Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Sep 28;12(19):6243.
doi: 10.3390/jcm12196243.

Predicting Survival for Veno-Arterial ECMO Using Conditional Inference Trees-A Multicenter Study

Julia Braun  1 , Sebastian D Sahli  2 , Donat R Spahn  2 , Daniel Röder  3 , Holger Neb  4 , Gösta Lotz  4 , Raed Aser  5 , Markus J Wilhelm  5 , Alexander Kaserer  2
Affiliations

Predicting Survival for Veno-Arterial ECMO Using Conditional Inference Trees-A Multicenter Study

Julia Braun et al. J Clin Med. .

Abstract

Background: Despite increasing use and understanding of the process, veno-arterial extracorporeal membrane oxygenation (VA-ECMO) therapy is still associated with considerable mortality. Personalized and quick survival predictions using machine learning methods can assist in clinical decision making before ECMO insertion.

Methods: This is a multicenter study to develop and validate an easy-to-use prognostic model to predict in-hospital mortality of VA-ECMO therapy, using unbiased recursive partitioning with conditional inference trees. We compared two sets with different numbers of variables (small and comprehensive), all of which were available just before ECMO initiation. The area under the curve (AUC), the cross-validated Brier score, and the error rate were applied to assess model performance. Data were collected retrospectively between 2007 and 2019.

Results: 837 patients were eligible for this study; 679 patients in the derivation cohort (median (IQR) age 60 (49 to 69) years; 187 (28%) female patients) and a total of 158 patients in two external validation cohorts (median (IQR) age 57 (49 to 65) and 70 (63 to 76) years). For the small data set, the model showed a cross-validated error rate of 35.79% and an AUC of 0.70 (95% confidence interval from 0.66 to 0.74). In the comprehensive data set, the error rate was the same with a value of 35.35%, with an AUC of 0.71 (95% confidence interval from 0.67 to 0.75). The mean Brier scores of the two models were 0.210 (small data set) and 0.211 (comprehensive data set). External validation showed an error rate of 43% and AUC of 0.60 (95% confidence interval from 0.52 to 0.69) using the small tree and an error rate of 35% with an AUC of 0.63 (95% confidence interval from 0.54 to 0.72) using the comprehensive tree. There were large differences between the two validation sets.

Conclusions: Conditional inference trees are able to augment prognostic clinical decision making for patients undergoing ECMO treatment. They may provide a degree of accuracy in mortality prediction and prognostic stratification using readily available variables.

Keywords: ECLS; VA ECMO; conditional inference trees; machine learning; predictors; unbiased recursive partitioning.

PubMed Disclaimer

Conflict of interest statement

J.B., S.D.S., D.R., H.N., G.L., R.A., and M.J.W. have no conflicts of interest to declare. DRS’ former academic department is receiving grant support from the Swiss National Science Foundation, Berne, Switzerland; the Swiss Society of Anesthesiology and Perioperative Medicine (SSAPM), Berne, Switzerland; the Swiss Foundation for Anesthesia Research, Zurich, Switzerland; and CSL Vifor (International) AG, St. Gallen, Switzerland. DRS is co-chair of the ABC-Trauma Faculty, sponsored by unrestricted educational grants from Alexion Pharma Germany GmbH, Munich, Germany; CSL Behring GmbH, Marburg, Germany; and LFB Biomédicaments, Courtaboeuf Cedex, France. DRS received honoraria/travel support for consulting or lecturing from: Alliance Rouge, Bern, Switzerland; Danube University of Krems, Austria; European Society of Anesthesiology and Intensive Care, Brussels, Belgium; International Foundation for Patient Blood Management, Basel, Switzerland; Korean Society of Anesthesiologists, Seoul, Korea; Network for the Advancement of Patient Blood Management, Haemostasis and Thrombosis, Paris, France; Society for the Advancement of Blood Management, Mount Royal NJ, Alexion Pharmaceuticals Inc., Boston, MA, USA; AstraZeneca AG, Baar, Switzerland; Bayer AG, Zürich, Switzerland; B. Braun Melsungen AG, Melsungen, Germany; Baxter AG, Glattpark, Switzerland; CSL Behring GmbH, Hattersheim am Main, Germany, and Berne, Switzerland; CSL Vifor (Switzerland) Villars-sur-Glâne, Switzerland; CSL Vifor (International), St. Gallen, Switzerland; Celgene International II Sàrl, Couvet, Switzerland; Daiichi Sankyo AG, Thalwil, Switzerland; Haemonetics, Braintree, MA, USA; iSEP, Nantes, France; LFB Biomédicaments, Courtaboeuf Cedex, France; Merck Sharp & Dohme, Kenilworth, New Jersey, USA; Novo Nordisk Health Care AG, Zurich, Switzerland; Octapharma AG, Lachen, Switzerland; Pharmacosmos A/S, Holbaek, Denmark; Pierre Fabre Pharma, Alschwil, Switzerland; Portola Schweiz GmbH, Aarau, Switzerland; Roche Diagnostics International Ltd., Reinach, Switzerland; Sarstedt AG & Co., Sevelen, Switzerland, and Nümbrecht, Germany; Shire Switzerland GmbH, Zug, Switzerland; Takeda, Glattpark, Switzerland; Werfen, Bedford, MA, USA; Zuellig Pharma Holdings, Singapore, Singapore. AK has received support from Bayer AG (Switzerland) and CSL Behring GmBH (Switzerland) for lecturing.

Figures

Figure 1
Figure 1
Decision Algorithm based on the Small Data Set. Numbers in square brackets represent the node numbers. Indications for ECMO therapy: postcardiotomy (PCT), cardiopulmonary resuscitation (CPR), refractory cardiogenic shock (RCS), and other. The category “other” included ECMO indications for lung transplantation and expansive thoracic surgery.
Figure 2
Figure 2
Decision Algorithm based on the Comprehensive Data Set. Numbers in square brackets represent the node numbers. Indications for ECMO therapy: postcardiotomy (PCT), cardiopulmonary resuscitation (CPR), refractory cardiogenic shock (RCS), and other. The category “other” included ECMO indications for lung transplantation and expansive thoracic surgery.
Figure 3
Figure 3
Calibration plots for ten-fold cross-validation. Left: derivation cohort (a) small and (b) comprehensive data set. Right: external validation variables (c) small and (d) comprehensive data set.
Figure 4
Figure 4
Receiver Operating Characteristic (ROC) curve (external validation) of the small ((a), left) and comprehensive ((b), right) decision trees. The predictions from the small tree (a) result in an Area Under the Curve (AUC) of 0.60 (95% confidence interval from 0.52 to 0.69), and the AUC from the comprehensive tree (b) is 0.63 (95% confidence interval from 0.54 to 0.72).
See this image and copyright information in PMC

References

    1. Eckman P.M., Katz J.N., El Banayosy A., Bohula E.A., Sun B., van Diepen S. Veno-Arterial Extracorporeal Membrane Oxygenation for Cardiogenic Shock: An Introduction for the Busy Clinician. Circulation. 2019;140:2019–2037. doi: 10.1161/CIRCULATIONAHA.119.034512. - DOI - PubMed
    1. Pineton de Chambrun M., Bréchot N., Combes A. Venoarterial extracorporeal membrane oxygenation in cardiogenic shock: Indications, mode of operation, and current evidence. Curr. Opin. Crit. Care. 2019;25:397–402. doi: 10.1097/MCC.0000000000000627. - DOI - PubMed
    1. Chung M., Zhao Y., Strom J.B., Shen C., Yeh R.W. Extracorporeal Membrane Oxygenation Use in Cardiogenic Shock: Impact of Age on In-Hospital Mortality, Length of Stay, and Costs. Crit. Care Med. 2019;47:e214–e221. doi: 10.1097/CCM.0000000000003631. - DOI - PMC - PubMed
    1. Fernando S.M., Qureshi D., Tanuseputro P., Fan E., Munshi L., Rochwerg B., Talarico R., Scales D.C., Brodie D., Dhanani S., et al. Mortality and costs following extracorporeal membrane oxygenation in critically ill adults: A population-based cohort study. Intensive Care Med. 2019;45:1580–1589. doi: 10.1007/s00134-019-05766-z. - DOI - PubMed
    1. Chen W.C., Huang K.Y., Yao C.W., Wu C.F., Liang S.J., Li C.H., Tu C.Y., Chen H.J. The modified SAVE score: Predicting survival using urgent veno-arterial extracorporeal membrane oxygenation within 24 hours of arrival at the emergency department. Crit. Care. 2016;20:336. doi: 10.1186/s13054-016-1520-1. - DOI - PMC - PubMed

LinkOut - more resources