Irvine Medical Center from 2015 to 2022. Data for patients who underwent surgery were captured from 2 different sources: High-fidelity physiological waveforms from all of the operating rooms were captured in real time and matched with electronic medical record data.

MOVER includes data from 58 799 unique patients and 83 468 surgeries. MOVER is available for download at https://doi.org/10.24432/C5VS5G, it can be downloaded by anyone who signs a data usage agreement (DUA), to restrict traffic to legitimate researchers.

To the best of our knowledge MOVER is the only freely available public data repository that contains electronic health record and high-fidelity physiological waveforms data for patients undergoing surgery.

MOVER is freely available to all researchers who sign a DUA, and we hope that it will accelerate the integration of AI into healthcare settings, ultimately leading to improved patient outcomes.

To systematically review contemporary prediction models for hospital mortality developed or validated in general medical patients.

We screened articles in five databases, from January 1, 2010, through April 7, 2022, and the bibliography of articles selected for final inclusion. We assessed the quality for risk of bias and applicability using the Prediction Model Risk of Bias Assessment Tool (PROBAST) and extracted data using the Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS) checklist. Two investigators independently screened each article, assessed quality, and extracted data.

From 20,424 unique articles, we identified 15 models in 8 studies across 10 countries. The studies included 280,793 general medical patients and 19,923 hospital deaths. Models included 7 early warning scores, 2 comorbidities indices, and 6 combination models. Ten models were studied in all general medical patients (general models) and 7 in general medical patients with infection (infection models). Of the 15 models, 13 were developed using logistic or Poisson regression and 2 using machine learning methods. Also, 4 of 15 models reported on handling of missing values. None of the infection models had high discrimination, whereas 4 of 10 general models had high discrimination (area under curve >0.8). Only 1 model appropriately assessed calibration. All models had high risk of bias; 4 of 10 general models and 5 of 7 infection models had low concern for applicability for general medical patients.

Mortality prediction models for general medical patients were sparse and differed in quality, applicability, and discrimination. These models require hospital-level validation and/or recalibration in general medical patients to guide mortality reduction interventions.

Ischemic stroke (IS) is a common and severe condition that requires intensive care unit (ICU) admission, with high mortality and variable prognosis. Accurate and reliable predictive tools that enable early risk stratification can facilitate interventions to improve patient outcomes; however, such tools are currently lacking. In this study, we developed and validated novel ensemble learning models based on soft voting and stacking methods to predict in-hospital mortality from IS in the ICU using two public databases: MIMIC-IV and eICU-CRD. Additionally, we identified the key predictors of mortality and developed a user-friendly online prediction tool for clinical use. The soft voting ensemble model, named ICU-ISPM, achieved an AUROC of 0.861 (95% CI: 0.829-0.892) and 0.844 (95% CI: 0.819-0.869) in the internal and external test cohorts, respectively. It significantly outperformed the APACHE scoring system and was more robust than individual models. ICU-ISPM obtained the highest performance compared to other models in similar studies. Using the SHAP method, the model was interpretable, revealing that GCS score, age, and intubation were the most important predictors of mortality. This model also provided a risk stratification system that can effectively distinguish between low-, medium-, and high-risk patients. Therefore, the ICU-ISPM is an accurate, reliable, interpretable, and clinically applicable tool, which is expected to assist clinicians in stratifying IS patients by the risk of mortality and rationally allocating medical resources. Based on ICU-ISPM, an online risk prediction tool was further developed, which was freely available at: http://ispm.idrblab.cn/.

Sepsis is one of the most common causes of death after surgery. Several conventional scoring systems have been developed to predict the outcome of sepsis; however, their predictive power is insufficient. The present study applies explainable machine-learning algorithms to improve the accuracy of predicting postoperative mortality in patients with sepsis caused by peritonitis.

We performed a retrospective analysis of data from demographic, clinical, and laboratory analyses, including the delta neutrophil index (DNI), WBC and neutrophil counts, and CRP level. Laboratory data were measured before surgery, 12-36 hours after surgery, and 60-84 hours after surgery. The primary study output was the probability of mortality. The areas under the receiver operating characteristic curves (AUCs) of several machine-learning algorithms using the Sequential Organ Failure Assessment (SOFA) and Simplified Acute Physiology Score (SAPS) 3 models were compared. 'SHapley Additive exPlanations' values were used to indicate the direction of the relationship between a variable and mortality.

The CatBoost model yielded the highest AUC (0.933) for mortality compared to SAPS3 and SOFA (0.860 and 0.867, respectively). Increased DNI on day 3, septic shock, use of norepinephrine therapy, and increased international normalized ratio on day 3 had the greatest impact on the model's prediction of mortality.

Machine-learning algorithms increase the accuracy of predicting postoperative mortality in patients with sepsis caused by peritonitis.

 Diagnosing appendicitis in young children (0-12 years) still poses a special difficulty despite the advent of radiological investigations. Few scoring models have evolved and been applied worldwide, but with significant fluctuations in accuracy upon validation.

 To utilize artificial intelligence (AI) techniques to develop and validate a diagnostic model based on clinical and laboratory parameters only (without imaging), in addition to prospective validation to confirm the findings.

 In Stage-I, observational data of children (0-12 years), referred for acute appendicitis (March 1, 2016-February 28, 2019, n = 166), was used for model development and evaluation using 10-fold cross-validation (XV) technique to simulate a prospective validation. In Stage-II, prospective validation of the model and the XV estimates were performed (March 1, 2019-November 30, 2021, n = 139).

 The developed model, AI Pediatric Appendicitis Decision-tree (AiPAD), is both accurate and explainable, with an XV estimation of average accuracy to be 93.5% ± 5.8 (91.4% positive predictive value [PPV] and 94.8% negative predictive value [NPV]). Prospective validation revealed that the model was indeed accurate and close to the XV evaluations, with an overall accuracy of 97.1% (96.7% PPV and 97.4% NPV).

 The AiPAD is validated, highly accurate, easy to comprehend, and offers an invaluable tool to use in diagnosing appendicitis in children without the need for imaging. Ultimately, this would lead to significant practical benefits, improved outcomes, and reduced costs.

Given the growing utilization of critical care services by an aging population, development of population-level risk models which predict intensive care unit (ICU) survivorship and mortality may offer advantages for researchers and health systems. Our objective was to develop a risk model for ICU survivorship and mortality among community dwelling older adults.

This was a population-based cohort study of 48,127 patients who were 50 years and older with at least one primary care visit between January 1, 2017, and December 31, 2017. We used electronic health record (EHR) data to identify variables predictive of ICU survivorship.

ICU admission and mortality within 2 years after index primary care visit date were used to divide patients into three groups of "alive without ICU admission", "ICU survivors," and "death." Multinomial logistic regression was used to identify EHR predictive variables for the three patient outcomes. Cross-validation by randomly splitting the data into derivation and validation data sets (60:40 split) was used to identify predictor variables and validate model performance using area under the receiver operating characteristics (AUC) curve. In our overall sample, 92.2% of patients were alive without ICU admission, 6.2% were admitted to the ICU at least once and survived, and 1.6% died. Greater deciles of age over 50 years, diagnoses of chronic obstructive pulmonary disorder or chronic heart failure, and laboratory abnormalities in alkaline phosphatase, hematocrit, and albumin contributed highest risk score weights for mortality. Risk scores derived from the model discriminated between patients that died versus remained alive without ICU admission (AUC = 0.858), and between ICU survivors versus alive without ICU admission (AUC = 0.765).

Our risk scores provide a feasible and scalable tool for researchers and health systems to identify patient cohorts at increased risk for ICU admission and survivorship. Further studies are needed to prospectively validate the risk scores in other patient populations.

Enhancing critical care efficacy involves evaluating and improving system functioning. Benchmarking, a retrospective comparison of results against standards, aids risk-adjusted assessment and helps healthcare providers identify areas for improvement based on observed and predicted outcomes. The last two decades have seen the development of several models using machine learning (ML) for clinical outcome prediction. ML is a field of artificial intelligence focused on creating algorithms that enable computers to learn from and make predictions or decisions based on data. This narrative review centers on key discoveries and outcomes to aid clinicians and researchers in selecting the optimal methodology for critical care benchmarking using ML.

We used PubMed to search the literature from 2003 to 2023 regarding predictive models utilizing ML for mortality (592 articles), length of stay (143 articles), or mechanical ventilation (195 articles). We supplemented the PubMed search with Google Scholar, making sure relevant articles were included. Given the narrative style, papers in the cohort were manually curated for a comprehensive reader perspective.

Our report presents comparative results for benchmarked outcomes and emphasizes advancements in feature types, preprocessing, model selection, and validation. It showcases instances where ML effectively tackled critical care outcome-prediction challenges, including nonlinear relationships, class imbalances, missing data, and documentation variability, leading to enhanced results.

Although ML has provided novel tools to improve the benchmarking of critical care outcomes, areas that require further research include class imbalance, fairness, improved calibration, generalizability, and long-term validation of published models.

The field of critical care-related artificial intelligence (AI) research is rapidly gaining interest. However, there is still a lack of comprehensive bibliometric studies that measure and analyze scientific publications on a global scale. Network charts have traditionally been used to highlight author collaborations and coword phenomena (ACCP). It is necessary to determine whether chord network charts (CNCs) can provide a better understanding of ACCP, thus requiring clarification. This study aimed to achieve 2 objectives: evaluate global research trends in AI in intensive care medicine on publication outputs, coauthorships between nations, citations, and co-occurrences of keywords; and demonstrate the use of CNCs for ACCP in bibliometric analysis.

The web of science database was searched for a total of 1992 documents published between 2013 and 2022. The document type was limited to articles and article reviews, and titles and abstracts were screened for eligibility. The characteristics of the publications, including preferred journals, leading research countries, international collaborations, top institutions, and major keywords, were analyzed using the category-journal rank-authorship-L-index score and trend analysis. The 100 most highly cited articles are also listed in detail.

Between 2018 and 2022, there was a sharp increase in publications, which accounted for 92.8% (1849/1992) of all papers included in the study. The United States and China were responsible for nearly 50% (936/1992) of the total publications. The leading countries, institutes, departments, authors, and journals in terms of publications were the US, Massachusetts Gen Hosp (US), Medical School, Zhongheng Zhang (China), and Science Reports. The top 3 primary keywords denoting research hotspots for AI in critically ill patients were mortality, model, and intensive care unit, with mortality having the highest burst strength (4.49). The keywords risk and system showed the highest growth trend (0.98) in counts over the past 4 years.

This study provides valuable insights into the potential for ACCP and future research opportunities. For AI-based clinical research to become widely accepted in critical care practice, collaborative research efforts are necessary to strengthen the maturity and robustness of AI-driven models using CNCs for display.

Common tasks encountered in epidemiology, including disease incidence estimation and causal inference, rely on predictive modelling. Constructing a predictive model can be thought of as learning a prediction function (a function that takes as input covariate data and outputs a predicted value). Many strategies for learning prediction functions from data (learners) are available, from parametric regressions to machine learning algorithms. It can be challenging to choose a learner, as it is impossible to know in advance which one is the most suitable for a particular dataset and prediction task. The super learner (SL) is an algorithm that alleviates concerns over selecting the one 'right' learner by providing the freedom to consider many, such as those recommended by collaborators, used in related research or specified by subject-matter experts. Also known as stacking, SL is an entirely prespecified and flexible approach for predictive modelling. To ensure the SL is well specified for learning the desired prediction function, the analyst does need to make a few important choices. In this educational article, we provide step-by-step guidelines for making these decisions, walking the reader through each of them and providing intuition along the way. In doing so, we aim to empower the analyst to tailor the SL specification to their prediction task, thereby ensuring their SL performs as well as possible. A flowchart provides a concise, easy-to-follow summary of key suggestions and heuristics, based on our accumulated experience and guided by SL optimality theory.

Conventional severity-of-illness scoring systems have shown suboptimal performance for predicting in-intensive care unit (ICU) mortality in patients with severe pneumonia. This study aimed to develop and validate machine learning (ML) models for mortality prediction in patients with severe pneumonia. This retrospective study evaluated patients admitted to the ICU for severe pneumonia between January 2016 and December 2021. The predictive performance was analyzed by comparing the area under the receiver operating characteristic curve (AU-ROC) of ML models to that of conventional severity-of-illness scoring systems. Three ML models were evaluated: (1) logistic regression with L2 regularization, (2) gradient-boosted decision tree (LightGBM), and (3) multilayer perceptron (MLP). Among the 816 pneumonia patients included, 223 (27.3%) patients died. All ML models significantly outperformed the Simplified Acute Physiology Score II (AU-ROC: 0.650 [0.584-0.716] vs 0.820 [0.771-0.869] for logistic regression vs 0.827 [0.777-0.876] for LightGBM 0.838 [0.791-0.884] for MLP; P < 0.001). In the analysis for NRI, the LightGBM and MLP models showed superior reclassification compared with the logistic regression model in predicting in-ICU mortality in all length of stay in the ICU subgroups; all age subgroups; all subgroups with any APACHE II score, PaO₂/FiO₂ ratio < 200; all subgroups with or without history of respiratory disease; with or without history of CVA or dementia; treatment with mechanical ventilation, and use of inotropic agents. In conclusion, the ML models have excellent performance in predicting in-ICU mortality in patients with severe pneumonia. Moreover, this study highlights the potential advantages of selecting individual ML models for predicting in-ICU mortality in different subgroups.

When clinicians assess the prognosis of patients in intensive care, they take imaging and non-imaging data into account. In contrast, many traditional machine learning models rely on only one of these modalities, limiting their potential in medical applications. This work proposes and evaluates a transformer-based neural network as a novel AI architecture that integrates multimodal patient data, i.e., imaging data (chest radiographs) and non-imaging data (clinical data). We evaluate the performance of our model in a retrospective study with 6,125 patients in intensive care. We show that the combined model (area under the receiver operating characteristic curve [AUROC] of 0.863) is superior to the radiographs-only model (AUROC = 0.811, p < 0.001) and the clinical data-only model (AUROC = 0.785, p < 0.001) when tasked with predicting in-hospital survival per patient. Furthermore, we demonstrate that our proposed model is robust in cases where not all (clinical) data points are available.