Procedural aspects of compassionate care such as the terminal extubation are understudied. We used machine learning methods to determine factors associated with the decision to extubate the critically ill patient at the end of life, and whether the terminal extubation shortens the dying process. We performed a secondary data analysis of a large, prospective, multicentre, cohort study, death prediction and physiology after removal of therapy (DePPaRT), which collected baseline data as well as ECG, pulse oximeter and arterial waveforms from WLST until 30 min after death. We analysed a priori defined factors associated with the decision to perform terminal extubation in WLST using the random forest method and logistic regression. Cox regression was used to analyse the effect of terminal extubation on time from WLST to death. A total of 616 patients were included into the analysis, out of which 396 (64.3%) were terminally extubated. The study centre, low or no vasopressor support, and good respiratory function were factors significantly associated with the decision to extubate. Unadjusted time to death did not differ between patients with and without extubation (median survival time extubated vs. not extubated: 60 [95% CI: 46; 76] vs. 58 [95% CI: 45; 75] min). In contrast, after adjustment for confounders, time to death of extubated patients was significantly shorter (49 [95% CI: 40; 62] vs. 85 [95% CI: 61; 115] min). The decision to terminally extubate is associated with specific centres and less respiratory and/or vasopressor support. In this context, terminal extubation was associated with a shorter time to death.

Canadian Donation and Transplant Research Program Edmonton Canada

Children's Hospital of Eastern Ontario Research Institute Ottawa ON Canada

Department of Anaesthesia and Intensive Care Medicine 3rd Faculty of Medicine Charles University and University Hospital Královské Vinohrady Prague Czech Republic

Department of Intensive Care Medicine Maastricht University Medical Centre and School of Health Maastricht The Netherlands

Faculty of Medicine Division of Critical Care Respiratory Epidemiology and Clinical Research Unit McGill University Health Centre Montreal QC Canada

Maastricht University Medical Centre and Heart and Vascular Centre Maastricht The Netherlands

Ottawa Hospital Research Institute Ottawa ON Canada

Pediatric Critical Care Montreal Children's Hospital Montreal QC Canada

System Development Canadian Blood Services Ottawa ON Canada

See more in PubMed

Vincent JL, et al. Assessment of the worldwide burden of critical illness: The Intensive Care Over Nations (ICON) audit. Lancet Respir. Med. 2014;2:380–386. doi: 10.1016/S2213-2600(14)70061-X. PubMed DOI

Avidan A, et al. Variations in end-of-life practices in intensive care units worldwide (Ethicus-2): A prospective observational study. Lancet Respir. Med. 2021;9:1101–1110. doi: 10.1016/S2213-2600(21)00261-7. PubMed DOI

Sprung CL, et al. Changes in end-of-life practices in European intensive care units from 1999 to 2016. JAMA. 2019;322:1692–1704. doi: 10.1001/jama.2019.14608. PubMed DOI PMC

Efstathiou N, et al. Terminal withdrawal of mechanical ventilation in adult intensive care units: A systematic review and narrative synthesis of perceptions, experiences and practices. Palliat. Med. 2020;34:1140–1164. doi: 10.1177/0269216320935002. PubMed DOI

Robert R, et al. Terminal weaning or immediate extubation for withdrawing mechanical ventilation in critically ill patients (the ARREVE observational study) Intensive Care Med. 2017;43:1793–1807. doi: 10.1007/s00134-017-4891-0. PubMed DOI

Downar J, Delaney JW, Hawryluck L, Kenny L. Guidelines for the withdrawal of life-sustaining measures. Intensive Care Med. 2016;42:1003–1017. doi: 10.1007/s00134-016-4330-7. PubMed DOI

von Elm E, et al. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: Guidelines for reporting observational studies. BMJ. 2007;335:806–808. doi: 10.1136/bmj.39335.541782.AD. PubMed DOI PMC

Knaus W, Draper E, Wagner D, Zimmerman J. APACHE II: A severity of disease classification system. Crit Care Med. 1985;13:818–829. doi: 10.1097/00003246-198510000-00009. PubMed DOI

R Core Team. R: A Language and Environment for Statistical Computing (2021).

RStudio Team. RStudio: Integrated Development Environment for R (2021).

Tracy MF, Chlan L, Savik K, Skaar DJ, Weinert C. A novel research method for determining sedative exposure in critically ill patients. Nurs. Res. 2019;68:73–79. doi: 10.1097/NNR.0000000000000322. PubMed DOI PMC

Stekhoven DJ, Buehlmann P. MissForest—non-parametric missing value imputation for mixed-type data. Bioinformatics. 2012;28:112–118. doi: 10.1093/bioinformatics/btr597. PubMed DOI

Lang M, et al. mlr3: A modern object-oriented machine learning framework in R. J. Open Source Softw. 2019 doi: 10.21105/joss.01903. DOI

Wright MN, Ziegler A. ranger: A fast implementation of random forests for high dimensional data in C++ and R. J. Stat. Softw. 2017;77:1–17. doi: 10.18637/jss.v077.i01. DOI

Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J. Stat. Softw. 2010;33:1–22. doi: 10.18637/jss.v033.i01. PubMed DOI PMC

Brier GW. Verification of forecasts expressed in terms of probability. Mon. Weather Rev. 1950;78(1):1–3. doi: 10.1175/1520-0493(1950)078<0001:vofeit>2.0.co;2. DOI

Fisher A, Rudin C, Dominici F. All models are wrong, but many are useful: Learning a variable’s importance by studying an entire class of prediction models simultaneously. J. Mach. Learn. Res. 2018;20:1–81. PubMed PMC

Apley DW, Zhu J. Visualizing the effects of predictor variables in black box supervised learning models. J. R. Stat. Soc. Ser. B Stat. Methodol. 2020;82:1059–1086. doi: 10.1111/rssb.12377. DOI

Molnar C, Bischl B, Casalicchio G. iml: An R package for Interpretable Machine Learning. JOSS. 2018;3:786. doi: 10.21105/joss.00786. DOI

Denz, R., Klaaßen-Mielke, R. & Timmesfeld, N. A Comparison of Different Methods to Adjust Survival Curves for Confounders. http://arxiv.org/abs/2203.10002 (2022). PubMed

Sprung CL, et al. End-of-life practices in European intensive care units: The Ethicus study. JAMA. 2003;290:790–797. doi: 10.1001/jama.290.6.790. PubMed DOI

Suntharalingam C, Sharples L, Dudley C, Bradley JA, Watson CJE. Time to cardiac death after withdrawal of life-sustaining treatment in potential organ donors. Am. J. Transpl, 2009;9:2157–2165. doi: 10.1111/j.1600-6143.2009.02758.x. PubMed DOI

Wind J, et al. Prediction of time of death after withdrawal of life-sustaining treatment in potential donors after cardiac death*. Crit. Care Med. 2012;40:766–769. doi: 10.1097/CCM.0b013e318232e2e7. PubMed DOI

Friedman JH. Greedy function approximation: A gradient boosting machine. Ann. Stat. 2001;29:1189–1232. doi: 10.1214/aos/1013203451. DOI

Molnar, C. Interpretable Machine Learning (2022).

Molnar, C. et al. General pitfalls of model-agnostic interpretation methods for machine learning models. 10.48550/arxiv.2007.04131 (2020).

Li H, Wu P, Wang Z, Mao J, Alsaadi FE, Zeng N. A generalized framework of feature learning enhanced convolutional neural network for pathology-image-oriented cancer diagnosis. Comput. Biol. Med. 2022;151:106265. doi: 10.1016/j.compbiomed.2022.106265. PubMed DOI