American Journal of Respiratory and Critical Care Medicine

Rationale: Studies examining survival outcomes after in-hospital cardiopulmonary arrest (CPA) among intensive care unit (ICU) patients requiring medications for hemodynamic support are limited.

Objectives: To examine outcomes of ICU patients who received cardiopulmonary resusitation.

Methods: We identified 49,656 adult patients with a first CPA occurring in an ICU between January 1, 2000 and August 26, 2008 within the National Registry of Cardiopulmonary Resuscitation. Survival outcomes of patients requiring hemodynamic support immediately before CPA were compared with those of patients who did not receive hemodynamic support (pressors), using multivariable logistic regression analyses to adjust for differences in demographics and clinical characteristics. Pressor medications included epinephrine, norepinephrine, phenylephrine, dopamine, dobutamine, and vasopressin.

Measurements and Main Results: The overall rate of survival to hospital discharge was 15.9%. Patients taking pressors before CPA were less likely to survive to discharge (9.3 vs. 21.2%; P < 0.0001). After multivariable adjustment, patients taking pressors before pulseless CPA were 55% less likely to survive to discharge (adjusted odds ratio [OR], 0.45; 95% confidence interval [CI], 0.42–0.48). Age equal to or greater than 65 years (adjusted OR, 0.77; 95% CI, 0.73–0.82), nonwhite race (adjusted OR, 0.58; 95% CI, 0.54–0.62), and mechanical ventilation (adjusted OR, 0.60; 95% CI, 0.56–0.63) were also variables that could be identified before CPA that were independently associated with lower survival. More than half of survivors were discharged to rehabilitation or extended care facilities. Only 3.9% of patients who had CPA despite pressors were discharged home from the hospital, as compared with 8.5% of patients with a CPA and not taking pressors (adjusted OR, 0.53; 95% CI, 0.49–0.59).

Conclusions: Although overall survival of ICU patients was 15.9%, patients requiring pressors and who experienced a CPA in an ICU were half as likely to survive to discharge and to be discharged home than patients not taking pressors. This study provides robust estimates of CPR outcomes of critically ill patients, and may assist clinicians to inform consent for this procedure.

Scientific Knowledge on the Subject

There are no large studies describing outcomes of critically ill patients who receive cardiopulmonary resuscitation (CPR).

What This Study Adds to the Field

This study analyzes the largest collected data source of CPR in the world, thereby providing robust estimates of outcomes after CPR in critically ill patients.

Despite its initial application to selected patient populations (1, 2), cardiopulmonary resuscitation (CPR) is now often administered to critically ill patients (3), even those who experience cardiopulmonary arrest (CPA) despite medications administered to reverse hypotension. Small single-center studies of CPR in the critically ill have demonstrated survival rates ranging from 0 to 47% (410). We hypothesized that patients who experience CPA despite administration of medications used to treat hypotension (pressors) are highly unlikely to respond to CPR. In a prior study from our institution, whereas 17% (of 83) critically ill patients who received CPR survived to hospital discharge, only 3.6% of patients taking pressors survived. The odds of death for patients taking pressors were 35 times those of patients not taking pressors (9). Our study was limited by its small sample size. Therefore, in the current study, we used data from the National Registry of Cardiopulmonary Resuscitation (NRCPR), the largest prospective registry of in-hospital cardiac arrests, to determine more definitively clinical factors associated with outcomes of critically ill patients who received CPR. Specifically, we hypothesize that critically ill patients who develop CPA and require CPR while taking pressors are much less likely to survive than those not taking pressors.

This study was approved by both the Institutional Review Committee of Bridgeport Hospital (Bridgeport, CT) and the research review committee of the National Registry of Cardiopulmonary Resuscitation. NRCPR is an American Heart Association–sponsored prospective, multisite, observational registry of hospital resuscitation. It includes events from more than 400 participating facilities that voluntarily submit performance data on patients receiving CPR in their institutions. The standardized Utstein-based data collection methods of the NRCPR have been described previously (11, 12). For the purposes of the NRCPR, CPA is defined as a resuscitation event that requires chest compressions or defibrillation and elicits an emergency resuscitation response by facility personnel (13).

We included for this analysis all CPAs of adults 18 years or older that occurred in intensive care units (ICUs) between January 1, 2000 and August 26, 2008. For patients with multiple CPAs, only the first event was included. We excluded patients who had a CPA outside of critical care units, and those with missing mortality data (Figure 1).

The primary independent variable was continuous intravenous administration of medications to support blood pressure (pressors), including phenylephrine, vasopressin, dobutamine, epinephrine, norepinephrine, and dopamine (>3 μg/kg/min) immediately before CPA. The primary outcome of interest was survival to hospital discharge. A secondary outcome of interest was discharge disposition (to home or an extended care or rehabilitation facility). Limited data were also available regarding neurological outcomes at hospital discharge in the form of “cerebral performance categories”: 1, normal or good status; 2, capable of activities of daily living; 3, severe disability and dependent for daily support; 4, coma or vegetative state; 5, brain death.

We evaluated potential confounding from a variety of patient variables, including age, sex, residence before hospitalization, comorbidities before CPA, cause of CPA, initial CPA rhythm, use of invasive mechanical ventilation before CPA, and time of CPA.

Statistical Analysis

Demographic and clinical characteristics of patients who took pressors were compared with those of patients who were not taking pressors at the time of CPA. Categorical variables are reported as percentages, and between-group comparisons were made by chi-square test. Continuous variables are reported as means ± standard deviation when variables were normally distributed, or as medians with the interquartile range when variables were not normally distributed. Differences between group means were compared by Student t test for normally distributed variables. Chi-square tests were used to compare group medians for nonnormally distributed variables.

To adjust for confounding, multivariate logistic regression analyses were performed to assess the relationship of use of pressor medications and survival in ICU patients with a pulseless CPA. Those with weak pulse who received chest compressions (“Other” in Figure 1) were excluded in multiple logistic regression analyses. Variables were included in the regression analyses if they demonstrated significant association with survival in univariate analyses. The Hosmer and Lemeshow goodness-of-fit test was performed to test the model's fitness. Because dobutamine can be administered for hypotension (i.e., as a pressor) or congestive heart failure without hypotension, and the NRCPR does not distinguish indications for its administration, two complete analyses were performed. The primary analysis, presented herein, includes patients who took dobutamine. Results of a second analysis excluding patients who took dobutamine are provided in the online supplement. Analyses were performed with SAS, version 8.2 (SAS Institute, Cary, NC).

Characteristics and Outcomes of the Overall Cohort

Of 49,656 ICU patients with a CPA, 67.7% were white, 58.4% were men, and the median age was 69 years (Table 1). Patients taking pressors constituted 44.1% of the study cohort. In addition, nearly half of the patients were mechanically ventilated (MV) and 50.7% of the CPAs occurred at night or during weekends. Sixty-six percent had pulseless electrical activity (PEA) or asystole and 18.9% had ventricular tachycardia or ventricular fibrillation (VT/VF) as the initial cardiac rhythm at the time of CPA.


Study Variable

Total (n = 49,656) [No. (%)]

Pressors (n = 21,894) [No. (%)]

No Pressors (n = 27,762) [No. (%)]

P Value
Age, yr
 18–404,269 (8.6)1,999 (9.1)2,270 (8.2)<0.0001
 41–505,247 (10.6)2,381 (10.9)2,866 (10.3)
 51–608,652 (17.4)3,914 (17.9)4,738 (17.1)
 61–7010,809 (21.8)4,860 (22.2)5,949 (21.4)
 71–8012,914 (26.0)5,588 (25.5)7,326 (26.4)
 81+7,762 (15.6)3,150 (14.4)4,612 (16.6)
 Unknown3 (0.01)2 (0.01)1 (0.00)
 Female20,666 (41.6)9,307 (42.5)11,359 (40.9)0.0003
 Male28,987 (58.4)12,585 (57.5)16,402 (59.1)
 White33,631 (67.7)14,472 (66.1)19,159 (69.0)<0.0001
 Nonwhite12,412 (25.0)5,818 (26.6)6,594 (23.8)
 Unknown3,613 (7.3)1,604 (7.3)2,009 (7.2)
Comorbidities before CPA
 Hepatic insufficiency4,693 (9.5)2,696 (12.3)1,947 (7.0)<0.0001
 Renal insufficiency17,548 (35.3)9,231 (42.2)8,317 (30.0)<0.0001
 Respiratory insufficiency24,030 (48.4)11,992 (54.8)12,038 (43.4)<0.0001
 Metastatic or hematological malignancy5,606 (11.3)2,586 (11.8)3,020 (10.9)0.002
 ≥Three comorbidities4,472 (9.0)2,501 (11.4)1,971 (7.1)<0.0001
 ≤Two comorbidities45,184 (91.0)19,393 (88.6)25,791 (92.9)
Immediate causes of CPA
 Hypothermia340 (0.7)230 (1.1)110 (0.4)<0.0001
 Active myocardial infarction4,328 (8.7)2,153 (9.8)2,175 (7.8)<0.0001
 Pulmonary embolism869 (1.8)309 (1.4)560 (2.0)<0.0001
 Septicemia9,717 (19.6)5,808 (26.5)3,909 (14.1)<0.0001
 ≥Three conditions8,171 (16.5)4,550 (20.8)3,621 (13.0)<0.0001
 ≤Two conditions41,485 (83.5)17,344 (79.2)24,141 (87.0)
Initial CPA rhythm
 PEA/asystole32,660 (65.8)14,909 (68.1)17,751 (63.9)<0.0001
 VT/VF9,387 (18.9)3,861 (17.6)5,226 (19.9)
 Other rhythms7,609 (15.3)3,124 (14.3)4,485 (16.2)
Mechanical ventilation
 Yes24,522 (49.4)14,196 (64.8)10,326 (37.2)<0.0001
 No25,133 (50.6)7,698 (35.2)17,435 (62.8)
Residency before admission
 Extended care facility9,398 (18.9)4,361 (19.9)5,037 (18.1)<0.0001
 Home36,092 (72.7)15,810 (72.2)20,282 (73.1)
Timing of CPA
 Night or weekend*25,176 (50.7)11,198 (51.1)13,978 (50.3)0.008
24,480 (49.3)
10,696 (48.9)
13,784 (49.7)

Definition of abbreviations: CPA = cardiopulmonary arrest; PEA = pulseless electrical activity; VT/VF = ventricular tachycardia/ventricular fibrillation.

*Night was defined as 11:00 p.m. to 7:00 a.m.; weekend was defined as 11:00 p.m. Friday to 7:00 a.m. Monday.

Overall, 15.9% of patients who had their first CPA in the ICU survived until hospital discharge. Survival rates after PEA (10.9%) were similar to those of patients with asystole (11.1%) but substantially lower than those for patients with VT (30.7%) or VF (34.2%). Patients with other primary rhythms, including supraventricular tachycardias, VT with a weak pulse, and bradycardias comprised 15.3% of the sample and 16.0% survived (Figure 1, and Appendix E1 in the online supplement).

Patients taking pressors were younger (63.9 vs. 65.0 yr; P < 0.0001); had more comorbidities, including hepatic (12.3 vs. 7.0%), renal (42.2 vs. 30.0%), or respiratory (54.8 vs. 43.4%) insufficiency; and were more likely to be mechanically ventilated (64.8 vs. 37.2%) compared with those not taking pressors (all P < 0.0001). Similarly, patients taking pressors were more likely to have septicemia as the specific immediate cause of CPA (26.5 vs. 14.1%; P < 0.001), to have three or more immediate contributory causes of CPA (20.8 vs. 13.0%; P < 0.0001), and were more likely to develop PEA or asystole (68.1 vs. 63.9%; P < 0.0001) compared with those not taking pressors (Table 1).

Comparing Outcomes: Pressors versus No Pressors

Patients taking pressors before CPA were less likely to survive to discharge than patients who did not require pressors (9.3 vs. 21.2%; P < 0.0001) (Figures 1 and 2, and Table 2). This difference in survival persisted when patients were stratified by the primary CPA rhythm (22.6 vs. 40.7% for VT/VF, P < 0.0001 and 5.9 vs. 15.2% for PEA/asystole, P < 0.0001) (Figure 1), and by MV status (7.6 vs. 16.1% for MV patients, P < 0.0001; 12.4 vs. 24.2% for non-MV patients, P < 0.0001) (Figure 2). These differences persisted when only patients with pulseless CPA were analyzed (see Appendix 2 in the online supplement). In addition, patients taking two or more pressors were less likely to survive than those taking a single pressor (6.4 vs. 11.5%, P < 0.001) (Figure 3). Survival ranged from 8% (for patients taking norepinephrine or vasopressin alone) to 23% (for those taking dobutamine alone).


Survived [n (%)]

Died [n (%)]

P Value
Age, yr
 18–40606 (14.2)3,663 (85.8)<0.0001
 41–50832 (15.9)4,415 (84.1)
 51–601,511 (17.5)7,141 (82.5)
 61–701,942 (18.0)8,867 (82.0)
 71–802,033 (15.7)10,881 (84.3)
 81+992 (12.8)6,770 (87.2)
 Female3,153 (15.3)17,513 (84.7)0.0004
 Male4,763 (16.4)24,224 (83.6)Control
 White5,934 (17.6)27,697 (82.4)<0.0001
 Nonwhite1,344 (10.8)11,068 (89.2)Control
Comorbidities before CPA
 Hepatic insufficiency361 (7.8)4,282 (92.2)<0.0001
 Renal insufficiency2,018 (11.5)15,530 (88.5)<0.0001
 Respiratory insufficiency3,110 (12.9)20,920 (87.1)<0.0001
 Metastatic or hematological malignancy461 (1.1)41,740 (98.9)<0.0001
 ≥Three comorbidities568 (12.7)3,904 (87.3)<0.0001
 ≤Two comorbidities7,348 (16.3)37,836 (83.7)Control
Immediate causes of CPA
 Hypothermia21 (6.2)319 (93.8)<0.0001
 Active myocardial infarction615 (14.2)3,713 (85.8)0.001
 Pulmonary embolism64 (7.4)805 (92.6)<0.0001
 Septicemia820 (8.4)8,897 (91.6)<0.0001
 ≥Three conditions847 (10.4)7,324 (89.6)<0.0001
 ≤Two conditions7,069 (17.0)34,416 (83.0)Control
Use of vasoactive medications
 Any medication2,033 (9.3)19,861 (90.7)<0.0001
 ≥Two medications614 (6.4)8,927 (93.6)<0.0001
 One medication1,419 (11.5)10,934 (88.5)<0.0001
Specific vasoactive medications
 Phenylephrine, norepinephrine, or vasopressin809 (6.4)11,828 (93.6)<0.0001
 Dopamine or dobutamine1,125 (12.8)7,631 (87.2)Control
Initial CPA rhythm
 PEA/asystole3,580 (11.0)29,080 (89.0)<0.0001
 VT/VF3,120 (33.2)6,267 (66.8)Control
 Other rhythms1,216 (16.0)6,393 (84.0)0.92
Mechanical ventilation
 Yes2,738 (11.2)21,784 (88.8)<0.0001
 No5,178 (20.6)19,955 (79.4)Control
Residence before admission
 Extended care facility1,462 (15.6)7,936 (84.4)0.26
 Home5,812 (16.1)30,280 (83.9)Control
Time of CPA
 Night or weekend*3,577 (14.2)21,599 (85.8)<0.0001
 Other time
4,339 (17.7)
20,141 (82.3)

Definition of abbreviations: CPA = cardiopulmonary arrest; PEA = pulseless electrical activity; VT/VF = ventricular tachycardia/ventricular fibrillation.

*Night was defined as 11:00 p.m. to 7:00 a.m.; weekend was defined as 11:00 p.m. Friday to 7:00 a.m. Monday.

After multivariable adjustment, the odds of survival for patients with pulseless CPA and taking pressors were less than half compared with patients not taking pressors (adjusted odds ratio [OR], 0.45; 95% confidence interval [95% CI], 0.42 to 0.48). Other independent predictors of survival to hospital discharge included race, mechanical ventilation, three or more immediate causes of CPA, age 65 years or more, and CPA on nights or weekends (Table 3). The Hosmer and Lemeshow goodness-of-fit test suggested a poor model fit (P = 0.003). When primary CPA rhythm, which may be considered a confounder or outcome, was included, model fit improved (P = 0.8) and survival estimates were not impacted substantially (see Appendix 3 in the online supplement). Similarly, when patients taking dobutamine alone—a group that had a substantially higher survival rate—were excluded, survival estimates were affected minimally (see Appendix 4 in online supplement).


Odds Ratio

95% Confidence Interval
Taking pressors0.450.42–0.48
Receiving mechanical ventilation0.600.56–0.63
≥Three immediate causes of CPA0.650.60–0.71
Age ≥ 65 yr0.770.73–0.82
CPA occurred at night or on weekend*

Definition of abbreviation: CPA = cardiopulmonary arrest.

Notes: In the initial models pressors, race, mechanical ventilation, number of immediate causes of CPA, age, time of CPA, sex, and number of comorbidities were included. In stepwise elimination, sex and number of comorbidities were not independently associated with survival. P = 0.003 for Hosmer and Lemeshow goodness-of-fit test.

*Night was defined as 11:00 p.m. to 7:00 a.m.; weekend was defined as 11:00 p.m. Friday to 7:00 a.m. Monday.

Discharge Destinations and Neurological Outcomes

Only 3.9% of CPA patients taking pressors were discharged home compared with 8.5% among those not taking pressors (adjusted OR, 0.53; 95% CI, 0.49–0.59; P < 0.0001). Only 3.3% of patients who had a CPA despite pressors were discharged home with favorable neurological outcome (normal and/or able to attend to their activities of daily living), as compared with 7.0% of patients with a CPA and not taking pressors (adjusted OR, 0.56; 95% CI, 0.51–0.62; P < 0.0001).

This study found a hospital survival rate of 15.9% in critically ill patients who suffered cardiopulmonary arrest. The odds of survival were 55% lower among patients taking pressor medications for hemodynamic support. Only 3.3% of patients with CPA despite pressors were discharged home with a favorable neurological outcome. Because NRCPR data are not sufficient to adjust for severity of illness, we could not determine whether pressors harm patients or simply mark disease severity. Nonetheless, given the large size of the NRCPR, this study provides the most contemporary and precise estimates of CPR survival for critically ill patients, including those requiring pressors, in the United States.

Treatment with pressors immediately before CPA was most strongly associated with outcomes, but race, time of CPR, initial CPA rhythm, and advanced age were also associated with poorer outcomes. These findings, in the critically ill subcohort of the NRCPR database, mirror those already reported for the overall cohort (11, 14, 15). It is important to note that in the previous report of the NRCPR investigators, excess mortality of nonwhite patients was attributable to “black patients receiving care at hospitals with lower rates of survival,” possibly related to “higher rates of delays in defibrillation time” (14). Although the absolute magnitude of the difference (14.2 vs. 17.7%) was less than that observed in the total NRCPR cohort that includes ward patients (14.7 vs. 19.8%) (15), critically ill patient survival was poorer on nights and weekends. ICU nursing staff ratios generally do not change at night or during weekends, whereas most hospitals do not have in-house intensivists at those times. Thus our study raises the possibility that less physician presence may contribute to lower survival from CPA at night or on weekends.

Our study also demonstrates that mechanical ventilation was associated with 40% lower odds of hospital survival after CPR. Because a substantial proportion of critically ill patients take pressors and/or mechanical ventilation at some point, these findings have significant clinical implications. A common model of critical care is to provide a “trial of life-sustaining therapies” until a patient declares himself as stabilizing, improving, or failing. CPA despite pressors or mechanical ventilation suggests that the trial is failing. It is not surprising that patients who have CPA despite pressors—most commonly used to treat shock—do not survive the higher dose pressors, ventilation, defibrillation, and compressions that comprise CPR. The illness that caused shock and CPA is not impacted by CPR; thus even if CPR is initially successful, such patients return to severe illness that has already declared itself refractory to aggressive support. Although, the aim of our study was to examine the outcomes of patients who suffered CPA despite pressors, we discovered other factors, including advanced age and CPA despite mechanical ventilation, that portend poor outcomes for critically ill patients.

Prognostic estimates, even when based on robust data, should be offered carefully, because population data of this nature are difficult to apply to specific cases. Although quantitative estimates of survival shape patients' decisions regarding CPR (16), they also consider quality of life after successful CPR. For example, 80% of patients opted out of CPR when survival was estimated as no more than 10% (16), and quality of life after survival was considered “extremely important” by more than one-third of patients (17). Our study is unique in that it examines outcomes beyond hospital survival, namely discharge destinations and gross neurological outcomes. Patients who had CPA and received CPR despite pressors were half as likely to survive hospitalization and less likely to have favorable neurological outcomes (3.3 vs. 7.0%). In many hospitals, CPR is provided to critically ill patients, including those in shock, often without patients' consent and based on small case series (310). Authors have suggested that intensivists should unilaterally withhold (3) or seek only assent (18) for futile life-sustaining therapies (3). Our data demonstrate that CPR is not uniformly futile in critically ill patients; overall, one in six survived to hospital discharge. Even in patients requiring pressors, 9% survived hospitalization and 4% were discharged home. Unfortunately, our study cannot aid in identifying this small minority. Because most patients wish to make their own informed medical decisions, even regarding CPR (19), our study provides critically ill patients and their families with the most robust outcomes data available, as they consider the risks, benefits, and alternatives of CPR, in the context of the patient's values (20, 21).

Our study should be interpreted in the context of the following limitations. The strengths of this study include the large hospital and patient sample size, and the use of standardized, rigorously collected data. Despite this, results from the NRCPR may not be generalizable to all hospitals or ICUs. For example, severity of illness and case mixes may vary remarkably among ICUs. Although it remains the most detailed database of CPR, the NRCPR depends on accurate data acquisition and transfer, and does not collect all data germane to this study. For example, information regarding doses of and duration of pressor infusions is not collected but could impact outcomes. Vital signs and detailed information about the severity of acute and chronic illnesses for computation of validated prognostic scores, such as the Acute Physiology and Chronic Health Evaluation (APACHE), were not available. Survival and quality of life 6–12 months after discharge are important considerations for patients and families, but are not available in the NRCPR. Finally, some medications (e.g., dobutamine) were commonly used during the study epoch in patients without shock (i.e., severe congestive heart failure). It is notable that patients with CPA and taking dobutamine alone were roughly twice as likely to survive compared with patients taking other pressors (see Figure 3). The NRCPR does not provide indications for administered pressors, but the data suggest that groups taking dobutamine alone have better underlying prognoses. Although survival estimates of patients taking pressors (as a group) decreased, excluding patients who took dobutamine did not substantially impact the associations noted in this study (see the online supplement). Notwithstanding these limitations, the NRCPR provides the most comprehensive and rigorously acquired/maintained database in the world, and thus provides the most robust estimates to date of outcomes of critically ill patients receiving CPR. Last, we used the established NRCPR definition for CPA, which includes some patients with “other” nonpulseless rhythms. Although we were surprised at the high mortality of this group, excluding them from analysis did not affect our findings (see the online supplement).

In conclusion, this study demonstrates that one of six critically ill patients who develop CPA in an ICU and receive CPR survives to hospital discharge. Mechanical ventilation, nonwhite race, age 65 years or more, and CPA on nights or weekends were among the variables independently associated with poorer prognosis. This study supports our initial hypothesis in that patients taking pressors were half as likely to survive as patients not requiring these medications for hemodynamic support.

The authors are grateful to the NRCPR research review board for permitting access to the database, and providing two rounds of peer review before final submission of this manuscript.

1. Kouwenhoven WB, Jude JR, Knickerbocker GG. Closed-chest cardiac massage. JAMA 1960;173:1064–1067.
2. Weisfeldt ML, Ornato JP. Closed-chest cardiac massage: progress measured by the exceptions. JAMA 2008;300:1582–1584.
3. Nader A, Seneff MG. In-hospital CPR: performing it better but less often. J Intensive Care Med 2009;24:208–209.
4. Myrianthefs P, Kalafati M, Lemonidou C, Minasidou E, Evagelopoulou P, Karatzas S, Baltopoulos G. Efficacy of CPR in a general, adult ICU. Resuscitation 2003;57:43–48.
5. Karetzky M, Zubair M, Parikh J. Cardiopulmonary resuscitation in intensive care unit and non-intensive care unit patients. Arch Intern Med 1995;155:1277–1280.
6. Smith DL, Kim K, Cairns BA, Fakhry SM, Meyer AA. Prospective analysis of outcome after cardiopulmonary resuscitation in critically ill surgical patients. J Am Coll Surg 1995;180:402–409.
7. Peterson MW, Geist LJ, Schwartz DA, Konicek S, Moseley PL. Outcome after cardiopulmonary resuscitation in a medical intensive care unit. Chest 1991;100:168–174.
8. Enohumah KO, Moerer O, Kirmse C, Bahr J, Neumann P, Qintel M. Outcome of cardiopulmonary resuscitation in intensive care units in a university hospital. Resuscitation 2006;71:161–170.
9. Grigoriyan A, Vazquez R, Pavlinskaya T, Bindelglass G, Rishi A, Amoateng-Adjepong Y, Manthous CA. Outcomes of CPR for patients on pressors or inotropes: a pilot study. J Crit Care 2009;24:415–418.
10. Chang S, Huang C, Shih C, Lee C, Chang W, Chen Y, Lee CH, Lin ZY, Tsai MS, Hsu CY, et al. Who survives cardiac arrest in intensive care units. J Crit Care 2009;24:408–414.
11. Peberdy MA, Kaye W, Ornato JP, Larkin GL, Nadkarni V, Mancini ME, Berg RA, Nichol G, Lane-Trultt T. Cardiopulmonary resuscitation of adults in the hospital: a report of 14 720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation 2003;58:297–308.
12. National Registry of Cardiopulmonary Resuscitation (NRCPR), American Heart Association. CPA data entry form. [Accessed June 2010]. Available from:
13. National Registry of Cardiopulmonary Resuscitation (NRCPR), American Heart Association. Inclusion/exclusion criteria [Accessed June 2010]. Available from:
14. Chan PS, Nichol G, Krumholz HM, Spertus JA, Jones PG, Peterson ED, Rathore SS, Nallamothu BK. Racial differences in survival after in-hospital cardiac arrest. JAMA 2009;302:1195–1201.
15. Perberdy MA, Ornato JP, Larkin GL, Braithwaite RS, Kashner TM, Carey SM, Meaney PA, Cen L, Nadkarni VM, Praestgaard AH, et al. Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008;299:785–792.
16. Murphy DJ, Burrows D, Santilli S, Kemp AW, Tenner S, Kreling B, Teno J. The influence of the probability of survival on patients' preferences regarding cardiopulmonary resuscitation. N Engl J Med 1994;330:545–549.
17. Heyland DK, Frank C, Groll D, Pichora D, Dodek D, Rocker G, Gafni A. Understanding cardiopulmonary resuscitation decision making: perspectives of seriously ill hospitalized patients and family members. Chest 2006;130:419–428.
18. Curtis JR, Burt RA. Point: the ethics of unilateral “do not resuscitate” orders: the role of “informed assent.” Chest 2007;132:748–751.
19. Nicolasora N, Pannala R, Mountantanakis S, Shanmugan B, Amoateng-Adjepong Y, Manthous CA. Hospitalized patients want to choose whether to receive life-sustaining therapies. J Hosp Med 2006;1:161–167.
20. Waisel DB, Truog RD. The cardiopulmonary resuscitation-not-indicated order: futility revisited. Ann Intern Med 1995;122:304–308.
21. Manthous CA. Counterpoint: is it ethical to order “do not resuscitate” without patient consent? Chest 2007;132:751–754.
Correspondence and requests for reprints should be addressed to Constantine A. Manthous, M.D., Bridgeport Hospital and Yale University School of Medicine, 267 Grant Street, Bridgeport, CT 06610. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record