Prediction of survival to discharge following cardiopulmonary resuscitation using classification and regression trees*

Mark H. Ebell, Anna M. Afonso, Romergryko G. Geocadin

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

OBJECTIVES:: To predict the likelihood that an inpatient who experiences cardiopulmonary arrest and undergoes cardiopulmonary resuscitation survives to discharge with good neurologic function or with mild deficits (Cerebral Performance Category score = 1). DESIGN:: Classification and Regression Trees were used to develop branching algorithms that optimize the ability of a series of tests to correctly classify patients into two or more groups. Data from 2007 to 2008 (n = 38,092) were used to develop candidate Classification and Regression Trees models to predict the outcome of inpatient cardiopulmonary resuscitation episodes and data from 2009 (n = 14,435) to evaluate the accuracy of the models and judge the degree of over fitting. Both supervised and unsupervised approaches to model development were used. SETTING:: 366 hospitals participating in the Get With the Guidelines-Resuscitation registry. SUBJECTS:: Adult inpatients experiencing an index episode of cardiopulmonary arrest and undergoing cardiopulmonary resuscitation in the hospital. MEASUREMENTS AND MAIN RESULTS:: The five candidate models had between 8 and 21 nodes and an area under the receiver operating characteristic curve from 0.718 to 0.766 in the derivation group and from 0.683 to 0.746 in the validation group. One of the supervised models had 14 nodes and classified 27.9% of patients as very unlikely to survive neurologically intact or with mild deficits (< 3%); the best unsupervised model had 11 nodes and classified 21.7% as very unlikely to survive. CONCLUSIONS:: We have developed and validated Classification and Regression Tree models that predict survival to discharge with good neurologic function or with mild deficits following in-hospital cardiopulmonary arrest. Models like this can assist physicians and patients who are considering do-not-resuscitate orders.

Original languageEnglish (US)
Pages (from-to)2688-2697
Number of pages10
JournalCritical care medicine
Volume41
Issue number12
DOIs
StatePublished - Dec 1 2013

Fingerprint

Cardiopulmonary Resuscitation
Heart Arrest
Inpatients
Nervous System
Survival
Resuscitation Orders
Resuscitation
ROC Curve
Registries
Guidelines
Physicians

Keywords

  • Cardiopulmonary arrest
  • Clinical prediction models
  • Do-not-resuscitate order
  • Medical futility
  • Resuscitation

ASJC Scopus subject areas

  • Critical Care and Intensive Care Medicine

Cite this

Prediction of survival to discharge following cardiopulmonary resuscitation using classification and regression trees*. / Ebell, Mark H.; Afonso, Anna M.; Geocadin, Romergryko G.

In: Critical care medicine, Vol. 41, No. 12, 01.12.2013, p. 2688-2697.

Research output: Contribution to journalArticle

Ebell, Mark H. ; Afonso, Anna M. ; Geocadin, Romergryko G. / Prediction of survival to discharge following cardiopulmonary resuscitation using classification and regression trees*. In: Critical care medicine. 2013 ; Vol. 41, No. 12. pp. 2688-2697.
@article{43a7eec1c28142438a4c91e7c3bc2650,
title = "Prediction of survival to discharge following cardiopulmonary resuscitation using classification and regression trees*",
abstract = "OBJECTIVES:: To predict the likelihood that an inpatient who experiences cardiopulmonary arrest and undergoes cardiopulmonary resuscitation survives to discharge with good neurologic function or with mild deficits (Cerebral Performance Category score = 1). DESIGN:: Classification and Regression Trees were used to develop branching algorithms that optimize the ability of a series of tests to correctly classify patients into two or more groups. Data from 2007 to 2008 (n = 38,092) were used to develop candidate Classification and Regression Trees models to predict the outcome of inpatient cardiopulmonary resuscitation episodes and data from 2009 (n = 14,435) to evaluate the accuracy of the models and judge the degree of over fitting. Both supervised and unsupervised approaches to model development were used. SETTING:: 366 hospitals participating in the Get With the Guidelines-Resuscitation registry. SUBJECTS:: Adult inpatients experiencing an index episode of cardiopulmonary arrest and undergoing cardiopulmonary resuscitation in the hospital. MEASUREMENTS AND MAIN RESULTS:: The five candidate models had between 8 and 21 nodes and an area under the receiver operating characteristic curve from 0.718 to 0.766 in the derivation group and from 0.683 to 0.746 in the validation group. One of the supervised models had 14 nodes and classified 27.9{\%} of patients as very unlikely to survive neurologically intact or with mild deficits (< 3{\%}); the best unsupervised model had 11 nodes and classified 21.7{\%} as very unlikely to survive. CONCLUSIONS:: We have developed and validated Classification and Regression Tree models that predict survival to discharge with good neurologic function or with mild deficits following in-hospital cardiopulmonary arrest. Models like this can assist physicians and patients who are considering do-not-resuscitate orders.",
keywords = "Cardiopulmonary arrest, Clinical prediction models, Do-not-resuscitate order, Medical futility, Resuscitation",
author = "Ebell, {Mark H.} and Afonso, {Anna M.} and Geocadin, {Romergryko G.}",
year = "2013",
month = "12",
day = "1",
doi = "10.1097/CCM.0b013e31829a708c",
language = "English (US)",
volume = "41",
pages = "2688--2697",
journal = "Critical Care Medicine",
issn = "0090-3493",
publisher = "Lippincott Williams and Wilkins",
number = "12",

}

TY - JOUR

T1 - Prediction of survival to discharge following cardiopulmonary resuscitation using classification and regression trees*

AU - Ebell, Mark H.

AU - Afonso, Anna M.

AU - Geocadin, Romergryko G.

PY - 2013/12/1

Y1 - 2013/12/1

N2 - OBJECTIVES:: To predict the likelihood that an inpatient who experiences cardiopulmonary arrest and undergoes cardiopulmonary resuscitation survives to discharge with good neurologic function or with mild deficits (Cerebral Performance Category score = 1). DESIGN:: Classification and Regression Trees were used to develop branching algorithms that optimize the ability of a series of tests to correctly classify patients into two or more groups. Data from 2007 to 2008 (n = 38,092) were used to develop candidate Classification and Regression Trees models to predict the outcome of inpatient cardiopulmonary resuscitation episodes and data from 2009 (n = 14,435) to evaluate the accuracy of the models and judge the degree of over fitting. Both supervised and unsupervised approaches to model development were used. SETTING:: 366 hospitals participating in the Get With the Guidelines-Resuscitation registry. SUBJECTS:: Adult inpatients experiencing an index episode of cardiopulmonary arrest and undergoing cardiopulmonary resuscitation in the hospital. MEASUREMENTS AND MAIN RESULTS:: The five candidate models had between 8 and 21 nodes and an area under the receiver operating characteristic curve from 0.718 to 0.766 in the derivation group and from 0.683 to 0.746 in the validation group. One of the supervised models had 14 nodes and classified 27.9% of patients as very unlikely to survive neurologically intact or with mild deficits (< 3%); the best unsupervised model had 11 nodes and classified 21.7% as very unlikely to survive. CONCLUSIONS:: We have developed and validated Classification and Regression Tree models that predict survival to discharge with good neurologic function or with mild deficits following in-hospital cardiopulmonary arrest. Models like this can assist physicians and patients who are considering do-not-resuscitate orders.

AB - OBJECTIVES:: To predict the likelihood that an inpatient who experiences cardiopulmonary arrest and undergoes cardiopulmonary resuscitation survives to discharge with good neurologic function or with mild deficits (Cerebral Performance Category score = 1). DESIGN:: Classification and Regression Trees were used to develop branching algorithms that optimize the ability of a series of tests to correctly classify patients into two or more groups. Data from 2007 to 2008 (n = 38,092) were used to develop candidate Classification and Regression Trees models to predict the outcome of inpatient cardiopulmonary resuscitation episodes and data from 2009 (n = 14,435) to evaluate the accuracy of the models and judge the degree of over fitting. Both supervised and unsupervised approaches to model development were used. SETTING:: 366 hospitals participating in the Get With the Guidelines-Resuscitation registry. SUBJECTS:: Adult inpatients experiencing an index episode of cardiopulmonary arrest and undergoing cardiopulmonary resuscitation in the hospital. MEASUREMENTS AND MAIN RESULTS:: The five candidate models had between 8 and 21 nodes and an area under the receiver operating characteristic curve from 0.718 to 0.766 in the derivation group and from 0.683 to 0.746 in the validation group. One of the supervised models had 14 nodes and classified 27.9% of patients as very unlikely to survive neurologically intact or with mild deficits (< 3%); the best unsupervised model had 11 nodes and classified 21.7% as very unlikely to survive. CONCLUSIONS:: We have developed and validated Classification and Regression Tree models that predict survival to discharge with good neurologic function or with mild deficits following in-hospital cardiopulmonary arrest. Models like this can assist physicians and patients who are considering do-not-resuscitate orders.

KW - Cardiopulmonary arrest

KW - Clinical prediction models

KW - Do-not-resuscitate order

KW - Medical futility

KW - Resuscitation

UR - http://www.scopus.com/inward/record.url?scp=84889244174&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84889244174&partnerID=8YFLogxK

U2 - 10.1097/CCM.0b013e31829a708c

DO - 10.1097/CCM.0b013e31829a708c

M3 - Article

AN - SCOPUS:84889244174

VL - 41

SP - 2688

EP - 2697

JO - Critical Care Medicine

JF - Critical Care Medicine

SN - 0090-3493

IS - 12

ER -