Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations
Background: Despite increasing utilization and survival benefit over the last decade, extracorporeal membrane oxygenation (ECMO) remains resource-intensive with significant complications and rehospitalization risk. We thus utilized machine learning (ML) to develop prediction models for 90-day nonele...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-06-01
|
| Series: | Surgery Open Science |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2589845024000538 |
| _version_ | 1797202970946830336 |
|---|---|
| author | Jeffrey Balian Sara Sakowitz, MS, MPH Arjun Verma, BS Amulya Vadlakonda, BS Emma Cruz Konmal Ali Peyman Benharash, MD |
| author_facet | Jeffrey Balian Sara Sakowitz, MS, MPH Arjun Verma, BS Amulya Vadlakonda, BS Emma Cruz Konmal Ali Peyman Benharash, MD |
| author_sort | Jeffrey Balian |
| collection | DOAJ |
| description | Background: Despite increasing utilization and survival benefit over the last decade, extracorporeal membrane oxygenation (ECMO) remains resource-intensive with significant complications and rehospitalization risk. We thus utilized machine learning (ML) to develop prediction models for 90-day nonelective readmission following ECMO. Methods: All adult patients receiving ECMO who survived index hospitalization were tabulated from the 2016–2020 Nationwide Readmissions Database. Extreme Gradient Boosting (XGBoost) models were developed to identify features associated with readmission following ECMO. Area under the receiver operating characteristic (AUROC), mean Average Precision (mAP), and the Brier score were calculated to estimate model performance relative to logistic regression (LR). Shapley Additive Explanation summary (SHAP) plots evaluated the relative impact of each factor on the model. An additional sensitivity analysis solely included patient comorbidities and indication for ECMO as potential model covariates. Results: Of ∼22,947 patients, 4495 (19.6 %) were readmitted nonelectively within 90 days. The XGBoost model exhibited superior discrimination (AUROC 0.64 vs 0.49), classification accuracy (mAP 0.30 vs 0.20) and calibration (Brier score 0.154 vs 0.165, all P < 0.001) in predicting readmission compared to LR. SHAP plots identified duration of index hospitalization, undergoing heart/lung transplantation, and Medicare insurance to be associated with increased odds of readmission. Upon sub-analysis, XGBoost demonstrated superior disclination compared to LR (AUROC 0.61 vs 0.60, P < 0.05). Chronic liver disease and frailty were linked with increased odds of nonelective readmission. Conclusions: ML outperformed LR in predicting readmission following ECMO. Future work is needed to identify other factors linked with readmission and further optimize post-ECMO care among this cohort. |
| first_indexed | 2024-04-24T08:11:54Z |
| format | Article |
| id | doaj.art-8a9048b8b6474128802178c16efa878e |
| institution | Directory Open Access Journal |
| issn | 2589-8450 |
| language | English |
| last_indexed | 2024-04-24T08:11:54Z |
| publishDate | 2024-06-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Surgery Open Science |
| spelling | doaj.art-8a9048b8b6474128802178c16efa878e2024-04-17T04:49:46ZengElsevierSurgery Open Science2589-84502024-06-0119125130Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizationsJeffrey Balian0Sara Sakowitz, MS, MPH1Arjun Verma, BS2Amulya Vadlakonda, BS3Emma Cruz4Konmal Ali5Peyman Benharash, MD6Cardiovascular Outcomes Research Laboratories (CORELAB), University of California, Los Angeles, CA, United States of AmericaCardiovascular Outcomes Research Laboratories (CORELAB), University of California, Los Angeles, CA, United States of AmericaCardiovascular Outcomes Research Laboratories (CORELAB), University of California, Los Angeles, CA, United States of AmericaCardiovascular Outcomes Research Laboratories (CORELAB), University of California, Los Angeles, CA, United States of AmericaCardiovascular Outcomes Research Laboratories (CORELAB), University of California, Los Angeles, CA, United States of AmericaCardiovascular Outcomes Research Laboratories (CORELAB), University of California, Los Angeles, CA, United States of AmericaCardiovascular Outcomes Research Laboratories (CORELAB), University of California, Los Angeles, CA, United States of America; Division of Cardiac Surgery, Department of Surgery, University of California, Los Angeles, CA, United States of America; Corresponding author at: UCLA Division of Cardiac Surgery, 64-249 Center for Health Sciences, Los Angeles, CA 90095, United States of America.Background: Despite increasing utilization and survival benefit over the last decade, extracorporeal membrane oxygenation (ECMO) remains resource-intensive with significant complications and rehospitalization risk. We thus utilized machine learning (ML) to develop prediction models for 90-day nonelective readmission following ECMO. Methods: All adult patients receiving ECMO who survived index hospitalization were tabulated from the 2016–2020 Nationwide Readmissions Database. Extreme Gradient Boosting (XGBoost) models were developed to identify features associated with readmission following ECMO. Area under the receiver operating characteristic (AUROC), mean Average Precision (mAP), and the Brier score were calculated to estimate model performance relative to logistic regression (LR). Shapley Additive Explanation summary (SHAP) plots evaluated the relative impact of each factor on the model. An additional sensitivity analysis solely included patient comorbidities and indication for ECMO as potential model covariates. Results: Of ∼22,947 patients, 4495 (19.6 %) were readmitted nonelectively within 90 days. The XGBoost model exhibited superior discrimination (AUROC 0.64 vs 0.49), classification accuracy (mAP 0.30 vs 0.20) and calibration (Brier score 0.154 vs 0.165, all P < 0.001) in predicting readmission compared to LR. SHAP plots identified duration of index hospitalization, undergoing heart/lung transplantation, and Medicare insurance to be associated with increased odds of readmission. Upon sub-analysis, XGBoost demonstrated superior disclination compared to LR (AUROC 0.61 vs 0.60, P < 0.05). Chronic liver disease and frailty were linked with increased odds of nonelective readmission. Conclusions: ML outperformed LR in predicting readmission following ECMO. Future work is needed to identify other factors linked with readmission and further optimize post-ECMO care among this cohort.http://www.sciencedirect.com/science/article/pii/S2589845024000538Extracorporeal membrane oxygenationECMOMachine learningXGBoostNational Readmissions DatabaseHCUP |
| spellingShingle | Jeffrey Balian Sara Sakowitz, MS, MPH Arjun Verma, BS Amulya Vadlakonda, BS Emma Cruz Konmal Ali Peyman Benharash, MD Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations Surgery Open Science Extracorporeal membrane oxygenation ECMO Machine learning XGBoost National Readmissions Database HCUP |
| title | Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations |
| title_full | Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations |
| title_fullStr | Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations |
| title_full_unstemmed | Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations |
| title_short | Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations |
| title_sort | machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations |
| topic | Extracorporeal membrane oxygenation ECMO Machine learning XGBoost National Readmissions Database HCUP |
| url | http://www.sciencedirect.com/science/article/pii/S2589845024000538 |
| work_keys_str_mv | AT jeffreybalian machinelearningbasedpredictivemodelingofreadmissionsfollowingextracorporealmembraneoxygenationhospitalizations AT sarasakowitzmsmph machinelearningbasedpredictivemodelingofreadmissionsfollowingextracorporealmembraneoxygenationhospitalizations AT arjunvermabs machinelearningbasedpredictivemodelingofreadmissionsfollowingextracorporealmembraneoxygenationhospitalizations AT amulyavadlakondabs machinelearningbasedpredictivemodelingofreadmissionsfollowingextracorporealmembraneoxygenationhospitalizations AT emmacruz machinelearningbasedpredictivemodelingofreadmissionsfollowingextracorporealmembraneoxygenationhospitalizations AT konmalali machinelearningbasedpredictivemodelingofreadmissionsfollowingextracorporealmembraneoxygenationhospitalizations AT peymanbenharashmd machinelearningbasedpredictivemodelingofreadmissionsfollowingextracorporealmembraneoxygenationhospitalizations |