Development and validation of a machine learning-based risk prediction model for stroke-associated pneumonia in older adult hemorrhagic stroke

Predictions aren't needed; PREVENTION IS! GET THERE! I'd have you all fired!

You've known about this problem for a long time. SOLVE IT! 

Just maybe this vaccine!

• Pneumonia Vaccine (3 posts to July 2020)

 

• 11% Stroke-associated pneumonia (2 posts to October 2020)

Is everyone in stroke so blitheringly stupid that they don't realize that you solve and prevent problems? Rather than lazily describing them? Serious question!

Send me hate mail on this: oc1dean@gmail.com. I'll print your complete statement with your name and my response in my blog. Or are you afraid to engage with my stroke-addled mind? Your patients need an explanation of why you aren't trying to get survivors recovered.

Why isn't your 'professional' solving stroke?

Laziness? Incompetence? Or just don't care? NO leadership? NO strategy? Not my job? Not my Problem!

 Development and validation of a machine learning-based risk prediction model for stroke-associated pneumonia in older adult hemorrhagic stroke

Yi Cao^1,2†Haipeng Deng^1†Shaoyun Liu^1†Xi Zeng¹Yangyang Gou²Weiting Zhang²Yixinyuan Li²Hua Yang^1*Min Peng^3*

• ¹Department of Neurosurgery, Affiliated Hospital of Guizhou Medical University, Guiyang, China
• ²School of Nursing, Guizhou Medical University, Guiyang, China
• ³Department of Nursing Quality Management, Affiliated Hospital of Guizhou Medical University, Guiyang, China

Objective: To develop and validate a machine learning (ML)-based model for predicting stroke-associated pneumonia (SAP) risk in older adult hemorrhagic stroke patients.

Methods: A retrospective collection of older adult hemorrhagic stroke patients from three tertiary hospitals in Guiyang, Guizhou Province (January 2019–December 2022) formed the modeling cohort, randomly split into training and internal validation sets (7:3 ratio). External validation utilized retrospective data from January–December 2023. After univariate and multivariate regression analyses, four ML models (Logistic Regression, XGBoost, Naive Bayes, and SVM) were constructed. Receiver operating characteristic (ROC) curves and area under the curve (AUC) were calculated for training and internal validation sets. Model performance was compared using Delong's test or Bootstrap test, while sensitivity, specificity, accuracy, precision, recall, and F1-score evaluated predictive efficacy. Calibration curves assessed model calibration. The optimal model underwent external validation using ROC and calibration curves.

Results: A total of 788 older adult hemorrhagic stroke patients were enrolled, divided into a training set (n = 462), an internal validation set (n = 196), and an external validation set (n = 130). The incidence of SAP in older adult patients with hemorrhagic stroke was 46.7% (368/788). Advanced age [OR = 1.064, 95% CI (1.024, 1.104)], smoking[OR = 2.488, 95% CI (1.460, 4.24)], low GCS score [OR = 0.675, 95% CI (0.553, 0.825)], low Braden score [OR = 0.741, 95% CI (0.640, 0.858)], and nasogastric tube [OR = 1.761, 95% CI (1.048, 2.960)] were identified as risk factors for SAP. Among the four machine learning algorithms evaluated [XGBoost, Logistic Regression (LR), Support Vector Machine (SVM), and Naive Bayes], the LR model demonstrated robust and consistent performance in predicting SAP among older adult patients with hemorrhagic stroke across multiple evaluation metrics. Furthermore, the model exhibited stable generalizability within the external validation cohort. Based on these findings, the LR framework was subsequently selected for external validation, accompanied by a nomogram visualization. The model achieved AUC values of 0.883 (training), 0.855 (internal validation), and 0.882 (external validation). The Hosmer-Lemeshow (H-L) test indicates that the calibration of the model is satisfactory in all three datasets, with P-values of 0.381, 0.142, and 0.066 respectively.

Conclusions: This study constructed and validated a risk prediction model for SAP in older adult patients with hemorrhagic stroke based on multi-center data. The results indicated that among the four machine learning algorithms (XGBoost, LR, SVM, and Naive Bayes), the LR model demonstrated the best and most stable predictive performance. Age, smoking, low GCS score, low Braden score, and nasogastric tube were identified as predictive factors for SAP in these patients. These indicators are easily obtainable in clinical practice and facilitate rapid bedside assessment. Through internal and external validation, the model was proven to have good generalization ability, and a nomogram was ultimately drawn to provide an objective and operational risk assessment tool for clinical nursing practice. It helps in the early identification of high-risk patients and guides targeted interventions, thereby reducing the incidence of SAP and improving patient prognosis.

1 Introduction

Stroke-associated pneumonia (SAP) refers to newly acquired pneumonia in non-mechanically ventilated patients within 7 days of stroke onset (1). First proposed by German scholar Hilker in 2003 (2), subsequent studies report its incidence rate ranging from 6.5 to 58.4%, with risk factors including advanced age, male sex, smoking, dysphagia, hyperglycemia, and lower Glasgow Coma Scale (GCS) scores (3–8). Compared to non-SAP patients, SAP significantly worsens prognosis, leading to increased disability and mortality rates, prolonged hospitalization, and elevated healthcare costs (3, 5, 8–10). Meanwhile, with the intensification of population aging in China, the nursing needs of older adult patients with hemorrhagic stroke are becoming increasingly prominent (11). Current nursing strategies have deficiencies in aspects such as infection prevention, individualized intervention, and uneven distribution of medical resources. Especially in the context of limited medical resources, there is a lack of validated tools to prioritize the identification of high-risk patients and optimize nursing priorities, which restricts the prevention and control of SAP.

Machine Learning (ML) a subset of artificial intelligence (AI), enables in-depth exploration and analysis of extensive datasets, offering novel methodologies and research frameworks for precise prediction. Its applications span diverse fields, particularly in medicine, where ML facilitates the development of automated tools for clinical decision-making based on multidimensional medical data (10, 12). Risk prediction models, initially applied in cardiothoracic surgery (13, 14), leverage patient-specific risk factors and ML algorithms to forecast disease progression, therapeutic responses, and outcomes. Recent studies have utilized ML to integrate vital signs, epidemiological data, and laboratory/imaging findings for diagnostic or prognostic purposes. However, there are currently few risk prediction models for SAP in older adult patients with hemorrhagic stroke based on ML algorithms. The absence of such models not only limits the clinical early-warning ability but also hinders the precise allocation of nursing resources. This need is particularly urgent considering the characteristics of older adult patients with multiple underlying diseases and a short window period for nursing intervention. In this study, the prediction of SAP in older adult patients with hemorrhagic stroke was defined as a binary classification problem. Therefore, four widely used ML algorithms for solving classification problems (Logistic regression, Naive Bayes, Support Vector Machine, and eXtreme Gradient Boosting algorithm) were selected to construct the risk prediction model. The effectiveness of the model was evaluated through internal and external validation, aiming to provide references for clinical nursing practice, prevention, and control.

More at link