Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study
Abstract Background Delirium is a severe complication in critical elderly patients. This study aimed to develop interpretable machine-learning (ML) models to predict acute delirium and identify risk factors for medical intervention in elderly patients in the intensive care unit (ICU). Patients and M...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2025-02-01
|
| Series: | Journal of Big Data |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s40537-025-01107-8 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850238001557274624 |
|---|---|
| author | Xin Hu Jun Luo Hong Liang Jingwei Yue Yeqing Qi Hui Liu |
| author_facet | Xin Hu Jun Luo Hong Liang Jingwei Yue Yeqing Qi Hui Liu |
| author_sort | Xin Hu |
| collection | DOAJ |
| description | Abstract Background Delirium is a severe complication in critical elderly patients. This study aimed to develop interpretable machine-learning (ML) models to predict acute delirium and identify risk factors for medical intervention in elderly patients in the intensive care unit (ICU). Patients and Methods elderly patients (age ≥ 65 and ≤ 89) were selected from electronic intensive care unit collaborative research database (eICU-CRD). Data of demographics and laboratory tests were collected on the first day of admission to ICU. Delirium in 7 days after admission was identified. Difference between delirium and non-delirium groups was demonstrated. Association between delirium and mortality was proved through Kaplan–Meier survival curve. Participants were randomly distributed into a training set and a validation set without replacement at a ratio of 7:3. Recursive feature elimination (RFE) was used to determine the number of variables adopted in the model. The predictive capability of the ML models was demonstrated by receiver operating characteristic (ROC) analysis and calibration curve analysis. The interpretability of the model was demonstrated with SHapley Additive ExPlanations (SHAP). Results a total of 66263 elderly patients were selected, and in which 6299 patients (9.5%) were identified as acute delirium (within 7d after admission). Hospital mortality in delirium group was higher than that in non-delirium group (16.32% vs. 10.63%, p = 0.000). The cumulative survival probability of non-delirium patients were significantly higher than that of delirium patients (p < 0.001). When 20 variables were adopted, RandomForest and Xgboost models showed the highest predictive capability with the area under curve (AUC) = 0.91. Calibration curve analysis also proved this result. Glascow Coma Scale (GCS), acute physical and chronic health evaluation IV (APACHE IV), and sepsis had the highest importance in ML models. Mechanical ventilation and temperature were also important risk factors of acute delirium. Conclusion Acute delirium is an independent risk of mortality in elderly patients in the intensive care unit. APACHEIV, sepsis, mechanical ventilation and temperature were important risk factors of acute delirium, which were potential targets for medication. |
| format | Article |
| id | doaj-art-c8261df155df41dca1ff01310f170313 |
| institution | OA Journals |
| issn | 2196-1115 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | Journal of Big Data |
| spelling | doaj-art-c8261df155df41dca1ff01310f1703132025-08-20T02:01:35ZengSpringerOpenJournal of Big Data2196-11152025-02-0112111710.1186/s40537-025-01107-8Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort studyXin Hu0Jun Luo1Hong Liang2Jingwei Yue3Yeqing Qi4Hui Liu5Department of Critical Care Medicine, the First Medical Center, Chinese PLA General HospitalDepartment of Critical Care Medicine, Xuanhan County People’s HospitalMedical Innovation Research Division, Chinese PLA General HospitalBeijing Institute of Radiation MedicineDepartment of Radiology, the First Medical Center, Chinese PLA General HospitalDepartment of Critical Care Medicine, the First Medical Center, Chinese PLA General HospitalAbstract Background Delirium is a severe complication in critical elderly patients. This study aimed to develop interpretable machine-learning (ML) models to predict acute delirium and identify risk factors for medical intervention in elderly patients in the intensive care unit (ICU). Patients and Methods elderly patients (age ≥ 65 and ≤ 89) were selected from electronic intensive care unit collaborative research database (eICU-CRD). Data of demographics and laboratory tests were collected on the first day of admission to ICU. Delirium in 7 days after admission was identified. Difference between delirium and non-delirium groups was demonstrated. Association between delirium and mortality was proved through Kaplan–Meier survival curve. Participants were randomly distributed into a training set and a validation set without replacement at a ratio of 7:3. Recursive feature elimination (RFE) was used to determine the number of variables adopted in the model. The predictive capability of the ML models was demonstrated by receiver operating characteristic (ROC) analysis and calibration curve analysis. The interpretability of the model was demonstrated with SHapley Additive ExPlanations (SHAP). Results a total of 66263 elderly patients were selected, and in which 6299 patients (9.5%) were identified as acute delirium (within 7d after admission). Hospital mortality in delirium group was higher than that in non-delirium group (16.32% vs. 10.63%, p = 0.000). The cumulative survival probability of non-delirium patients were significantly higher than that of delirium patients (p < 0.001). When 20 variables were adopted, RandomForest and Xgboost models showed the highest predictive capability with the area under curve (AUC) = 0.91. Calibration curve analysis also proved this result. Glascow Coma Scale (GCS), acute physical and chronic health evaluation IV (APACHE IV), and sepsis had the highest importance in ML models. Mechanical ventilation and temperature were also important risk factors of acute delirium. Conclusion Acute delirium is an independent risk of mortality in elderly patients in the intensive care unit. APACHEIV, sepsis, mechanical ventilation and temperature were important risk factors of acute delirium, which were potential targets for medication.https://doi.org/10.1186/s40537-025-01107-8DeliriumXGboostRandomForestElderly patientseICU-CRD |
| spellingShingle | Xin Hu Jun Luo Hong Liang Jingwei Yue Yeqing Qi Hui Liu Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study Journal of Big Data Delirium XGboost RandomForest Elderly patients eICU-CRD |
| title | Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study |
| title_full | Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study |
| title_fullStr | Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study |
| title_full_unstemmed | Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study |
| title_short | Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study |
| title_sort | interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units a retrospective cohort study |
| topic | Delirium XGboost RandomForest Elderly patients eICU-CRD |
| url | https://doi.org/10.1186/s40537-025-01107-8 |
| work_keys_str_mv | AT xinhu interpretablemachinelearningmodeltopredicttheacuteoccurrenceofdeliriuminelderlypatientsintheintensivecareunitsaretrospectivecohortstudy AT junluo interpretablemachinelearningmodeltopredicttheacuteoccurrenceofdeliriuminelderlypatientsintheintensivecareunitsaretrospectivecohortstudy AT hongliang interpretablemachinelearningmodeltopredicttheacuteoccurrenceofdeliriuminelderlypatientsintheintensivecareunitsaretrospectivecohortstudy AT jingweiyue interpretablemachinelearningmodeltopredicttheacuteoccurrenceofdeliriuminelderlypatientsintheintensivecareunitsaretrospectivecohortstudy AT yeqingqi interpretablemachinelearningmodeltopredicttheacuteoccurrenceofdeliriuminelderlypatientsintheintensivecareunitsaretrospectivecohortstudy AT huiliu interpretablemachinelearningmodeltopredicttheacuteoccurrenceofdeliriuminelderlypatientsintheintensivecareunitsaretrospectivecohortstudy |