Using machine learning to predict stroke‐associated pneumonia in Chinese acute ischaemic stroke patients

Li, X.; Wu, Min; Sun, Chao; Zhao, Zheng; Wang, F.; Zheng, Xuqin; Ge, Weihong; Zhou, Junshan; Zou, Jianjun

doi:10.1111/ene.14295

Cited by 37 publications

(22 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This model assesses risk factors, including age, atrial fibrillation, dysphagia, sex, and previous stroke severity. In a total of 3,160 Chinese AIS patients, Li et al ( 16 ) used machine learning methods to develop a model with high sensitivity and specificity to predict SAP. Interestingly, a study found that reduced vitamin D was a potential risk factor of SAP ( 17 ).…”

Section: Introductionmentioning

confidence: 99%

High Monocyte-To-Lymphocyte Ratio Is Associated With Stroke-Associated Pneumonia

et al. 2020

View full text Add to dashboard Cite

Purpose: Stroke-associated pneumonia (SAP), a common complication in acute ischemic stroke (AIS) patients, is associated with poor prognosis after AIS. Inflammation plays an important role in the development of SAP. In this study, we aimed to explore the association between the monocyte-to-lymphocyte ratio (MLR) and SAP in AIS patients. Methods: We continuously enrolled 972 AIS patients. SAP was diagnosed by two trained neurologists and confirmed by radiography, meeting the modified Centers for Disease Control and Prevention criteria. MLR values were measured for all participants, and all patients were evenly classified into three tertiles according to the MLR levels. We used the values that Youden's index max points corresponded to represent the optimal cutoffs, which represented the balance in sensitivity and specificity. Results: 104 (10.7%) patients were diagnosed with SAP. SAP patients showed a significant increased ( P < 0.001) MLR when compared with non-SAP. The optimal cutoff points of MLR were (T1) <0.2513, (T2) 0.2513–0.3843, and (T3) > 0.3843. The incidence of SAP was significantly higher in the third MLR tertile than the first and second MLR tertiles (21.7 vs. 4 vs. 6.5%, respectively, P < 0.001). After adjusting for confounding and risk factors, multivariate regression analysis showed that the third MLR tertile was an independent variable predicting the occurrence of SAP (odds ratio = 3.503, 95%CI = 1.066–11.515, P = 0.039). Conclusions: Our study showed that higher MLR was significantly associated with SAP in AIS patients. MLR is beneficial for clinicians to recognize patients with a high risk of SAP at an early stage and is an effective way to improve clinical care of SAP patients. Higher MLR could be a helpful and valid biomarker for predicting SAP in clinical practice.

show abstract

Section: Introductionmentioning

confidence: 99%

High Monocyte-To-Lymphocyte Ratio Is Associated With Stroke-Associated Pneumonia

et al. 2020

View full text Add to dashboard Cite

show abstract

“…SAP not only increases the length of hospital stay and hospital costs [4,5], but is also an important risk factor for poor outcomes after acute stroke [6,7]. Therefore, it is important to nd a scale that is effective in predicting SAP and can help clinicians take early preventative measures to reduce the incidence of SAP [8,9]. Based on the con rmed risk factors for SAP, including an older age, the male sex, dysphagia, preadmission dependency, stroke severity, etc.…”

Section: Introductionmentioning

confidence: 99%

The Feasibility of the Braden Scale for Assessing Pneumonia After Spontaneous Intracerebral Hemorrhage

Ding

Zhao

et al. 2020

Preprint

View full text Add to dashboard Cite

Background Stroke-associated pneumonia (SAP) is an infection that commonly occurs in patients with spontaneous intracerebral hemorrhage (ICH) and causes serious burdens. The six subscales of the Braden scale seem to be related to the occurrence of pneumonia. We aimed to evaluate the feasibility of the Braden scale for predicting SAP after spontaneous ICH. Methods Patients with spontaneous ICH who were admitted to Jingjiang People’s Hospital and Zhoukou Central Hospital were retrospectively included and divided into two groups: the pneumonia and no pneumonia groups. The Braden scale and ICH-APS-A scale scores and demographic and clinical characteristics were collected, and the differences between the two groups were compared with statistical analyses. Receiver operating characteristic (ROC) curve analysis was used to assess the predictive validity of the Braden scale for SAP after ICH. Results A total of 629 patients with ICH were included in this study, 150 (23.8%) of whom developed SAP. There were significant differences in age and fasting blood glucose level between the two groups, and the ICH-ASP-A score in the pneumonia group was significantly higher than that in the no pneumonia group (8.5 ± 3.9 vs 4.9 ± 3.2, P < 0.01). The mean score on the Braden scale in the pneumonia group was significantly lower than that in the no pneumonia group (16.5 ± 2.6, P < 0.01). The area under the curve (AUC) for the ICH-ASP-A score for the prediction of pneumonia after spontaneous ICH was 0.755 (95% CI = 0.712–0.798). When the cutoff point was 8 points, the sensitivity was 56.7% and the specificity was 77.5%. The AUC for the Braden scale for the prediction of pneumonia after ICH was 0.760 (95% CI = 0.717–0.804). When the cutoff point was 15 points, the sensitivity was 74.3% and the specificity was 64.7%. Conclusions The Braden scale is effective in predicting pneumonia after ICH. The Braden scale, with a cutoff point of 15 points, is a valid clinical grading scale for predicting SAP after spontaneous ICH.

show abstract

“…ML models are usually constructed based on high volume data recorded in the electronic patient record (EPR) systems and its deep learning ability allows ML models to capture complex, nonlinear relationships, even previously unknown correlations in big data, digging deeper into clinical data [14], and shows promising potential in clinical scenes where large amount of data were collected and integrated every day. Recently, Li and colleagues [15] have developed a model using ML methods to predict stroke-associated pneumonia in Chinese patients with acute ischemic stroke. In addition, ML was used to predict severe pneumonia during posttransplant hospitalization in recipients of a kidney transplant [16].…”

Section: Introductionmentioning

confidence: 99%

Development and performance assessment of novel machine learning models to predict pneumonia after liver transplantation

Chen

Yang²,

Gao

et al. 2021

Respir Res

View full text Add to dashboard Cite

Background Pneumonia is the most frequently encountered postoperative pulmonary complications (PPC) after orthotopic liver transplantation (OLT), which cause high morbidity and mortality rates. We aimed to develop a model to predict postoperative pneumonia in OLT patients using machine learning (ML) methods. Methods Data of 786 adult patients underwent OLT at the Third Affiliated Hospital of Sun Yat-sen University from January 2015 to September 2019 was retrospectively extracted from electronic medical records and randomly subdivided into a training set and a testing set. With the training set, six ML models including logistic regression (LR), support vector machine (SVM), random forest (RF), adaptive boosting (AdaBoost), extreme gradient boosting (XGBoost) and gradient boosting machine (GBM) were developed. These models were assessed by the area under curve (AUC) of receiver operating characteristic on the testing set. The related risk factors and outcomes of pneumonia were also probed based on the chosen model. Results 591 OLT patients were eventually included and 253 (42.81%) were diagnosed with postoperative pneumonia, which was associated with increased postoperative hospitalization and mortality (P < 0.05). Among the six ML models, XGBoost model performed best. The AUC of XGBoost model on the testing set was 0.734 (sensitivity: 52.6%; specificity: 77.5%). Pneumonia was notably associated with 14 items features: INR, HCT, PLT, ALB, ALT, FIB, WBC, PT, serum Na+, TBIL, anesthesia time, preoperative length of stay, total fluid transfusion and operation time. Conclusion Our study firstly demonstrated that the XGBoost model with 14 common variables might predict postoperative pneumonia in OLT patients.

show abstract

Using machine learning to predict stroke‐associated pneumonia in Chinese acute ischaemic stroke patients

Cited by 37 publications

References 32 publications

High Monocyte-To-Lymphocyte Ratio Is Associated With Stroke-Associated Pneumonia

High Monocyte-To-Lymphocyte Ratio Is Associated With Stroke-Associated Pneumonia

The Feasibility of the Braden Scale for Assessing Pneumonia After Spontaneous Intracerebral Hemorrhage

Development and performance assessment of novel machine learning models to predict pneumonia after liver transplantation

Contact Info

Product

Resources

About