0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Original Investigation |

Development and Validation of the Good Outcome Following Attempted Resuscitation (GO-FAR) Score to Predict Neurologically Intact Survival After In-Hospital Cardiopulmonary Resuscitation FREE

Mark H. Ebell, MD, MS1,2; Woncheol Jang, PhD3; Ye Shen, PhD1; Romergryko G. Geocadin, MD4; for the Get With the Guidelines–Resuscitation Investigators
[+] Author Affiliations
1Department of Epidemiology and Biostatistics, College of Public Health, University of Georgia, Athens
2Institute for Evidence-Based Practice in the Health Professions, University of Georgia, Athens
3Department of Statistics, Seoul National University, Seoul, Republic of Korea
4Departments of Neurology, Anesthesiology–Critical Care Medicine, and Neurosurgery, Johns Hopkins University School of Medicine, Baltimore, Maryland
JAMA Intern Med. 2013;173(20):1872-1878. doi:10.1001/jamainternmed.2013.10037.
Text Size: A A A
Published online

Importance  Informing patients and providers of the likelihood of survival after in-hospital cardiac arrest (IHCA), neurologically intact or with minimal deficits, may be useful when discussing do-not-attempt-resuscitation orders.

Objective  To develop a simple prearrest point score that can identify patients unlikely to survive IHCA, neurologically intact or with minimal deficits.

Design, Setting, and Participants  The study included 51 240 inpatients experiencing an index episode of IHCA between January 1, 2007, and December 31, 2009, in 366 hospitals participating in the Get With the Guidelines–Resuscitation registry. Dividing data into training (44.4%), test (22.2%), and validation (33.4%) data sets, we used multivariate methods to select the best independent predictors of good neurologic outcome, created a series of candidate decision models, and used the test data set to select the model that best classified patients as having a very low (<1%), low (1%-3%), average (>3%-15%), or higher than average (>15%) likelihood of survival after in-hospital cardiopulmonary resuscitation for IHCA with good neurologic status. The final model was evaluated using the validation data set.

Main Outcomes and Measures  Survival to discharge after in-hospital cardiopulmonary resuscitation for IHCA with good neurologic status (neurologically intact or with minimal deficits) based on a Cerebral Performance Category score of 1.

Results  The best performing model was a simple point score based on 13 prearrest variables. The C statistic was 0.78 when applied to the validation set. It identified the likelihood of a good outcome as very low in 9.4% of patients (good outcome in 0.9%), low in 18.9% (good outcome in 1.7%), average in 54.0% (good outcome in 9.4%), and above average in 17.7% (good outcome in 27.5%). Overall, the score can identify more than one-quarter of patients as having a low or very low likelihood of survival to discharge, neurologically intact or with minimal deficits after IHCA (good outcome in 1.4%).

Conclusions and Relevance  The Good Outcome Following Attempted Resuscitation (GO-FAR) scoring system identifies patients who are unlikely to benefit from a resuscitation attempt should they experience IHCA. This information can be used as part of a shared decision regarding do-not-attempt-resuscitation orders.

Figures in this Article

In-hospital cardiac arrest (IHCA) is common in the inpatient setting. In 1 study of 13 400 patients admitted over 2 years at a teaching hospital, approximately 2.5% experienced an episode of IHCA.1 Data from more than 400 hospitals participating in Get With the Guidelines Registry–Resuscitation (GWTG-R) found that there are an estimated 192 000 treated episodes of IHCA annually in US hospitals, an event rate of 0.92 per 1000 patient days.2Overall survival to discharge after treated IHCA is approximately 18%, with about half of those patients neurologically intact or with only mild neurologic deficits at discharge.3,4 Because IHCA is common and has potentially devastating outcomes, discussion of patient wishes regarding resuscitation or “code status” is important. Do-not-attempt-resuscitation (DNAR) orders are written when cardiopulmonary resuscitation (CPR) is unlikely to meet the patient’s treatment goals of long-term, neurologically intact survival.5 Presumably, information about prognosis would be helpful as part of a shared decision regarding DNAR orders.

When patients and their physicians are discussing a DNAR order, a question that may arise is, “If I experience cardiopulmonary arrest and receive CPR, how likely am I to survive to discharge?” However, previous attempts to develop clinical decision rules that predict survival to discharge after IHCA have been limited by small sample size, inclusion of periarrest variables not knowable before arrest, and inconsistent validation.613

In this study, we use data from the national GWTG-R of IHCA episodes (formerly the National Registry of Cardiopulmonary Resuscitation) to develop and validate a clinical decision rule for prediction of survival to discharge with good neurologic status. Patients and their physicians could potentially use this information when making decisions about DNAR orders.

Population Studied

The GWTG-R has gathered data on more than 160 000 episodes of IHCA in children and adults at 366 hospitals in the United States, beginning in 2000. Details of this data set, including data collection methods using the Utstein template,14 have been published elsewhere.3 We limited our analysis to adults experiencing an initial (index) episode of pulseless IHCA between January 1, 2007, and December 31, 2009. Index arrests were used because they are most relevant when discussing DNAR orders at admission, and only the most recent 3 full years of data were used to most closely reflect current outcomes and clinical practice. We excluded survivors who did not have an assessment of cerebral performance at discharge (3787 of all 27 088 survivors [14.0%]). The final cohort for this study had 51 240 patients and was randomly divided into data sets for training (44.4% of cases), testing (22.2%), and final validation (33.4%) of the clinical decision rule. The mean age of patients experiencing IHCA was 65 years, 58.3% were male, and of those whose race was known, 72.8% were white, 23.3% were black, and the rest were another race. The mean duration of resuscitations was 20.8 minutes.

Outcome (Dependent) Variables

The Cerebral Performance Category (CPC) score was measured as part of the GWTG-R data set at the time of admission and at hospital discharge for survivors. A CPC score of 1 represents good cerebral performance. The patient is conscious, alert, and able to work but might have mild neurologic or psychological deficits (such as mild dysphagia or minor cranial nerve abnormalities). Patients with a CPC score of 2 have moderate cerebral disability and are able to live independently and work in a sheltered environment. Disabilities may include hemiplegia, seizures, ataxia, dysphagia, or permanent memory or mental changes. Patients with CPC scores of 3 through 5 progress through severe cerebral disability, coma or vegetative state, and finally brain death. In a previous study of IHCA survivors, this scoring system was shown to have good validity and reliability.15 The dependent variable was a discharge CPC score other than 1.

Predictor (Independent) Variables

Bivariate analysis used a Pearson χ2 test. Only variables knowable before IHCA were included in our study because periarrest variables (eg, duration of resuscitation, location of arrest, or initial rhythm) cannot be used to inform DNAR decision making before the event. A previous meta-analysis by Ebell and Afonso4 identified the following prearrest predictors of failure to survive: metastatic or hematologic cancer (odds ratio [OR], 3.9-4.8), age more than 70, 75, or 80 years (OR, 1.5-2.7), black race (OR, 2.1), altered mental status (OR, 2.2), dependency for activities of daily living (OR, 3.2-7.0), impaired renal function (OR, 1.9), hypotension on admission (OR, 1.8), and admission for pneumonia (OR, 1.7), trauma (OR, 1.7), or medical noncardiac diagnosis (OR, 2.2). Cardiovascular diagnoses and comorbid conditions were associated with improved survival (OR, 0.23-0.53).

The GWTG-R data set included all of the above variables. In addition, we included a number of other prearrest variables from GWTG-R based on clinical reasoning: acute central nervous system (CNS) nonstroke event, acute stroke, arrhythmia, depressed CNS status, congestive heart failure at this or a prior admission, diabetes mellitus, hepatic insufficiency, human immunodeficiency virus positivity, AIDS, metabolic or electrolyte abnormality, myocardial infarction at this or a previous admission, pneumonia, renal insufficiency or dialysis, respiratory insufficiency, septicemia, CPC score at admission, residence before admission (home vs other), and illness category (ie, medical noncardiac). The specific definitions used by GWTG-R are summarized for the final set of predictor variables in Table 1. Because the focus of this study was on individual risk factors, hospital characteristics were not included as predictors, nor was the location at the time of resuscitation because that might differ from the location at the time of a discussion about DNAR orders. Although African American race was associated with a lower likelihood of survival, we thought this was probably due to differences in resuscitation quality rather than inherent biological factors.9,16 Depressed CNS status on admission was not included as a predictor variable because this information was captured more reliably by the CPC score at admission. Age was categorized as less than 70, 70 to 74, 75 to 79, 80 to 84, or 85 years or more based on the results of a previous meta-analysis.4 Interaction terms were explored based on previous meta-analyses, including age × cancer and age × renal insufficiency.

Table Graphic Jump LocationTable 1.  Definitions of Predictor Variablesa
Multivariate Analysis

With a large data set such as the GWTG-R, many predictor variables will have a statistically significant association with survival to discharge. However, these associations may not be clinically significant, defined as a clinically meaningful difference in survival between those with and without the clinical characteristic. Therefore, in addition to the variables identified in the previous meta-analysis, only those with an absolute survival difference of at least 3% between patients with and without the variable of interest were included as candidate predictor variables for the multivariate analysis. Clinical characteristics present in less than 3% of patients, such as a diagnosis of AIDS, were also excluded in the interest of parsimony. The final list of predictor variables for the multivariate analysis is shown in Table 2.

Table Graphic Jump LocationTable 2.  Predictor Variables Initially Included in Multivariate Model Based on Bivariate Analysis

We used 2 approaches to identify a parsimonious set of independent variables for the final multivariate prediction model, the Bayesian information criterion (BIC) and the least absolute shrinkage and selection operator (LASSO).1719 Once the variables were selected, a second logistic regression was performed using only those selected predictor variables. We then used the β coefficients of each model to create complex and simplified point scores for each approach.

Point Score Development and Evaluation

For each of the 4 candidate point scores, we identified cutoffs using the training data set to create the following groups according to the likelihood of survival to discharge neurologically intact or with minimal deficit: very low (<1%), low (1%-3%), average (>3%-15%), and higher than average (>15%). Because medical futility has previously been defined both as a likelihood of success less than 3%20 and as a likelihood of success less than 1%,21 we present data for both cutoffs. We evaluated the 4 candidate point scores using the test data set to identify the best performing point score, based on our judgment regarding parsimony, simplicity, the greatest number of patients classified in the very low and low survival groups, and the least loss in accuracy when applied to the test set. The accuracy of this final point score was evaluated using the validation data set, which had been held aside and not used in the model development process.

We used Stata software (version 11.0; StataCorp) for bivariate analysis (Pearson χ2 for dichotomous variables), for logistic regression, and to determine the area under the receiver operating characteristic (ROC) curve and the Hosmer-Lemeshow statistic for each model. The area under the ROC curve is also called the C statistic. We used Microsoft Excel software (Microsoft) to calculate the BIC and the predictive values and likelihood ratios for each point score, and R software (version 2.14; R Project for Statistical Computing) to calculate the least absolute shrinkage and selection operator.

This study was approved by the Human Subjects Committee of the University of Georgia.

The overall rate of survival to discharge with a CPC score of 1 was 10.4%. The bivariate analysis is shown in Table 2; variables were excluded if they were present in less than 3% of patients or if there was less than a 3% survival difference between those with and without the clinical characteristic of interest.

Using the BIC to guide variable selection resulted in a final model with 13 variables (Table 2). We performed a second logistic regression using only those 13 final independent variables; the β coefficients for each predictor are shown in Table 3. The full model had a C statistic of 0.800 and a Hosmer-Lemeshow χ28 statistic of 11.39 (P = .18); the Hosmer-Lemeshow graph is shown in the Figure and shows evidence of good calibration. The β coefficients were multiplied by 10 and rounded to create the final score shown in Table 3.

Table Graphic Jump LocationTable 3.  Results of Multivariate Analysis for Final Modela
Place holder to copy figure label and caption
Figure.
Hosmer-Lemeshow Graph for the Full Model

The Hosmer-Lemeshow graph is shown for the full logistic regression model, which uses variable selection based on the Bayesian information criterion.

Graphic Jump Location

Table 4 shows the performance of the final point score in the training, test, and validation data sets, as well as for the entire data set. The score performed similarly in each data set, with no loss of overall accuracy as measured by the C statistic from training and test sets (0.77) to validation set (0.78) (see eFigure in the Supplement for the ROC curve). Compared with the other candidate scores evaluated by using the test data set (eTable in the Supplement), the final score classified the highest percentage of patients in both the low and very low likelihood groups and had the lowest percentage of survival to discharge with a CPC score of 1 in those groups. We call this new clinical prediction rule the Good Outcome Following Attempted Resuscitation (GO-FAR) score.

Table Graphic Jump LocationTable 4.  Performance of the Final GO-FAR Score in the Training, Test, and Validation Sets

In the validation data set, the GO-FAR score identified 9.4% and 18.9% of patients, respectively, as having a very low or low likelihood of a good outcome (0.9% or 1.7% survival to discharge with a CPC score of 1). Overall, the GO-FAR score identified 28.3% of patients as having a low or very low likelihood of a good outcome, with only 1.4% surviving to discharge with a CPC score of 1 in this group.

We have developed and validated a simple scoring system, the GO-FAR score, that can identify hospitalized patients having a very low, low, average, or higher than average likelihood of surviving to discharge neurologically intact or with minimal deficits following CPR for IHCA. This outcome was defined as a CPC score of 1, which means the patient can lead a normal life and has only minimal neurologic or psychological deficits (eg, mild dysphagia, nonincapacitating hemiparesis, or minor cranial nerve abnormalities). This information could be useful when counseling patients regarding their DNAR status. Because the clinical prediction rule uses information that is known at the time of hospital admission, it could also be built into the admissions process and used to identify patients who have little possibility of benefitting from CPR. Previous research has shown that patients significantly overestimate their likelihood of survival to discharge after CPR for IHCA. In 2 large surveys of community-dwelling adults, participants estimated their likelihood of survival to discharge after in-hospital CPR at more than 50% and did not seem to understand that CPR routinely involved intubation and cardioversion.22,23 Although physicians could provide guidance regarding the likelihood of survival to discharge, a previous study found that the physicians’ survival estimates were inaccurate.24

Our study had several limitations. First, outcomes at the hospitals participating in the GWTG-R may differ from those of nonparticipating hospitals. However, the large number of hospitals from diverse communities makes that unlikely. Second, we were limited by the clinical variables gathered at admission by the GWTG-R. Other factors (eg, nonmetastatic cancer and serum albumin and hemoglobin levels) may be important predictors but could not be included in our model. Finally, we had data only for patients who actually underwent CPR and had an assessment of CPC score at discharge. Patients who already had a DNAR order written would not become part of the GWTG-R data set, creating a possible systematic bias. For example, if patients with metastatic cancer and poor functional status generally received a DNAR order on admission and those with metastatic cancer but good functional status who were undergoing active treatment did not, the association of this variable with the final outcome might be attenuated. Although one could gather information on all patients rather than just those who undergo CPR, this would not help answer our question because we cannot know patients’ outcomes after CPR unless they actually undergo CPR after cardiopulmonary arrest.

The strengths of the study include the large, diverse sample; a rigorous process for developing the model and determining the cutoffs for risk groups without using the final validation set; and the use of only the prearrest variables known at the time of hospital admission.

The Patient Self-Determination Act of 199025 requires that patients admitted to a health care facility in the United States be asked about the presence of an advance directive such as a durable power of attorney or living will. In some cases, this may include a preference for a DNAR order. The GO-FAR score could be used to provide objective information about the expected outcome after in-hospital CPR, which would be integrated with the patient’s values and expectations to optimize decision making about DNAR orders. We do not intend to make a value judgment for patients. In fact, we expect that some patients may decide that resuscitation should be attempted regardless of the expected outcome, and many patients may consider survival with moderate or even severe neurologic impairment to be a good or acceptable outcome.

We are developing a web-based tool that could be used to provide patients with general information about CPR and its typical outcomes, assess their values and preferences, and generate patient-specific estimates of outcome (very low, low, average, or higher than average likelihood of surviving after CPR for IHCA neurologically intact or with minimal deficits). In the validation group, patients classified as having a very low or low survival had only a 1.4% likelihood of survival to discharge neurologically intact or with minimal deficit. This information could be shared with a physician or nurse to facilitate a discussion about DNAR orders.

In summary, we have developed and successfully validated a simple clinical rule for predicting survival after inpatient CPR, neurologically intact or with minimal deficits. The GO-FAR score identified 9.4% of patients as having a very low likelihood of a good outcome after CPR (good outcome in 0.9% [14 of 1624]) and another 18.9% as having a low likelihood (good outcome in 1.7% [54 of 3247]). We believe that this scoring system can provide useful information for patients and physicians and, when integrated with information on patients’ values, beliefs, and goals, can productively inform end-of-life decisions and reduce unnecessary suffering.

Accepted for Publication: June 25, 2013.

Corresponding Author: Mark H. Ebell, MD, MS, Department of Epidemiology and Biostatistics, College of Public Health, University of Georgia, 233 Miller Hall, UGA Health Sciences Campus, Athens, GA 30602 (ebell@uga.edu).

Published Online: September 9, 2013. doi:10.1001/jamainternmed.2013.10037.

Author Contributions:Study concept and design: Ebell, Geocadin.

Analysis and interpretation of data: All authors.

Drafting of the manuscript: Ebell, Jang.

Critical revision of the manuscript for important intellectual content: Ebell, Shen, Geocadin.

Statistical analysis: Ebell, Jang, Shen.

Administration, technical, and material support: Ebell, Geocadin.

Study supervision: Ebell, Geocadin.

Conflict of Interest Disclosures: None reported.

Group Information: The Get With the Guidelines–Resuscitation Investigators include Paul Chan, MD, MSc, Saint Luke’s Mid America Heart Institute, Kansas City, MO; Comilla Sasson, MD, MS, and Steven Bradley, MD, MPH, University of Colorado, Denver; Michael W. Donnino, MD, Beth Israel Deaconess Medical Center, Boston, MA; Dana P. Edelson, MD, MS, University of Chicago, Chicago IL; Robert T. Faillace, MD, ScM, Geisinger Healthcare System, Danville, PA; Raina Merchant, MD, MSHP, University of Pennsylvania School of Medicine, Philadelphia; Vincent N. Mosesso Jr, MD, University of Pittsburgh School of Medicine, Pittsburgh, PA; Joseph P. Ornato, MD, and Mary Ann Peberdy, MD, Virginia Commonwealth University, Richmond.

Previous Presentation: This study was presented at the annual meeting of the North American Primary Care Research Group; November 9-12, 2013; Ottawa, Ontario, Canada.

Saklayen  M, Liss  H, Markert  R.  In-hospital cardiopulmonary resuscitation: survival in 1 hospital and literature review. Medicine (Baltimore). 1995;74(4):163-175.
PubMed   |  Link to Article
Merchant  RM, Yang  L, Becker  LB,  et al; American Heart Association Get With the Guidelines-Resuscitation Investigators.  Incidence of treated cardiac arrest in hospitalized patients in the United States. Crit Care Med. 2011;39(11):2401-2406.
PubMed   |  Link to Article
Peberdy  MA, Kaye  W, Ornato  JP,  et al.  Cardiopulmonary resuscitation of adults in the hospital: a report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation. 2003;58(3):297-308.
PubMed   |  Link to Article
Ebell  MH, Afonso  AM.  Pre-arrest predictors of failure to survive after in-hospital cardiopulmonary resuscitation: a meta-analysis. Fam Pract. 2011;28(5):505-515. doi:10.1093/fampra/cmr023.
PubMed   |  Link to Article
Ebell  MH, Doukas  DJ, Smith  MA.  The do-not-resuscitate order: a comparison of physician and patient preferences and decision-making. Am J Med. 1991;91(3):255-260.
PubMed   |  Link to Article
Ebell  MH, Preston  PS.  The effect of the APACHE II score and selected clinical variables on survival following cardiopulmonary resuscitation. Fam Med. 1993;25(3):191-196.
PubMed
O’Keeffe  S, Ebell  MH.  Prediction of failure to survive following in-hospital cardiopulmonary resuscitation: comparison of two predictive instruments. Resuscitation. 1994;28(1):21-25.
PubMed   |  Link to Article
Ebell  MH, Kruse  JA, Smith  M, Novak  J, Drader-Wilcox  J.  Failure of three decision rules to predict the outcome of in-hospital cardiopulmonary resuscitation. Med Decis Making. 1997;17(2):171-177.
PubMed   |  Link to Article
Ebell  MH, Smith  M, Kruse  JA, Drader-Wilcox  J, Novak  J.  Effect of race on survival following in-hospital cardiopulmonary resuscitation. J Fam Pract. 1995;40(6):571-577.
PubMed
George  AL  Jr, Folk  BP  III, Crecelius  PL, Campbell  WB.  Pre-arrest morbidity and other correlates of survival after in-hospital cardiopulmonary arrest. Am J Med. 1989;87(1):28-34.
PubMed   |  Link to Article
Bowker  L, Stewart  K.  Predicting unsuccessful cardiopulmonary resuscitation (CPR): a comparison of three morbidity scores. Resuscitation. 1999;40(2):89-95.
PubMed   |  Link to Article
de Vos  R, Koster  RW, De Haan  RJ, Oosting  H, van der Wouw  PA, Lampe-Schoenmaeckers  AJ.  In-hospital cardiopulmonary resuscitation: prearrest morbidity and outcome. Arch Intern Med. 1999;159(8):845-850.
PubMed   |  Link to Article
Danciu  SC, Klein  L, Hosseini  MM, Ibrahim  L, Coyle  BW, Kehoe  RF.  A predictive model for survival after in-hospital cardiopulmonary arrest. Resuscitation. 2004;62(1):35-42.
PubMed   |  Link to Article
Cummins  RO, Chamberlain  D, Hazinski  MF,  et al; American Heart Association.  Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: the in-hospital “Utstein style.” Circulation. 1997;95(8):2213-2239.
PubMed   |  Link to Article
Stiell  IG, Nesbitt  LP, Nichol  G,  et al; OPALS Study Group.  Comparison of the Cerebral Performance Category Score and the Health Utilities Index for survivors of cardiac arrest. Ann Emerg Med. 2009;53(2):241-248.
PubMed   |  Link to Article
Chan  PS, Nichol  G, Krumholz  HM,  et al; American Heart Association National Registry of Cardiopulmonary Resuscitation (NRCPR) Investigators.  Racial differences in survival after in-hospital cardiac arrest. JAMA. 2009;302(11):1195-1201.
PubMed   |  Link to Article
Akaike  H.  A new look at the statistical model identification. IEEE Trans Automat Contr. 1974;19(6):716-723. doi:10.1109/TAC.1974.1100705.
Link to Article
Umbach  DM, Wilcox  AJ.  A technique for measuring epidemiologically useful features of birthweight distributions. Stat Med. 1996;15(13):1333-1348.
PubMed   |  Link to Article
Tibshirani  R.  Regression shrinkage and selection via the lasso. J R Stat Soc. 1996;58(1):267-288.
Schneiderman  LJ, Jecker  NS, Jonsen  AR.  Medical futility: its meaning and ethical implications. Ann Intern Med. 1990;112(12):949-954.
PubMed   |  Link to Article
Ebell  MH.  When everything is too much: quantitative approaches to the issue of futility. Arch Fam Med. 1995;4(4):352-356.
PubMed   |  Link to Article
Marco  CA, Larkin  GL.  Public education regarding resuscitation: effects of a multimedia intervention. Ann Emerg Med. 2003;42(2):256-260.
PubMed   |  Link to Article
Marco  CA, Larkin  GL.  Cardiopulmonary resuscitation: knowledge and opinions among the U.S. general public: state of the science-fiction. Resuscitation. 2008;79(3):490-498.
PubMed   |  Link to Article
Ebell  MH, Bergus  GR, Warbasse  L, Bloomer  R.  The inability of physicians to predict the outcome of in-hospital resuscitation. J Gen Intern Med. 1996;11(1):16-22.
PubMed   |  Link to Article
Patient Self-Determination Act of 1990. 101st Congress (1989-1990). http://thomas.loc.gov/cgi-bin/query/z?c101:H.R.4449.IH. Accessed August 1, 2013.

Figures

Place holder to copy figure label and caption
Figure.
Hosmer-Lemeshow Graph for the Full Model

The Hosmer-Lemeshow graph is shown for the full logistic regression model, which uses variable selection based on the Bayesian information criterion.

Graphic Jump Location

Tables

Table Graphic Jump LocationTable 1.  Definitions of Predictor Variablesa
Table Graphic Jump LocationTable 2.  Predictor Variables Initially Included in Multivariate Model Based on Bivariate Analysis
Table Graphic Jump LocationTable 3.  Results of Multivariate Analysis for Final Modela
Table Graphic Jump LocationTable 4.  Performance of the Final GO-FAR Score in the Training, Test, and Validation Sets

References

Saklayen  M, Liss  H, Markert  R.  In-hospital cardiopulmonary resuscitation: survival in 1 hospital and literature review. Medicine (Baltimore). 1995;74(4):163-175.
PubMed   |  Link to Article
Merchant  RM, Yang  L, Becker  LB,  et al; American Heart Association Get With the Guidelines-Resuscitation Investigators.  Incidence of treated cardiac arrest in hospitalized patients in the United States. Crit Care Med. 2011;39(11):2401-2406.
PubMed   |  Link to Article
Peberdy  MA, Kaye  W, Ornato  JP,  et al.  Cardiopulmonary resuscitation of adults in the hospital: a report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation. 2003;58(3):297-308.
PubMed   |  Link to Article
Ebell  MH, Afonso  AM.  Pre-arrest predictors of failure to survive after in-hospital cardiopulmonary resuscitation: a meta-analysis. Fam Pract. 2011;28(5):505-515. doi:10.1093/fampra/cmr023.
PubMed   |  Link to Article
Ebell  MH, Doukas  DJ, Smith  MA.  The do-not-resuscitate order: a comparison of physician and patient preferences and decision-making. Am J Med. 1991;91(3):255-260.
PubMed   |  Link to Article
Ebell  MH, Preston  PS.  The effect of the APACHE II score and selected clinical variables on survival following cardiopulmonary resuscitation. Fam Med. 1993;25(3):191-196.
PubMed
O’Keeffe  S, Ebell  MH.  Prediction of failure to survive following in-hospital cardiopulmonary resuscitation: comparison of two predictive instruments. Resuscitation. 1994;28(1):21-25.
PubMed   |  Link to Article
Ebell  MH, Kruse  JA, Smith  M, Novak  J, Drader-Wilcox  J.  Failure of three decision rules to predict the outcome of in-hospital cardiopulmonary resuscitation. Med Decis Making. 1997;17(2):171-177.
PubMed   |  Link to Article
Ebell  MH, Smith  M, Kruse  JA, Drader-Wilcox  J, Novak  J.  Effect of race on survival following in-hospital cardiopulmonary resuscitation. J Fam Pract. 1995;40(6):571-577.
PubMed
George  AL  Jr, Folk  BP  III, Crecelius  PL, Campbell  WB.  Pre-arrest morbidity and other correlates of survival after in-hospital cardiopulmonary arrest. Am J Med. 1989;87(1):28-34.
PubMed   |  Link to Article
Bowker  L, Stewart  K.  Predicting unsuccessful cardiopulmonary resuscitation (CPR): a comparison of three morbidity scores. Resuscitation. 1999;40(2):89-95.
PubMed   |  Link to Article
de Vos  R, Koster  RW, De Haan  RJ, Oosting  H, van der Wouw  PA, Lampe-Schoenmaeckers  AJ.  In-hospital cardiopulmonary resuscitation: prearrest morbidity and outcome. Arch Intern Med. 1999;159(8):845-850.
PubMed   |  Link to Article
Danciu  SC, Klein  L, Hosseini  MM, Ibrahim  L, Coyle  BW, Kehoe  RF.  A predictive model for survival after in-hospital cardiopulmonary arrest. Resuscitation. 2004;62(1):35-42.
PubMed   |  Link to Article
Cummins  RO, Chamberlain  D, Hazinski  MF,  et al; American Heart Association.  Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: the in-hospital “Utstein style.” Circulation. 1997;95(8):2213-2239.
PubMed   |  Link to Article
Stiell  IG, Nesbitt  LP, Nichol  G,  et al; OPALS Study Group.  Comparison of the Cerebral Performance Category Score and the Health Utilities Index for survivors of cardiac arrest. Ann Emerg Med. 2009;53(2):241-248.
PubMed   |  Link to Article
Chan  PS, Nichol  G, Krumholz  HM,  et al; American Heart Association National Registry of Cardiopulmonary Resuscitation (NRCPR) Investigators.  Racial differences in survival after in-hospital cardiac arrest. JAMA. 2009;302(11):1195-1201.
PubMed   |  Link to Article
Akaike  H.  A new look at the statistical model identification. IEEE Trans Automat Contr. 1974;19(6):716-723. doi:10.1109/TAC.1974.1100705.
Link to Article
Umbach  DM, Wilcox  AJ.  A technique for measuring epidemiologically useful features of birthweight distributions. Stat Med. 1996;15(13):1333-1348.
PubMed   |  Link to Article
Tibshirani  R.  Regression shrinkage and selection via the lasso. J R Stat Soc. 1996;58(1):267-288.
Schneiderman  LJ, Jecker  NS, Jonsen  AR.  Medical futility: its meaning and ethical implications. Ann Intern Med. 1990;112(12):949-954.
PubMed   |  Link to Article
Ebell  MH.  When everything is too much: quantitative approaches to the issue of futility. Arch Fam Med. 1995;4(4):352-356.
PubMed   |  Link to Article
Marco  CA, Larkin  GL.  Public education regarding resuscitation: effects of a multimedia intervention. Ann Emerg Med. 2003;42(2):256-260.
PubMed   |  Link to Article
Marco  CA, Larkin  GL.  Cardiopulmonary resuscitation: knowledge and opinions among the U.S. general public: state of the science-fiction. Resuscitation. 2008;79(3):490-498.
PubMed   |  Link to Article
Ebell  MH, Bergus  GR, Warbasse  L, Bloomer  R.  The inability of physicians to predict the outcome of in-hospital resuscitation. J Gen Intern Med. 1996;11(1):16-22.
PubMed   |  Link to Article
Patient Self-Determination Act of 1990. 101st Congress (1989-1990). http://thomas.loc.gov/cgi-bin/query/z?c101:H.R.4449.IH. Accessed August 1, 2013.

Correspondence

CME


You need to register in order to view this quiz.
Submit a Comment

Multimedia

Supplement.

eFigure. Receiver Operating Characteristic Curve for the Good Outcome Following Attempted Resuscitation (GO-FAR) Score Using Data From the Validation Set

eTable. Performance of the 4 Candidate Models in the Training and Test Sets

Supplemental Content

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

See Also...
Articles Related By Topic
Related Collections
PubMed Articles