Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data
Hospital intensive care units (ICUs) care for severely ill patients, many of whom require some form of organ support. Clinicians in ICUs are often challenged with integrating large volumes of continuously recorded physiological and clinical data in order to diagnose and treat patients. In this work,...
| Main Authors: | , , , , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Institute of Electrical and Electronics Engineers (IEEE)
2019
|
| Online Access: | https://hdl.handle.net/1721.1/123313 |
| _version_ | 1811098374880886784 |
|---|---|
| author | Ren, Oliver Johnson, Alistair Edward William Lehman, Eric P. Komorowski, Matthieu Aboab, Jerome Emile Francois Leon Tang, Fengyi Shahn, Zach Sow, Daby Sow, Daby Mark, Roger G Lehman, Li-wei |
| author2 | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science |
| author_facet | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Ren, Oliver Johnson, Alistair Edward William Lehman, Eric P. Komorowski, Matthieu Aboab, Jerome Emile Francois Leon Tang, Fengyi Shahn, Zach Sow, Daby Sow, Daby Mark, Roger G Lehman, Li-wei |
| author_sort | Ren, Oliver |
| collection | MIT |
| description | Hospital intensive care units (ICUs) care for severely ill patients, many of whom require some form of organ support. Clinicians in ICUs are often challenged with integrating large volumes of continuously recorded physiological and clinical data in order to diagnose and treat patients. In this work, we focus on developing interpretable models for predicting unexpected respiratory decompensation requiring intubation in ICU patients. Predicting need for intubation could have important implications for the patient and medical staff and potentially enable timely interventions for improved patient outcome. Using data from adult ICU patients from the Medical Information Mart for Intensive Care (MIMIC)-III database, we developed gradient boosting models for predicting intubation onset. In a cohort of 12,470 patients, of whom 1,067 were intubated (8.55%), we achieved an area under the receiver operating characteristic curve (AUROC) of 0.89, with 95% confidence interval (CI) 0.87 - 0.91, when predicting intubation 3 hours ahead of time, a significant increase (p<0.001) over the AUROC achieved using several baselines, including logistic regression (0.81, 95% CI 0.78 - 0.84) and neural networks (0.80, 95% CI 0.77 - 0.83]). Finally, we conducted feature importance analysis using gradient boosting and derived useful insights in understanding the relative importance of clinical vs. biological variables in predicting impending respiratory decompensation in ICUs. |
| first_indexed | 2024-09-23T17:14:03Z |
| format | Article |
| id | mit-1721.1/123313 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T17:14:03Z |
| publishDate | 2019 |
| publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| record_format | dspace |
| spelling | mit-1721.1/1233132022-10-03T11:16:44Z Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data Ren, Oliver Johnson, Alistair Edward William Lehman, Eric P. Komorowski, Matthieu Aboab, Jerome Emile Francois Leon Tang, Fengyi Shahn, Zach Sow, Daby Sow, Daby Mark, Roger G Lehman, Li-wei Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology. Institute for Medical Engineering & Science Hospital intensive care units (ICUs) care for severely ill patients, many of whom require some form of organ support. Clinicians in ICUs are often challenged with integrating large volumes of continuously recorded physiological and clinical data in order to diagnose and treat patients. In this work, we focus on developing interpretable models for predicting unexpected respiratory decompensation requiring intubation in ICU patients. Predicting need for intubation could have important implications for the patient and medical staff and potentially enable timely interventions for improved patient outcome. Using data from adult ICU patients from the Medical Information Mart for Intensive Care (MIMIC)-III database, we developed gradient boosting models for predicting intubation onset. In a cohort of 12,470 patients, of whom 1,067 were intubated (8.55%), we achieved an area under the receiver operating characteristic curve (AUROC) of 0.89, with 95% confidence interval (CI) 0.87 - 0.91, when predicting intubation 3 hours ahead of time, a significant increase (p<0.001) over the AUROC achieved using several baselines, including logistic regression (0.81, 95% CI 0.78 - 0.84) and neural networks (0.80, 95% CI 0.77 - 0.83]). Finally, we conducted feature importance analysis using gradient boosting and derived useful insights in understanding the relative importance of clinical vs. biological variables in predicting impending respiratory decompensation in ICUs. National Institutes of Health (U.S.) (Grant R01-EB017205) National Institutes of Health (U.S.) (Grant R01-EB001659) National Institutes of Health (U.S.) (Grant R01GM104987) 2019-12-20T00:22:14Z 2019-12-20T00:22:14Z 2018-07 2018-06 2019-12-04T19:26:17Z Article http://purl.org/eprint/type/ConferencePaper 9781538653777 2575-2634 https://hdl.handle.net/1721.1/123313 Ren, Oliver et al. "Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data." 2018 IEEE International Conference on Healthcare Informatics (ICHI), June 2018, New York, New York,USA, Institute of Electrical and Electronics Engineers (IEEE), July 2018 © 2018 IEEE en http://dx.doi.org/10.1109/ichi.2018.00024 2018 IEEE International Conference on Healthcare Informatics (ICHI) Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Institute of Electrical and Electronics Engineers (IEEE) Prof. Mark via Courtney Crummett |
| spellingShingle | Ren, Oliver Johnson, Alistair Edward William Lehman, Eric P. Komorowski, Matthieu Aboab, Jerome Emile Francois Leon Tang, Fengyi Shahn, Zach Sow, Daby Sow, Daby Mark, Roger G Lehman, Li-wei Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data |
| title | Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data |
| title_full | Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data |
| title_fullStr | Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data |
| title_full_unstemmed | Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data |
| title_short | Predicting and Understanding Unexpected Respiratory Decompensation in Critical Care Using Sparse and Heterogeneous Clinical Data |
| title_sort | predicting and understanding unexpected respiratory decompensation in critical care using sparse and heterogeneous clinical data |
| url | https://hdl.handle.net/1721.1/123313 |
| work_keys_str_mv | AT renoliver predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT johnsonalistairedwardwilliam predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT lehmanericp predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT komorowskimatthieu predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT aboabjeromeemilefrancoisleon predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT tangfengyi predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT shahnzach predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT sowdaby predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT sowdaby predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT markrogerg predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata AT lehmanliwei predictingandunderstandingunexpectedrespiratorydecompensationincriticalcareusingsparseandheterogeneousclinicaldata |