Predicting Intensive Care Unit Admission Research Articles

Differentiating Kawasaki Disease and Multisystem Inflammatory Syndrome in Children Using Blood Composite Scores: Insights into Clinical Outcomes and Predictive Indices.

The study sought to assess the clinical utility of complete blood count-derived composite scores, suggesting their potential as markers of inflammation and disease severity in Kawasaki disease (KD) and multisystem inflammatory syndrome in children (MIS-C) with Kawasaki-like features. This retrospective study analyzed data from 71 KD and 73 MIS-C patients and 70 healthy controls. The KD group showed a higher rate of coronary involvement (26.7% vs. 10.9%), while the MIS-C group had a higher intensive care unit (ICU) admission rate (34.2% vs. 2.8%). Platelet counts, lymphocyte counts, mean platelet volume (MPV), MPV/Lymphocyte (MPVLR), and MPV/Platelet (MPVPR) ratios demonstrated the highest specificities in distinguishing MIS-C than KD (84.5%, 83.1%, 91.1%, 88.7%, and 88.7%, respectively). Monocyte counts, MPV, and MPVPR demonstrated the highest specificities to predictive ICU admission in the MIS-C group (83.3%, 89.6%, and 89.6%, respectively). Lymphocyte counts, platelet/lymphocyte ratio (PLR), neutrophil/lymphocyte ratio (NLR), MPVLR, and Systemic Immune-Inflammation Index (SII) parameters were found to have high negative predictive values for predicting KD patients without coronary artery lesions (CALs) (85.7%, 86.1%, 87.1%, 87.1%, and 85.7%, respectively)., Systemic Inflammation Response Index (SIRI), MPVPR, and CRP were independently predictive of ICU admission in the MIS-C group, and lymphocyte count and IVIG resistance were also identified as significant predictors of CALs in the KD group. NLR, MPVLR, MPVPR, and NPR indices effectively differentiate MIS-C from KD and predict ICU admission in MIS-C. NLR, PLR, MPVLR, and SII are valuable in excluding CALs in KD with high negative predictive values. In addition, SIRI and MPVLR were independent predictors of ICU admission in MIS-C, and lymphocyte count was identified as an independent predictor of CALs in KD.

Predicting the risk of intensive care unit admission in patients with COVID-19 presenting in the emergency room: Development and evaluation of the CROSS score.

Existing risk evaluation tools underperform in predicting intensive care unit (ICU) admission for patients with the Coronavirus Disease 2019 (COVID-19). This study aimed to develop and evaluate an accurate and calculator-free clinical tool for predicting ICU admission at emergency room (ER) presentation. Data from patients with COVID-19 in a nationwide German cohort (March 2020-January 2023) were analyzed. Candidate predictors were selected based on literature and clinical expertise. A risk score, predicting ICU admission within seven days of ER presentation, was developed using elastic net logistic regression on a northern German cohort (derivation cohort), evaluated on a southern German cohort (evaluation cohort) and externally validated on a Colombian cohort. Performance was evaluated through discrimination, calibration, and clinical utility against existing tools. ICU admission rates within seven days were 30.8% (derivation cohort, n=1295, median age 60, 38.1% female), 28.1% (evaluation cohort, n=1123, median age 58, 36.9% female), and 30.3% (Colombian cohort, n=780, median age 57, 38.8% female). The 11-point CROSS score, based on Confusion, Respiratory rate, Oxygen Saturation (with or without concurrent supplemental oxygen), and oxygen Supplementation, demonstrated good discrimination (area under the curve (AUC): 0.77 in the evaluation cohort; 0.69 in the Colombian cohort), good calibration, and superior clinical utility compared to existing tools. Mortality-predicting tools performed poorly in predicting ICU admission risk for patients with COVID-19. The calculator-free CROSS score effectively predicts ICU admission for patients with COVID-19 in the ER. Further studies are needed to assess its generalizability in other settings. Mortality-predicting tools are not recommended for ICU admission prediction.

Estimation of blood lactate and bicarbonate levels after stored blood transfusion to predict ICU admission in patients undergoing major head and neck surgeries: A prospective observational study

Abstract Background and Aims: Major head neck surgeries are often associated with major blood loss requiring blood transfusion. However, in spite of transfusion, patients usually suffer adverse postoperative outcomes. Biomarkers can help in identifying such events early. This observational study was conducted to compare blood lactate and bicarbonate levels as predictors of adverse postoperative outcomes. Materials and Methods: Forty-eight adult American Society of Anesthesiologists Physical Status I–III patients met the inclusion criteria. Intraoperative blood loss was managed with stored blood transfusion as per transfusion trigger. Blood lactate and bicarbonate levels were measured preoperatively (Tbas), at the immediate postoperative period (T0), and at 8 h (T8), 16 h (T16), and 24 h (T24) postoperatively. Outcomes such as need for intensive care unit (ICU) admission, length of ICU stay, intraoperative blood transfusion, re-exploration rate, and mortality were recorded. Results: Blood transfusions and ICU admissions were required in 19 (39.6%) and 24 (50%) patients, respectively. Lactate levels of patients requiring blood transfusion and admission to ICU rose significantly from their baseline (1.30 ± 0.41 mmol/l) to 2.80 ± 1.14 mmol/l at the immediate postoperative period, which fell to 2.06 ± 0.78 mmol/l at 24 h postoperatively, compared to other patients who did not require transfusion and ICU admission (P < 0.001). The bicarbonate value did not show any significant change from its baseline (22.68 ± 1.83 mEq/l) at all time points (P = 0.8). In addition, no significant difference was noted regarding ICU admissions (P = 0.659) or blood transfusions (P = 0.788). Conclusions: Following major head and neck surgeries, blood lactate level is a good predictor, but bicarbonate is a poor predictor of the need for blood transfusions and ICU admission. Intraoperative blood transfusion failed to prevent rise in blood lactate level, which is taken as a surrogate marker of tissue hypoxia.

Predicting Intensive Care Unit Admission in COVID-19-Infected Pregnant Women Using Machine Learning.

Background: The rapid onset of COVID-19 placed immense strain on many already overstretched healthcare systems. The unique physiological changes in pregnancy, amplified by the complex effects of COVID-19 in pregnant women, rendered prioritization of infected expectant mothers more challenging. This work aims to use state-of-the-art machine learning techniques to predict whether a COVID-19-infected pregnant woman will be admitted to ICU (Intensive Care Unit). Methods: A retrospective study using data from COVID-19-infected women admitted to one hospital in Astana and one in Shymkent, Kazakhstan, from May to July 2021. The developed machine learning platform implements and compares the performance of eight binary classifiers, including Gaussian naïve Bayes, K-nearest neighbors, logistic regression with L2 regularization, random forest, AdaBoost, gradient boosting, eXtreme gradient boosting, and linear discriminant analysis. Results: Data from 1292 pregnant women with COVID-19 were analyzed. Of them, 10.4% were admitted to ICU. Logistic regression with L2 regularization achieved the highest F1-score during the model selection phase while achieving an AUC of 0.84 on the test set during the evaluation stage. Furthermore, the feature importance analysis conducted by calculating Shapley Additive Explanation values points to leucocyte counts, C-reactive protein, pregnancy week, and eGFR and hemoglobin as the most important features for predicting ICU admission. Conclusions: The predictive model obtained here may be an efficient support tool for prioritizing care of COVID-19-infected pregnant women in clinical practice.

Factors associated with intensive care unit admission due to diabetic ketoacidosis in adults: a validated predictive model.

The objective of this study was to develop a predictive model capable of determining the need for intensive care unit (ICU) admission of patients with diabetic ketoacidosis (DKA) during their assessment in the Emergency Department. This is an observational study of consecutive cases including all adult patients diagnosed with DKA at a tertiary hospital between 2010 and 2024. Variables from medical history, physical examination, and laboratory tests at admission were collected and studied for their association with ICU admission. The sample was divided into two randomized parts: one to build a logistic regression model and another to validate it. Two hundred and thirty-one DKA events were included. Individuals had a mean age of 49.6 ± 19.9years and 50.2% were male. Forty-eight point five percent of cases required ICU admission, and 30-day mortality was 4.8%. The best model to predict ICU admission included Glasgow Coma Scale (odds ratio [OR] = 0.64, p = 0.003), pH (OR = 0.0088, p = 0.005), bilirubin (OR = 0.13, p = 0.036), bicarbonate (OR = 0.0091, p = 0.013), and pH-bicarbonate interaction (OR = 3.78, p = 0.015). The model had an R2 of 0.561, and the area under the curve (AUC) in the validation cohort was 0.842. Internal validation by bootstrap resampling showed an AUC = 0.871. Variables associated with the severity of acidosis in patients with DKA predict the need for ICU admission better and earlier than other clinical variables.

Predicting ICU Admission for Hospitalized COVID-19 Patients by Artificial Neural Network Combined with Elastic Net

Background: Machine learning models could assist physicians in identifying high-risk COVID-19 patients. This study aimed to predict the Intensive Care Unit (ICU) admission in COVID-19 hospitalized patients by the Artificial Neural Network (ANN) model combined with Elastic Net algorithm. Methods: In this prospective study, the data of 139 COVID-19 patients admitted to Imam Reza Hospital in Tabriz between 20 March and 5 April 2020, were analyzed. The Elastic Net method was used to choose features with high importance. The chosen variables were standardized and ANN was fitted to the data with one hidden layer based on the descending gradient algorithm. To validate the model, the training and test group method with a ratio of 70 to 30 was used. The model’s predictive power was reported by calculating the overall accuracy, sensitivity, specificity, and Area Under the ROC curve (AUC). Results: According to the results of the Elastic Net, the ANN model was constructed based on age, sex, body mass index, diabetes mellitus status, history of heart disease, and vital signs including systolic and diastolic blood pressure, saturated oxygen level, pulse rate, respiration rate, and body temperature. The overall accuracy of this model was 93.15%, sensitivity 80%, specificity 95.8%, and AUC 0.90. Saturated oxygen level, pulse rate, and age were the most important and predictive variables. Conclusion: In the investigated sample of patients admitted with COVID-19, the fitted ANN model had acceptable performance to predict the ICU admission. This finding could be useful for physicians and policy makers.

Serum lactate is associated with an ICU admission in patients presenting with seizure at the emergency department.

Patients presenting with seizure at the emergency department (ED) may have morbidity and mortality. Serum lactate has been shown to be a predictor of diagnosis of seizure versus syncope and mortality in patients presenting with seizure at the ED. There is limited data on using serum lactate as a predictor of the intensive care unit (ICU) admission in patients presenting with seizure at the ED. This study aimed to evaluate if serum lactate can be a predictor of the ICU admission in this setting. This was a retrospective cohort study enrolled adult patients aged 18years or more, diagnosed with seizures or convulsive status epilepticus at the ED, and who had undergone testing for serum venous lactate. Clinical factors predictive of ICU admission were computed using logistic regression analysis. Of the 288 patients who met the study criteria, 45 patients (15.63%) required ICU admission. Two factors were independently associated with the ICU admission: convulsive status epilepticus and serum lactate. Both factors had adjusted odds ratio (95% confidence interval) of 4.793 (2.119, 10.844) and 1.008 (1.001, 1.016), respectively. Serum lactate of over 63mg/dL had specificity of 80.25% and sensitivity of 35.56% with the area under ROC curve of 63.30% (95% confidence interval of 54.77%, 71.84%). Serum lactate and convulsive status epilepticus were two predictors of ICU admission in patients presenting with seizure at the ED.

D-dimer as a Predictor of ICU Admission and Mortality in COVID-19 Patients: Insights From a Two-Year Retrospective Study From a Tertiary Care Center in South India.

Introduction Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-induced pneumonitis results in a prothrombotic and hypercoagulable state. Prognostic indicators are crucial for identifying patients at risk of complications. D-dimer, a degradation product of cross-linked fibrin, is a specific marker for thrombosis. Elevated D-dimer levels have been strongly correlated with poor prognosis and increased severity of illness in COVID-19 patients. Given D-dimer's eight-hour half-life, periodic measurement is necessary to track disease progression. This study aimed to analyze and derive threshold and peak D-dimer values to predict outcomes in COVID-19 patients, comparing those treated in isolation wards to those requiring intensive care. Methods This two-year retrospective observational study included patients above 18 years with confirmed COVID-19. Patients were categorized into those treated in isolation wards and those admitted to the intensive care unit (ICU). Based on the outcome, they were further divided into survivors and non-survivors. Demographic and outcome-related data were collected from the hospital's laboratory information system. Serial D-dimer measurements were taken at eight time points. Statistical analysis was performed using the Mann-Whitney test for laboratory values and the chi-square test for demographic data. Receiver operating characteristic (ROC) curve analysis was utilized to derive critical D-dimer values. The area under the curve (AUC) was calculated for initial and peak D-dimer values. Results Of 2.149 patients with confirmed COVID-19, 811 (38%) presented with elevated D-dimer levels. ICU admission was required for 239 patients, either due to direct admission or worsening conditions. An initial D-dimer value of ≥0.93 mg/L FEU indicated the need for ICU admission, while a peak D-dimer value of 5.65 mg/L FEU predicted mortality. The AUC for the initial D-dimer was 0.60 (95% CI: 0.55-0.64), indicating moderate discriminatory power. The AUC for the peak D-dimer was 0.58 (95% CI: 0.54-0.62), suggesting lower predictive accuracy for peak values. Sensitivity was high for both initial (0.925) and peak (0.960) D-dimer values, although specificity was lower, especially for the peak D-dimer (0.486), resulting in a higher rate of false positives. Among the ICU patients, the age range was 27-97 years, with a mean of 53.5 years. Males were more affected than females (71% vs. 29%), with a male-to-female ratio of 1.4:1. Of the ICU patients, 64.8% recovered, while 35.2% succumbed to the disease. Younger patients (mean age: 50.5 ± 12 years) recovered faster than older patients (mean age: 64 ± 16 years), with a significant difference in recovery time (p < 0.001). Gender did not significantly impact outcomes (p = 0.743). Survivors spent less time in the ICU (3-7 days) compared to non-survivors (4-14 days) (p = 0.041). Conclusion Serial D-dimer monitoring is essential for predicting outcomes and guiding treatment in COVID-19 patients. Initial and peak D-dimer values can help identify patients requiring intensive care and those at risk of mortality, allowing for timely interventions. D-dimer levels should be integrated into routine clinical assessments for managing COVID-19 patients.

NEWS, NEWS2, and qSOFA accuracy in predicting sepsis-related mortality in acute myeloid leukemia: a retrospective single-center analysis.

Acute myeloid leukemia (AML) treated with intensive chemotherapy carries a high risk of severe infection. The development of reliable assessment tools to promptly identify patients at risk of developing critical illness is essential to prevent delays in intensive care unit (ICU) admission. This study evaluated the accuracy of quick Sequential Organ Failure Assessment (qSOFA) score, National Early Warning Score (NEWS), and NEWS2 score in predicting ICU admission and sepsis-related mortality in this population. A retrospective analysis was conducted, including 365 episodes of febrile neutropenia in 126 patients. The results showed that all three scores-qSOFA, NEWS, and NEWS2-demonstrated good accuracy for all outcomes, with area under the receiver-operating characteristic curve values for sepsis-related mortality of 0.812, 0.858, and 0.848, respectively. In addition, the scores exhibited excellent accuracy in predicting ICU admission and the composite outcome of ICU admission or sepsis-related mortality. To our knowledge, this is the first study to evaluate the accuracy of NEWS in a population of patients with AML who did not undergo stem cell transplantation. These findings suggest that NEWS and NEWS2 are effective tools for identifying patients with AML at high risk of clinical deterioration during febrile neutropenia, supporting their use in clinical practice.

Development and Validation of a Risk Score for Predicting ICU Admission in Adults with New-Onset Encephalitis.

Timely intensive care unit (ICU) admission for patients with encephalitis is associated with better prognosis. Therefore, our aim was to create a risk score predicting ICU admission in adults with encephalitis, which could aid in optimal management and resource allocation. We initially identified variables that would be most predictive of ICU admission among 372 patients with encephalitis from two hospital systems in Houston, Texas (cohort 1), who met the International Encephalitis Consortium (IEC) criteria from 2005 to 2023. Subsequently, we used a binary logistic regression model to create a risk score for ICU admission, which we then validated externally using a separate cohort of patients from two hospitals in Baltimore, Maryland (cohort 2), who met the IEC criteria from 2006 to 2022. Of 634 patients with encephalitis, 255 (40%) were admitted to the ICU, including 45 of 113 (39.8%) patients with an autoimmune cause, 100 of 272 (36.7%) with an infectious cause, and 110 of 249 (44.1%) with an unknown cause (p = 0.225). After conducting a multivariate analysis in cohort 1, we found that the presence of focal neurological signs, new-onset seizure, a Full Outline of Unresponsiveness score ≤ 14, leukocytosis, and a history of chronic kidney disease at admission were associated with an increased risk of ICU admission. The resultant clinical score for predicting ICU admission had an area under the receiver operating characteristic curve (AUROC) of 0.77 (95% confidence interval [CI] 0.72-0.82, p < 0.001). Patients were classified into three risk categories for ICU admission: low risk (score 0, 12.5%), intermediate risk (scores 1-5, 49.5%), and high risk (scores 6-8, 87.5%). External validation in cohort 2 yielded an AUROC of 0.76 (95% CI 0.69-0.83, p < 0.001). ICU admission is common in patients with encephalitis, regardless of etiology. Our risk score, encompassing neurologic and systemic factors, may aid physicians in decisions regarding intensity of care for adult patients with encephalitis upon hospital admission.

Lactate to albumin ratio as a prognosis predictor in gastrointestinal bleeding in the emergency department.

Gastrointestinal bleeding (GIB) is a common cause of emergency department (ED) visits, and has various prognoses. This study aimed to verify the prognostic ability of the lactate/albumin (L/A) ratio in patients with GIB compared with the AIMS65 score and blood urea nitrogen/albumin (B/A) ratio. This retrospective study included patients with GIB symptoms who visited the ED in 2019. Baseline characteristics and laboratory data were obtained to calculate the L/A and B/A ratios and AIMS65 score. Each score was evaluated as a predictor of intensive care unit (ICU) admission and in-hospital mortality by using the area under the receiver operating characteristic (AUROC) curve. The multivariate logistic regression analysis revealed that the L/A ratio significantly predicted ICU admission and in-hospital mortality. The AUROC curve for predicting ICU admission were 0.788, 0.695, and 0.586 for the L/A, B/A, and AIMS65 scores, respectively, while the AUROC curve for predicting in-hospital mortality were 0.807, 0.799, and 0.683 for the L/A, B/A, and AIMS65 scores, respectively. The L/A ratio, which consists of the serum lactate and albumin levels, showed superior performance relative to the B/A ratio and AIMS65 score in predicting the prognosis of patients with GIB.

Hematological Markers as Predictors of ICU Admission in COVID-19 Patients: A Case-Control Study From a Tertiary Hospital.

COVID-19 illness severity ranges from mild- to life-threatening cases necessitating critical care. Rapid prediction of disease severity and the need for critical care support in COVID-19 patients remain essential, not only for current management but also for preparedness in future pandemics. This study aimed to assess hematological parameters as predictors of intensive care unit (ICU) admission and survival in COVID-19 patients, providing insights applicable to a broad range of infectious diseases. A case-control study was conducted at Hospital Raja Perempuan Zainab II, a tertiary referral hospital in Kelantan, Malaysia, from March 2020 to August 2021. Demographics, clinical, and laboratory data were retrieved from patients' medical records. Statistical analyses, including the Chi-square (χ2) test, independent t-tests, and simple and multiple logistic regressions, were used to analyze the data. A receiver operating characteristic (ROC) curve analysis was conducted to assess the accuracy of the predictors. The median age was 51 years, with females comprising 56.7% (n=148) and males 43.3% (n=113). A total of 88.5% of patients were admitted to non-ICU wards, with a mortality rate of 5.7%. Significant differences were observed in the distribution of hematological parameters between ICU-admitted and non-admitted patients. Neutrophil (OR: 23.96, 95% CI: 7.296-78.675) and white blood cell (WBC) count (OR: 36.677, 95% CI: 2.086-644.889) were the most significant predictors for ICU admission and survival, respectively. WBC and neutrophil counts exhibited high predictive value for ICU admission, while WBC, neutrophil, lymphocyte, and immature granulocyte (IG) counts were significant predictors of survival status among COVID-19 patients. These findings underscore the continued relevance of hematological markers in managing severe respiratory infections and improving critical care triage, with implications for current and future healthcare challenges.

An explainable machine learning-based model to predict intensive care unit admission among patients with community-acquired pneumonia and connective tissue disease

BackgroundThere is no individualized prediction model for intensive care unit (ICU) admission on patients with community-acquired pneumonia (CAP) and connective tissue disease (CTD) so far. In this study, we aimed to establish a machine learning-based model for predicting the need for ICU admission among those patients.MethodsThis was a retrospective study on patients admitted into a University Hospital in China between November 2008 and November 2021. Patients were included if they were diagnosed with CAP and CTD during admission and hospitalization. Data related to demographics, CTD types, comorbidities, vital signs and laboratory results during the first 24 h of hospitalization were collected. The baseline variables were screened to identify potential predictors via three methods, including univariate analysis, least absolute shrinkage and selection operator (Lasso) regression and Boruta algorithm. Nine supervised machine learning algorithms were used to build prediction models. We evaluated the performances of differentiation, calibration, and clinical utility of all models to determine the optimal model. The Shapley Additive Explanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) techniques were performed to interpret the optimal model.ResultsThe included patients were randomly divided into the training set (1070 patients) and the testing set (459 patients) at a ratio of 70:30. The intersection results of three feature selection approaches yielded 16 predictors. The eXtreme gradient boosting (XGBoost) model achieved the highest area under the receiver operating characteristic curve (AUC) (0.941) and accuracy (0.913) among various models. The calibration curve and decision curve analysis (DCA) both suggested that the XGBoost model outperformed other models. The SHAP summary plots illustrated the top 6 features with the greatest importance, including higher N-terminal pro-B-type natriuretic peptide (NT-proBNP) and C-reactive protein (CRP), lower level of CD4 + T cell, lymphocyte and serum sodium, and positive serum (1,3)-β-D-glucan test (G test).ConclusionWe successfully developed, evaluated and explained a machine learning-based model for predicting ICU admission in patients with CAP and CTD. The XGBoost model could be clinical referenced after external validation and improvement.

Validation of the Korean Version of the Clinical Frailty Scale-Adjusted Korean Triage and Acuity Scale for Older Patients in the Emergency Department.

Background and Objectives: The Clinical Frailty Scale (CFS), used to screen for prehospital frailty in patients aged >65 years, is simple, time-efficient, and has been validated in emergency departments (EDs). In this study, we analyzed whether the Korean Triage and Acuity Scale (KTAS) classification by level in older patients determined to have frailty based on the Korean version of the CFS increases the triage performance of the current KTAS. Materials and Methods: The primary outcome was 30-day in-hospital mortality, and secondary outcomes were hospital and intensive care unit (ICU) admissions. This study retrospectively analyzed prospectively collected data from three ED centers. Patients with a CFS score ranging from five (mildly frail) to nine (terminally ill) were categorized into the frailty group. We upgraded the KTAS classification of the frailty group by one level of urgency and defined this as the CFS-KTAS. Results: The cutoff values for predicting admission were three and two for the KTAS and CFS-KTAS, respectively. A significant difference was observed in the area under the receiver operating characteristic (AUROC) curve between the KTAS and CFS-KTAS. To predict ICU admission, the cutoff score was two for both scales. A significant difference was observed in the AUROC curve between the KTAS and CFS-KTAS. For predicting in-hospital mortality, the cutoff score was two for both scales. A significant difference was observed in the AUROC curve between the KTAS and CFS-KTAS. Conclusions: This study showed that the CFS-adjusted KTAS has a more useful prognostic value than the KTAS alone for predicting hospital outcomes in older patients.

A machine learning framework for interpretable predictions in patient pathways: The case of predicting ICU admission for patients with symptoms of sepsis

Proactive analysis of patient pathways helps healthcare providers anticipate treatment-related risks, identify outcomes, and allocate resources. Machine learning (ML) can leverage a patient’s complete health history to make informed decisions about future events. However, previous work has mostly relied on so-called black-box models, which are unintelligible to humans, making it difficult for clinicians to apply such models. Our work introduces PatWay-Net, an ML framework designed for interpretable predictions of admission to the intensive care unit (ICU) for patients with symptoms of sepsis. We propose a novel type of recurrent neural network and combine it with multi-layer perceptrons to process the patient pathways and produce predictive yet interpretable results. We demonstrate its utility through a comprehensive dashboard that visualizes patient health trajectories, predictive outcomes, and associated risks. Our evaluation includes both predictive performance – where PatWay-Net outperforms standard models such as decision trees, random forests, and gradient-boosted decision trees – and clinical utility, validated through structured interviews with clinicians. By providing improved predictive accuracy along with interpretable and actionable insights, PatWay-Net serves as a valuable tool for healthcare decision support in the critical case of patients with symptoms of sepsis.

Development of Prediction Model for Intensive Care Unit Admission Based on Heart Rate Variability: A Case-Control Matched Analysis.

This study aimed to develop a predictive model for intensive care unit (ICU) admission by using heart rate variability (HRV) data. This retrospective case-control study used two datasets (emergency department [ED] patients admitted to the ICU, and patients in the operating room without ICU admission) from a single academic tertiary hospital. HRV metrics were measured every 5 min using R-peak-to-R-peak (R-R) intervals. We developed a generalized linear mixed model to predict ICU admission and assessed the area under the receiver operating characteristic curve (AUC). Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated from the coefficients. We analyzed 610 (ICU: 122; non-ICU: 488) patients, and the factors influencing the odds of ICU admission included a history of diabetes mellitus (OR [95% CI]: 3.33 [1.71-6.48]); a higher heart rate (OR [95% CI]: 3.40 [2.97-3.90] per 10-unit increase); a higher root mean square of successive R-R interval differences (RMSSD; OR [95% CI]: 1.36 [1.22-1.51] per 10-unit increase); and a lower standard deviation of R-R intervals (SDRR; OR [95% CI], 0.68 [0.60-0.78] per 10-unit increase). The final model achieved an AUC of 0.947 (95% CI: 0.906-0.987). The developed model effectively predicted ICU admission among a mixed population from the ED and operating room.

Obesity Surgery Mortality Risk Score as a Predictor for Intensive Care Unit Admission in Patients Undergoing Laparoscopic Bariatric Surgery.

Background: Laparoscopic bariatric surgery provides many benefits including lower postoperative pain scores, reduced opioid consumption, shorter hospital stays, and improved quality of recovery. However, the anaesthetic management of obese patients requires caution in determining postoperative risk and in planning adequate postoperative pathways. Currently, there are no specific indications for intensive care unit (ICU) admission in this surgical population and most decisions are made on a case-by-case basis. The aim of this study is to investigate whether Obesity Surgery Mortality Risk Score (OS-MRS) is able to predict ICU admission in patients undergoing laparoscopic bariatric surgery (LBS). Methods: We retrospectively reviewed data of patients who underwent LBS during a 2-year period (2017-2019). The collected data included demographics, comorbidities and surgery-related variables. Postoperative ICU admission was decided via bariatric anaesthesiologists' evaluations, based on the high risk of postoperative cardiac or respiratory complications. Anaesthesia protocol was standardized. Logistic regression was used for statistical analysis. Results: ICU admission was required in 2% (n = 15) of the 763 patients. The intermediate risk group of the OS-MRS was detected in 84% of patients, while the American Society of Anaesthesiologists class III was reported in 80% of patients. A greater OS-MRS (p = 0.01), advanced age (p = 0.04), male gender (p = 0.001), longer duration of surgery (p = 0.0001), increased number of patient comorbidities (p = 0.002), and previous abdominal surgeries (p = 0.003) were predictive factors for ICU admission. Conclusions: ICU admission in obese patients undergoing LBS is predicted by OS-MRS together with age, male gender, number of comorbidities, previous abdominal surgeries, and duration of surgery.

Prediction of Admission to Intensive Care Unit and 1-Year Mortality After Acute Pancreatitis With Walled-Off Pancreatic Necrosis: A Retrospective, Single-Center Cohort Study.

Pancreatic walled-off necrosis (WON) carries significant mortality and morbidity risks, often necessitating intensive care unit (ICU) admission. This retrospective study aimed to evaluate whether routine biochemical parameters at the time of the index endoscopic procedure could predict ICU admission and 1-year mortality following endoscopic treatment of WON. We retrospectively identified 201 consecutive patients who underwent endoscopic drainage for WON between January 1, 2010, and December 31, 2020. Associations between routine biochemical blood tests and outcomes were assessed using logistic regression models. Within 1 year of the index endoscopy, 31 patients (15.4%) died, and 40 (19.9%) were admitted to the ICU due to sepsis. Preoperative electrolyte disturbances were more prevalent among ICU-admitted patients and nonsurvivors. Hyperkalemia, hypoalbuminemia, and elevated urea were significant predictors of 1-year mortality, while hypernatremia, elevated serum creatinine, and hypoalbuminemia predicted ICU admission. Predictive models exhibited good discriminative ability, with an AUC of 0.84 (95% CI,0,75-0.93) for 1-year mortality and 0.86 (95%CI, 0.79-0.92) for ICU admission. Preoperative imbalances in routine blood tests effectively predict adverse outcomes in endoscopically treated WON patients.

Assessing the predictive value of neutrophil percentage to albumin ratio for ICU admission in ischemic stroke patients.

Acute ischemic stroke (AIS) remains a substantial global health challenge, contributing to increased morbidity, disability, and mortality. This study aimed at investigating the predictive value of the neutrophil percentage to albumin ratio (NPAR) in determining intensive care unit (ICU) admission among AIS patients. A retrospective observational study was conducted, involving AIS cases admitted to a tertiary hospital in Jordan between 2015 and 2020. Lab data were collected upon admission, and the primary outcome was ICU admission during hospitalization. Descriptive and inferential analyses were performed using SPSS version 29. In this study involving 364 AIS patients, a subset of 77 (21.2%) required admission to the ICU during their hospital stay, most frequently within the first week of admission. Univariable analysis revealed significantly higher NPAR levels in ICU-admitted ischemic stroke patients compared to those who were not admitted (23.3 vs. 15.7, p < 0.001), and multivariable regression models confirmed that higher NPAR (≥19.107) independently predicted ICU admission in ischemic stroke patients (adjusted odds ratio [aOR] = 4.85, 95% CI: 1.83-12.83). Additionally, lower GCS scores and higher neutrophil-to-lymphocyte ratio (NLR) were also associated with increased likelihood of ICU admission. In terms of predictive performance, NPAR showed the highest accuracy with an AUC of 0.885, sensitivity of 0.805, and specificity of 0.854, using a cutoff value of 19.107. NPAR exhibits an AUC of 0.058, significantly outperforming NLR (Z = 2.782, p = 0.005). NPAR emerged as a robust independent predictor of ICU admission in ischemic stroke patients, surpassing the predictive performance of the NLR.

Shock and modified shock indices in predicting poisoning severity and outcomes in acute aluminum phosphide poisoned patients.

Severe refractory hypotension and cardiogenic shock are the main contributors to death in acute aluminum phosphide (ALP) poisoning. Shock index (SI) and modified shock index (MSI) are easily obtained parameters that reflect shock at an early stage. This study aimed to evaluate the role of SI and MSI in the prediction of the severity and outcomes of acute ALP poisoned patients. This cross sectional study was conducted on patients admitted to Tanta University Poison Control Centre with acute ALP poisoning from April 2022 to March 2023. Socio-demographics and toxicological data were taken, findings of clinical examination and laboratory investigations were recoded, SI was calculated by dividing heart rate over systolic blood pressure, and MSI was obtained by dividing heart rate over mean arterial pressure. Poisoning severity was assessed using poisoning severity score (PSS). Patients were divided into groups according to intensive care unit (ICU) admission and mortality. The study enrolled 94 patients. The median values of SI and MSI were significantly higher in ICU-admitted patients and non-survivors rather than their comparable groups. Significant positive correlations were observed between each of SI and MSI and PSS. At cut-off >1.14, SI conveyed fair performance to predict ICU admission and mortality (AUC=0.710 and 0.739, respectively). Similarly, MSI had fair performance to predict ICU admission (AUC=0.731) and mortality (AUC=0.744) at cut-off >1.47 and >1.5, respectively. Both SI and MSI could be considered simple bedside adjuncts to predict ICU admission and mortality in acute ALP poisoning.