Development and external validation of a nomogram for predicting postoperative pneumonia in aneurysmal subarachnoid hemorrhage

Interpretable machine learning for predicting optimal surgical timing in polytrauma patients with TBI and fractures to reduce postoperative infection risk

This retrospective study leverages machine learning to determine the optimal timing for fracture reconstruction surgery in polytrauma patients, focusing on those with concomitant traumatic brain injury. The analysis included 218 patients admitted to Qilu Hospital of Shandong University from July 2011 to April 2024. Demographic data, physiological status, and non-invasive test indicators were collected. Feature selection via the Boruta and LASSO algorithms preceded the construction of predictive models using Random Forest, Decision Tree, K-Nearest Neighbors, Support Vector Machine, LightGBM, and XGBoost. The Random Forest model excelled in the training set, with an AUC-ROC of 0.828 and accuracy of 0.745, and sustained high performance in the validation set (AUC-ROC: 0.840; Accuracy: 0.813). The final model was informed by eight critical factors, including the Glasgow Coma Scale score, calcium levels, D-dimer, hemoglobin, platelet count, LDL-cholesterol, prothrombin time-international normalized ratio, and prior surgeries. SHAP and LIME algorithms were utilized for model interpretation, elucidating the importance and predictive thresholds of the variables. The application of machine learning in this study provided precise predictions for optimal surgical conditions and timing in polytrauma patients with traumatic brain injury and fractures. This study's findings provide a foundation for personalized surgical planning, potentially reducing postoperative infections and improving patient prognoses.

Association between estimated pulse wave velocity and the risk of mortality in patients with subarachnoid hemorrhage: a retrospective cohort study based on the MIMIC database

BACKGROUND

The estimated pulse wave velocity (ePWV) is a recently developed, simple and useful tool to measure arterial stiffness and to predict long-term cardiovascular mortality. However, the association of ePWV with mortality risk in patients with subarachnoid hemorrhage (SAH) is unclear. Herein, this study aims to assess the potential prediction value of ePWV on short- and long-term mortality of SAH patients.

METHODS

Data of adult patients with no traumatic SAH were extracted from the Medical Information Mart for Intensive Care (MIMIC) III and IV database in this retrospective cohort study. Weighted univariate and multivariable Cox regression analyses were used to explore the associations of ePWV levels with 30-day mortality and 1-year mortality in SAH patients. The evaluation indexes were hazard ratios (HRs) and 95% confidence intervals (CIs). In addition, subgroup analyses of age, the sequential organ failure assessment (SOFA) score, surgery, atrial fibrillation (AF), renal failure (RF), hepatic diseases, chronic obstructive pulmonary disease (COPD), sepsis, hypertension, and diabetes mellitus (DM) were also performed.

RESULTS

Among 1,481 eligible patients, 339 died within 30 days and 435 died within 1 year. After adjusting for covariates, ePWV ≥ 12.10 was associated with higher risk of both 30-day mortality (HR = 1.77, 95%CI: 1.17-2.67) and 1-year mortality (HR = 1.97, 95%CI: 1.36-2.85), compared to ePWV < 10.12. The receiver operator characteristic (ROC) curves showed that compared to single SOFA score, ePWV combined with SOFA score had a relative superior predictive performance on both 30-day mortality and 1-year mortality, with the area under the curves (AUCs) of 0.740 vs. 0.664 and 0.754 vs. 0.658. This positive relationship between ePWV and mortality risk was also found in age ≥ 65 years old, SOFA score < 2, non-surgery, non-hepatic diseases, non-COPD, non-hypertension, non-DM, and sepsis subgroups.

CONCLUSION

Baseline ePWV level may have potential prediction value on short- and long-term mortality in SAH patients. However, the application of ePWV in SAH prognosis needs further clarification.

Development and validation of a machine-learning model for predicting postoperative pneumonia in aneurysmal subarachnoid hemorrhage

Pneumonia is a common postoperative complication in patients with aneurysmal subarachnoid hemorrhage (aSAH), which is associated with poor prognosis and increased mortality. The aim of this study was to develop a predictive model for postoperative pneumonia (POP) in patients with aSAH. A retrospective analysis was conducted on 308 patients with aSAH who underwent surgery at the Neurosurgery Department of the First Affiliated Hospital of Soochow University. Univariate and multivariate logistic regression and lasso regression analysis were used to analyze the risk factors for POP. Receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) were used to evaluate the constructed model. Finally, the effectiveness of modeling these six variables in different machine learning methods was investigated. In our patient cohort, 23.4% (n = 72/308) of patients experienced POP. Univariate, multivariate logistic regression analysis and lasso regression analysis revealed age, Hunt-Hess grade, mechanical ventilation, leukocyte count, lymphocyte count, and platelet count as independent risk factors for POP. Subsequently, these six factors were used to build the final model. We found that age, Hunt-Hess grade, mechanical ventilation, leukocyte count, lymphocyte count, and platelet count were independent risk factors for POP in patients with aSAH. Through validation and comparison with other studies and machine learning models, our novel predictive model has demonstrated high efficacy in effectively predicting the likelihood of pneumonia during the hospitalization of aSAH patients.

New grading scale based on early factors for predicting delayed cerebral ischemia in patients with aneurysmal subarachnoid hemorrhage: a multicenter retrospective study

Delayed cerebral ischemia (DCI) could lead to poor clinical outcome(s). The aim of the present study was to establish and validate a predictive model for DCI after aneurysmal subarachnoid hemorrhage (aSAH) based on clinical data. Data from a series of 217 consecutive patients with aSAH were reviewed and analyzed. Related risk factors within 72 h after aSAH were analyzed depending on whether DCI recurred. Least absolute shrinkage and selection operator (LASSO) analysis was performed to reduce data dimensions and screen for optimal predictors. Multivariable logistic regression was used to establish a predictive model and construct a nomogram. Receiver operating characteristic (ROC) and calibration curves were generated to assess the discriminative ability and goodness of fit of the model. Decision curve analysis was applied to evaluated the clinical applicability of the predictive model. LASSO regression identified 4 independent predictors, including Subarachnoid Hemorrhage Early Brain Edema Score (i.e., "SEBES"), World Federation of Neurosurgical Societies scale score (i.e., "WFNS"), modified Fisher Scale score, and intraventricular hemorrhage (IVH), which were incorporated into logistic regression to develop a nomogram. After verification, the area under the ROC curve for the model was 0.860. The calibration curve indicated that the predictive probability of the new model was in good agreement with the actual probability, and decision curve analysis demonstrated the clinical applicability of the model within a specified range. The prediction model could precisely calculate the probability of DCI after aSAH, and may contribute to better clinical decision-making and personalized treatment to achieve better outcomes.

Development and performance assessment of novel machine learning models for predicting postoperative pneumonia in aneurysmal subarachnoid hemorrhage patients: external validation in MIMIC-IV

Background

Postoperative pneumonia (POP) is one of the primary complications after aneurysmal subarachnoid hemorrhage (aSAH) and is associated with postoperative mortality, extended hospital stay, and increased medical fee. Early identification of pneumonia and more aggressive treatment can improve patient outcomes. We aimed to develop a model to predict POP in aSAH patients using machine learning (ML) methods.

Methods

This internal cohort study included 706 patients with aSAH undergoing intracranial aneurysm embolization or aneurysm clipping. The cohort was randomly split into a train set (80%) and a testing set (20%). Perioperative information was collected from participants to establish 6 machine learning models for predicting POP after surgical treatment. The area under the receiver operating characteristic curve (AUC), precision-recall curve were used to assess the accuracy, discriminative power, and clinical validity of the predictions. The final model was validated using an external validation set of 97 samples from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database.

Results

In this study, 15.01% of patients in the training set and 12.06% in the testing set with POP after underwent surgery. Multivariate logistic regression analysis showed that mechanical ventilation time (MVT), Glasgow Coma Scale (GCS), Smoking history, albumin level, neutrophil-to-albumin Ratio (NAR), c-reactive protein (CRP)-to-albumin ratio (CAR) were independent predictors of POP. The logistic regression (LR) model presented significantly better predictive performance (AUC: 0.91) than other models and also performed well in the external validation set (AUC: 0.89).

Conclusion

A machine learning model for predicting POP in aSAH patients was successfully developed using a machine learning algorithm based on six perioperative variables, which could guide high-risk POP patients to take appropriate preventive measures.