دورية أكاديمية
Explainable Machine Learning on AmsterdamUMCdb for ICU Discharge Decision Support: Uniting Intensivists and Data Scientists.
| العنوان: | Explainable Machine Learning on AmsterdamUMCdb for ICU Discharge Decision Support: Uniting Intensivists and Data Scientists. |
|---|---|
| المؤلفون: | Thoral PJ; Department of Intensive Care Medicine, Laboratory for Critical Care Computational Intelligence (LCCCI), Amsterdam Medical Data Science (AMDS), Amsterdam UMC, Vrije Universiteit, Amsterdam, The Netherlands., Fornasa M; Pacmed BV, Amsterdam, The Netherlands., de Bruin DP; Pacmed BV, Amsterdam, The Netherlands., Tonutti M; Pacmed BV, Amsterdam, The Netherlands., Hovenkamp H; Pacmed BV, Amsterdam, The Netherlands., Driessen RH; Department of Intensive Care Medicine, Laboratory for Critical Care Computational Intelligence (LCCCI), Amsterdam Medical Data Science (AMDS), Amsterdam UMC, Vrije Universiteit, Amsterdam, The Netherlands., Girbes ARJ; Department of Intensive Care Medicine, Laboratory for Critical Care Computational Intelligence (LCCCI), Amsterdam Medical Data Science (AMDS), Amsterdam UMC, Vrije Universiteit, Amsterdam, The Netherlands., Hoogendoorn M; Computational Intelligence Group, Department of Computer Science, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands., Elbers PWG; Department of Intensive Care Medicine, Laboratory for Critical Care Computational Intelligence (LCCCI), Amsterdam Medical Data Science (AMDS), Amsterdam UMC, Vrije Universiteit, Amsterdam, The Netherlands. |
| المصدر: | Critical care explorations [Crit Care Explor] 2021 Sep 10; Vol. 3 (9), pp. e0529. Date of Electronic Publication: 2021 Sep 10 (Print Publication: 2021). |
| نوع المنشور: | Journal Article |
| اللغة: | English |
| بيانات الدورية: | Publisher: Wolters Kluwer Health Country of Publication: United States NLM ID: 101746347 Publication Model: eCollection Cited Medium: Internet ISSN: 2639-8028 (Electronic) Linking ISSN: 26398028 NLM ISO Abbreviation: Crit Care Explor Subsets: PubMed not MEDLINE |
| أسماء مطبوعة: | Original Publication: Philadelphia, PA : Wolters Kluwer Health, [2019]- |
| مستخلص: | Unexpected ICU readmission is associated with longer length of stay and increased mortality. To prevent ICU readmission and death after ICU discharge, our team of intensivists and data scientists aimed to use AmsterdamUMCdb to develop an explainable machine learning-based real-time bedside decision support tool. Derivation Cohort: Data from patients admitted to a mixed surgical-medical academic medical center ICU from 2004 to 2016. Validation Cohort: Data from 2016 to 2019 from the same center. Prediction Model: Patient characteristics, clinical observations, physiologic measurements, laboratory studies, and treatment data were considered as model features. Different supervised learning algorithms were trained to predict ICU readmission and/or death, both within 7 days from ICU discharge, using 10-fold cross-validation. Feature importance was determined using SHapley Additive exPlanations, and readmission probability-time curves were constructed to identify subgroups. Explainability was established by presenting individualized risk trends and feature importance. Results: Our final derivation dataset included 14,105 admissions. The combined readmission/mortality rate within 7 days of ICU discharge was 5.3%. Using Gradient Boosting, the model achieved an area under the receiver operating characteristic curve of 0.78 (95% CI, 0.75-0.81) and an area under the precision-recall curve of 0.19 on the validation cohort ( n = 3,929). The most predictive features included common physiologic parameters but also less apparent variables like nutritional support. At a 6% risk threshold, the model showed a sensitivity (recall) of 0.72, specificity of 0.70, and a positive predictive value (precision) of 0.15. Impact analysis using probability-time curves and the 6% risk threshold identified specific patient groups at risk and the potential of a change in discharge management to reduce relative risk by 14%. Conclusions: We developed an explainable machine learning model that may aid in identifying patients at high risk for readmission and mortality after ICU discharge using the first freely available European critical care database, AmsterdamUMCdb. Impact analysis showed that a relative risk reduction of 14% could be achievable, which might have significant impact on patients and society. ICU data sharing facilitates collaboration between intensivists and data scientists to accelerate model development. Competing Interests: Amsterdam University Medical Centers is entitled to royalties from Pacmed. Mr. Hovenkamp is cofounder of Pacmed, a Dutch technology company in health care. The remaining authors have disclosed that they do not have any potential conflicts of interest. (Copyright © 2021 The Authors. Published by Wolters Kluwer Health, Inc. on behalf of the Society of Critical Care Medicine.) |
| References: | BMJ Open. 2019 Mar 7;9(3):e025925. (PMID: 30850412) QJM. 2001 Oct;94(10):521-6. (PMID: 11588210) Br J Anaesth. 2010 Apr;104(4):459-64. (PMID: 20185517) J Trauma. 2006 May;60(5):1027-31. (PMID: 16688065) BMJ Open. 2017 Sep 15;7(9):e017199. (PMID: 28918412) Ann Am Thorac Soc. 2018 Jul;15(7):846-853. (PMID: 29787309) Crit Care Med. 2018 May;46(5):749-756. (PMID: 29394182) BMJ Open. 2017 Dec 26;7(12):e018322. (PMID: 29282265) Crit Care Med. 2008 Mar;36(3):676-82. (PMID: 18431260) J Intensive Care Med. 2019 Nov-Dec;34(11-12):924-929. (PMID: 30270722) Annu Int Conf IEEE Eng Med Biol Soc. 2017 Jul;2017:2570-2573. (PMID: 29060424) AMIA Annu Symp Proc. 2017 Feb 10;2016:827-836. (PMID: 28269879) Crit Care Med. 2021 Jun 1;49(6):e563-e577. (PMID: 33625129) Yonsei Med J. 2015 Mar;56(2):543-9. (PMID: 25684007) Anaesth Intensive Care. 2018 Jan;46(1):88-96. (PMID: 29361261) JPEN J Parenter Enteral Nutr. 2012 Jan;36(1):60-8. (PMID: 22167076) J Intensive Care. 2015 Jun 26;3(1):30. (PMID: 26157581) Med Decis Making. 2006 Nov-Dec;26(6):565-74. (PMID: 17099194) Crit Care. 2013 Jun 29;17(3):R102. (PMID: 23718698) BMJ. 2001 May 26;322(7297):1274-6. (PMID: 11375229) Crit Care Med. 2017 Feb;45(2):290-297. (PMID: 27632681) Intensive Care Med. 2010 Jul;36(7):1196-201. (PMID: 20221748) Crit Care Resusc. 2010 Jun;12(2):83-9. (PMID: 20513215) Crit Care. 2006;10(6):R179. (PMID: 17192174) Crit Care Med. 1998 Nov;26(11):1834-41. (PMID: 9824076) Nat Biomed Eng. 2018 Oct;2(10):749-760. (PMID: 31001455) Crit Care Med. 2018 Mar;46(3):418-424. (PMID: 29474323) Eur Urol. 2018 Dec;74(6):796-804. (PMID: 30241973) Anaesthesiol Intensive Ther. 2015;47(5):457-63. (PMID: 26459228) Crit Care Med. 2016 Sep;44(9):1704-9. (PMID: 27071066) PLoS One. 2012;7(11):e48758. (PMID: 23144958) PLoS One. 2015 Mar 04;10(3):e0118432. (PMID: 25738806) Am J Bioeth. 2020 Nov;20(11):20-22. (PMID: 33103983) Artif Intell Med. 2019 Apr;95:27-37. (PMID: 30213670) J Crit Care. 2012 Aug;27(4):422.e1-9. (PMID: 22172798) |
| فهرسة مساهمة: | Keywords: decision support techniques; information dissemination; machine learning; mortality; patient discharge; patient readmission |
| تواريخ الأحداث: | Date Created: 20210930 Latest Revision: 20220427 |
| رمز التحديث: | 20231215 |
| مُعرف محوري في PubMed: | PMC8437217 |
| DOI: | 10.1097/CCE.0000000000000529 |
| PMID: | 34589713 |
| قاعدة البيانات: | MEDLINE |
Full Text Finder كن أول من يترك تعليقا!