Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery
Abstract Hip fractures among the elderly population continue to present significant risks and high mortality rates despite advancements in surgical procedures. In this study, we developed machine learning (ML) algorithms to estimate 30-day mortality risk post-hip fracture surgery in the elderly usin...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-98713-6 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849399672813125632 |
|---|---|
| author | Fouad Trad Bassel Isber Ryan Yammine Khaled Hatoum Dana Obeid Mohammad Chahine Rachid Haidar Ghada El-Hajj Fuleihan Ali Chehab |
| author_facet | Fouad Trad Bassel Isber Ryan Yammine Khaled Hatoum Dana Obeid Mohammad Chahine Rachid Haidar Ghada El-Hajj Fuleihan Ali Chehab |
| author_sort | Fouad Trad |
| collection | DOAJ |
| description | Abstract Hip fractures among the elderly population continue to present significant risks and high mortality rates despite advancements in surgical procedures. In this study, we developed machine learning (ML) algorithms to estimate 30-day mortality risk post-hip fracture surgery in the elderly using data from the National Surgical Quality Improvement Program (NSQIP 2012–2017, n = 62,492 patients). Our approach involves two models: one estimating the patients’ 30-day mortality risk based on pre-operative conditions, and another considering both pre-operative and post-operative factors. We performed comprehensive data cleaning and preprocessing, then applied tenfold cross-validation with randomized search to the training set to identify optimal hyperparameters for various machine learning models. We used logistic regression, Naive Bayes, random forest, AdaBoost, XGBoost, CatBoost, Gradient Boosting, and LightGBM. The models’ performances were evaluated on the test set using the Area Under the Receiver Operating Characteristic Curve (AUC). The best pre-operative model was AdaBoost, achieving an AUC of 0.792 with 29 features (predictors), and the best post-operative model was CatBoost, achieving an AUC of 0.885 with 45 features. After modeling, we derived feature importance for each of the two models and decreased the number of features to reach a parsimonious highly performing model. The pre-operative model achieves an AUC of 0.725 with the eight most important features and the post-operative model achieves an AUC of 0.8529 with the six most important features. To ensure the models’ decision-making is compatible with clinical decisions and common practices, we applied explainability techniques such as SHAP to reveal the patterns learned by the models. These patterns were found to be clinically plausible. In summary, our approach involving data preprocessing, model tuning, feature selection, and explainability achieved state-of-the-art performance in predicting 30-day mortality rates following hip fractures surgery using a limited set of features, making it highly applicable in clinical settings. |
| format | Article |
| id | doaj-art-236c96008e6e43cd87d0b35940e2f12c |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-236c96008e6e43cd87d0b35940e2f12c2025-08-20T03:38:16ZengNature PortfolioScientific Reports2045-23222025-07-0115111410.1038/s41598-025-98713-6Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgeryFouad Trad0Bassel Isber1Ryan Yammine2Khaled Hatoum3Dana Obeid4Mohammad Chahine5Rachid Haidar6Ghada El-Hajj Fuleihan7Ali Chehab8Electrical and Computer Engineering Department, American University of BeirutElectrical and Computer Engineering Department, American University of BeirutCalcium Metabolism and Osteoporosis Program, WHO Collaborating Center for Metabolic Bone Disorders, Division of Endocrinology and Metabolism, Department of Internal Medicine, American University of Beirut Medical CenterElectrical and Computer Engineering Department, American University of BeirutElectrical and Computer Engineering Department, American University of BeirutElectrical and Computer Engineering Department, American University of BeirutDepartment of Surgery, American University of Beirut Medical CenterCalcium Metabolism and Osteoporosis Program, WHO Collaborating Center for Metabolic Bone Disorders, Division of Endocrinology and Metabolism, Department of Internal Medicine, American University of Beirut Medical CenterElectrical and Computer Engineering Department, American University of BeirutAbstract Hip fractures among the elderly population continue to present significant risks and high mortality rates despite advancements in surgical procedures. In this study, we developed machine learning (ML) algorithms to estimate 30-day mortality risk post-hip fracture surgery in the elderly using data from the National Surgical Quality Improvement Program (NSQIP 2012–2017, n = 62,492 patients). Our approach involves two models: one estimating the patients’ 30-day mortality risk based on pre-operative conditions, and another considering both pre-operative and post-operative factors. We performed comprehensive data cleaning and preprocessing, then applied tenfold cross-validation with randomized search to the training set to identify optimal hyperparameters for various machine learning models. We used logistic regression, Naive Bayes, random forest, AdaBoost, XGBoost, CatBoost, Gradient Boosting, and LightGBM. The models’ performances were evaluated on the test set using the Area Under the Receiver Operating Characteristic Curve (AUC). The best pre-operative model was AdaBoost, achieving an AUC of 0.792 with 29 features (predictors), and the best post-operative model was CatBoost, achieving an AUC of 0.885 with 45 features. After modeling, we derived feature importance for each of the two models and decreased the number of features to reach a parsimonious highly performing model. The pre-operative model achieves an AUC of 0.725 with the eight most important features and the post-operative model achieves an AUC of 0.8529 with the six most important features. To ensure the models’ decision-making is compatible with clinical decisions and common practices, we applied explainability techniques such as SHAP to reveal the patterns learned by the models. These patterns were found to be clinically plausible. In summary, our approach involving data preprocessing, model tuning, feature selection, and explainability achieved state-of-the-art performance in predicting 30-day mortality rates following hip fractures surgery using a limited set of features, making it highly applicable in clinical settings.https://doi.org/10.1038/s41598-025-98713-6Machine learningHip fracturePre-operative mortality risk scorePost-operative mortality risk scoreAI explainability |
| spellingShingle | Fouad Trad Bassel Isber Ryan Yammine Khaled Hatoum Dana Obeid Mohammad Chahine Rachid Haidar Ghada El-Hajj Fuleihan Ali Chehab Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery Scientific Reports Machine learning Hip fracture Pre-operative mortality risk score Post-operative mortality risk score AI explainability |
| title | Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery |
| title_full | Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery |
| title_fullStr | Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery |
| title_full_unstemmed | Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery |
| title_short | Parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery |
| title_sort | parsimonious and explainable machine learning for predicting mortality in patients post hip fracture surgery |
| topic | Machine learning Hip fracture Pre-operative mortality risk score Post-operative mortality risk score AI explainability |
| url | https://doi.org/10.1038/s41598-025-98713-6 |
| work_keys_str_mv | AT fouadtrad parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT basselisber parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT ryanyammine parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT khaledhatoum parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT danaobeid parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT mohammadchahine parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT rachidhaidar parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT ghadaelhajjfuleihan parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery AT alichehab parsimoniousandexplainablemachinelearningforpredictingmortalityinpatientsposthipfracturesurgery |