This Article Abstract Résumé de cet Article Full Text (PDF) Additional Material Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sinuff, T. Search for Related Content PubMed PubMed Citation Articles by Sinuff, T.

DNR directives are established early in mechanically ventilated intensive care unit patients

[Les directives PDR sont établies tôt chez les patients sous ventilation mécanique à l’unité des soins intensifs]

Tasnim Sinuff, MD^*, Deborah J. Cook, MD^*,, Graeme M. Rocker, DM, Lauren E. Griffith, MSc, Stephen D. Walter, PhD, Malcolm M. Fisher, MD, Peter M. Dodek, MD^¶, Peter Sjokvist, MD, Ellen McDonald, RN, John C. Marshall, MD^**, Peter A. Kraus, MD^*, Mitchell M. Levy, MD, Neil M. Lazar, MD and Gordon H. Guyatt, MD^*, for The Level Of Care Study Investigators & The Canadian Critical Care Trials Group

^* From the Departments of Medicine and
Clinical Epidemiology & Biostatistics, McMaster University, Hamilton, Ontario, Canada;
the Department of Medicine, Dalhousie University, Halifax, Nova Scotia, Canada;
the Intensive Therapy Unit, Royal North Shore Hospital, University of Sydney, Sydney, Australia;
^¶ the Program of Critical Care Medicine and Center for Health Evaluation and Outcome Sciences, University of British Columbia, Vancouver, British Columbia, Canada; the Department of Anesthesia & Intensive Care, Huddinge University, Stockholm, Sweden;
^** the Department of Surgery, University of Toronto, Toronto, Ontario, Canada;
the Department of Medicine, Brown University, Providence, Rhode Island, USA; and
the Department of Medicine, University of Toronto and University Health Network, Toronto, Ontario, Canada.

Address correspondence to: Dr. D. J. Cook, Department of Medicine & Epidemiology and Biostatistics, McMaster University Health Sciences Center, Room 2C11, 1200 Main Street West, Hamilton, Ontario L8N 3Z5, Canada. Phone: 905-525-9140, ext. 22900; Fax: 905-524-3841; E-mail: debcook{at}mcmaster.ca

	Abstract

TOP Abstract Introduction Methods Results Discussion References

 
Purpose: Setting treatment goals in the intensive care unit (ICU) often involves resuscitation decisions. Our objective was to study the rate of establishing do-not-resuscitate (DNR) directives, determinants, and outcomes of those directives for mechanically ventilated patients.

Methods: In a multicentre observational study, we included consecutive adults with no DNR directives within 24 hr of ICU admission who were mechanically ventilated for at least 48 hr. We identified the rate with which DNR directives were established, and factors associated with these directives.

Results: Among 765 patients, DNR directives were established for 231 (30.2%) patients; 143 (62.1%) of these were established within the first week. Factors independently associated with a DNR directive were: patient age [ 75 yr (hazard ratio [HR] 2.3, 95% confidence interval 1.5–3.4], 65 to 74 yr (HR 1.8, 1.2–2.7), 50 to 64 yr (HR 1.4, 1.0–2.2) relative to < 50 yr); medical rather than surgical diagnosis (HR 1.8, 1.3–2.5); multiple organ dysfunction score (HR 1.7 for each five-point increment, 1.4–2.0); physician prediction of ICU survival [< 10% (HR 15.0, 6.7–33.6)], 10 to 40% [(HR 5.0, 2.3–11.2), 41 to 60% (HR 4.0, 1.8–9.0) relative to > 90%]; and physician perception of patient preference to limit life support (no advanced life support [(HR 5.8, 3.6–9.4) or partial advanced life support (HR 3.2, 2.2–4.6) compared to full measures].

Conclusion: One third of mechanically ventilated patients had DNR directives established early during their ICU stay after the first 24 hr of admission. The strongest predictors of DNR directives were physician prediction of low probability of survival, physician perception of patient preference to limit life support, organ dysfunction, medical diagnosis and age.

	Introduction

TOP Abstract Introduction Methods Results Discussion References

 
SETTING realistic treatment goals in the intensive care unit (ICU) often involves decisions about whether to attempt cardiopulmonary resuscitation (CPR) in the event of a cardiopulmonary arrest. Hospital survival rates of ICU patients who receive CPR range from 0 to 33%.^1–7 Decisions to provide or forgo CPR in the ICU should reflect patients’ values and preferences. The influence of physicians’ judgements about the utility of CPR on patients’ resuscitation preferences remains uncertain, as does the effect of physicians’ perceptions of patients’ survival and of patients’ preferences on decisions to offer CPR.

Before making decisions about CPR and other life-support measures, many patients want to discuss their prognosis with physicians.^8–12 Despite discussions, physicians and patient surrogates are still sometimes unaware of, or are unable to predict, patient preferences regarding CPR.^12–16 Moreover, the concurrence between physician or surrogate judgements and patients’ CPR preferences is low.^10,12,14,16 For critically ill patients already receiving life-support who cannot communicate their values and preferences, it is often unclear whether CPR would be congruent with their wishes. In the ICU, decisions about CPR are frequently delayed until late in the course of critical illness¹² and sometimes actually follow rather than precede cardiopulmonary arrest and successful resuscitation.¹⁷

When critically ill patients elect not to undergo CPR in the event of a cardiopulmonary arrest, their wishes are documented as do-not-resuscitate (DNR) directives. Prospective studies have shown that 9 to 13% of patients had DNR orders written at some point during their ICU stay.^18–20 Also, the frequency of new DNR orders in the ICU appears to have increased over the past decade.^21,22 We previously found that explicit CPR directives were established for only 11% of ICU patients within the first 24 hr of their ICU admission; half of these were directives not to resuscitate.²³ Increased age, illness severity, and inability to participate in decision-making were the strongest predictors of a DNR directive during the first 24 hr of ICU admission.²³ The clinical decisions that lead to establishing DNR directives may change with time during the ICU course. To further understand this process, the objectives of this study were to examine the rate, determinants and outcomes of DNR directives for mechanically ventilated patients after the first 24 hr of their ICU stay.

	Methods

TOP Abstract Introduction Methods Results Discussion References

 
Patients and procedures
Among the 3,099 critically ill patients admitted to the 15 ICUs (11 in Canada, two in the US, one in Sweden and one in Australia) between May 1995 and September 1998, of the originally published study,²³ we included consecutive patients in this study if they met the following criteria: age greater than 18 yr, predicted ICU length of stay greater than 72 hr, and mechanical ventilation for at least 48 hr. Patients were excluded if they had a DNR directive established before or within the first 24 hr of ICU admission. If patients had more than one ICU admission, we included only the last. The need for informed consent was waived due to the observational nature of the study. Ethics approval was obtained by all Institutional Ethics Review Boards.

We collected baseline data including patient sex, age, APACHE II score²⁴ during the first ICU day, and admission diagnosis. Daily, we calculated the multiple organ dysfunction score (MODS),^25,26 and recorded whether life support (mechanical ventilation, inotropes or vasopressors, dialysis) was administered, withheld or withdrawn, the patient’s ability to participate in decision-making, and whether a DNR directive was established. In addition, we documented daily the attending physicians’ clinical judgements, including their estimate of the patient’s prior function, their perception of the patient’s preference regarding limitation of life-support, their prediction of the patient’s ICU and hospital survival, and their prediction of the patient’s functional and cognitive status one month after hospital discharge. The primary outcome was a DNR directive established during the ICU stay, more than 24 hr after ICU admission. All patients were followed until hospital discharge.

Statistical analysis
We present continuous variables as means and standard deviations (±SD) or medians and interquartile range (IQR) for skewed data. We report rates using proportions and 95% confidence intervals. Using life table analysis, we established the hazard rate for establishing DNR directives. We compared the differences in important patient characteristics and outcomes between patients with and without DNR directives. To further characterize patients for whom DNR directives were established in the ICU, we categorized these patients into those who underwent withdrawal of at least one form of life-support and those who did not.

To determine the factors associated with DNR directives established during the ICU admission, we used a Cox proportional hazards regression analysis for time-dependent outcomes and time-dependent explanatory variables.²⁷ We identified six baseline factors potentially associated with DNR decisions including patient age (< 50 yr, 50–64 yr, 65–74 yr, 75 yr), sex, APACHE II score (five unit interval), medical vs surgical admitting diagnosis, physician’s estimation of patient’s prior functional status (poor, moderate, good), and patient’s ability to participate in decisions. In addition, daily factors were included in the univariable analysis, excluding values 24 hr before the DNR directive: illness severity factors [MODS (five unit interval), inotropes, hemodialysis], physician’s clinical judgements [physician prediction of ICU and hospital survival (< 10, 10–40%, 41–60%, > 60%)], physician prediction of patient’s future functional status and future cognitive function one month after hospital discharge (don’t believe patient will leave hospital, severely limited, somewhat limited, totally independent), physician perception of patient preference to limit lifesupport (no advanced life-support, partial advanced life-support, full advanced life-support), and centre. We tested for all pairwise interactions among the daily factors in the multivariable model and their interactions with time, and tested the proportional hazards assumption of all variables using the time-dependent covariate function.²⁷

All factors in the univariable analysis associated with DNR directives with a P value < 0.1 were entered into the full multivariable model. We then used a forward stepwise likelihood ratio method for the final multivariable analysis. We ordered the variables to include patient factors, centre, and then physician clinical judgements. Variables were retained in the final model if P < 0.01. These analyses were performed using SPSS 11.0^TM (Chicago, IL, USA).

	Results

TOP Abstract Introduction Methods Results Discussion References

 
Of 3,099 patients included in the original cohort,²³ 851 were predicted to have an ICU stay of > 72 hr and were mechanically ventilated for > 48 hr. Of these 851 patients, 86 were excluded as they had an explicit CPR directive established either before or during the first 24 hr of ICU admission. Thus, we included 765 patients in this study with a mean age of 60.0 (±17.7) yr, and a mean APACHE II score of 21.2 (±8.5). Other baseline characteristics are presented in Table I.

View larger version (27K): [in this window] [in a new window]   FIGURE We present the day in the ICU on which new DNR directives were established for 231 of the total of 765 mechanically ventilated patients. The peak rate of establishing DNR directives was on day four of the ICU stay. DNR = do-not-resuscitate; ICU = intensive care unit.

	Discussion

TOP Abstract Introduction Methods Results Discussion References

 
In this multicentre study of mechanically ventilated patients with no DNR directive within 24 hr of admission to ICU, we found that a DNR directive was established for 30% of patients. Approximately 50% of the DNR directives were established in the first week of the ICU stay. In our study, DNR directives were established earlier than in studies from the 1990’s, in which they were often documented late in the course of critical illness, just before withholding or withdrawal of lifesupport measures,^12,19,28 or just after a cardiopulmonary arrest.¹⁷ Our results are similar to a recent large European study in which limitation of life sustaining therapy occurred a median of three days after ICU admission.²⁹ The current trend towards earlier decisions about CPR compared to the past may represent temporal changes in ICU practice, particularly regarding end-of-life care.

We found that the factor most strongly associated with the decision to establish a DNR directive was the physician’s prediction of patients’ survival. It is surprising that it was not only physician predictions of high likelihood of death that was associated with DNR directives, but also moderate likelihood of death. While the odds of a DNR decision increased fifteen-fold when clinicians predicted ICU survival of < 10% (compared to probability of surviving of > 90%), the odds increased fivefold with predictions of between 10 and 40% probability of ICU survival, fourfold with 41 to 60% probability of ICU survival, and almost twofold with 61 to 90% probability of ICU survival (Table A, available as Additional Material at www.cja-jca.org). We hypothesize that regardless of the patient’s probability of surviving critical illness, if and when physicians believe that CPR is unlikely to be successful in the event of a cardiopulmonary arrest during the ICU course, they may convey their beliefs to families in a manner that significantly influences the DNR decision. Thus, over the course of the ICU admission, a model of shared decision-making may evolve between physicians and surrogate decision-makers such that a DNR directive is ultimately established.

The association of DNR directives with age is another important finding of this study. This relationship was not restricted to the elderly, but was also true of patients over 50 yr of age. We previously found that age was strongly associated with DNR decisions made within 24 hr of ICU admission.²³ In contrast, in the previous study we did not find that age was an independent determinant of the decision to withdraw mechanical ventilation.³⁰ These findings suggest differences in the process of decision-making between DNR directives and life-support withdrawal, in that age appears to influence resuscitation plans in the event of a cardiopulmonary arrest, but is much less likely to influence the more complex decisions to withdraw life-support.

We previously found that severe premorbid functional impairment was associated with DNR directives established during the first 24 hr of ICU admission.³¹ However, in the current study, we found no independent association between prior functional status and DNR directives. An inconsistent relationship between functional status and resuscitation decisions has been identified in other studies, some of which have found an association,¹³ while others have not.³² One potential explanation for the discrepancy between the determinants of initial and subsequent resuscitation decisions in the ICU is that prior functional status may influence initial plans to limit resuscitation in the event of cardiopulmonary arrest. However, as the course of critical illness evolves and the ICU team becomes familiar with the patient, resuscitation decisions in the event of a cardiopulmonary arrest, and subsequent withdrawal of life-support are influenced more strongly by perceptions of patient preferences, and physician predictions about survival and future cognitive function.³⁰

There are other important differences and similarities between the determinants of DNR directives established within the first 24 hr of ICU admission²³ compared with those following 24 hr of ICU admission, as we have found in this study. Similar to the first 24 hr of ICU admission,²³ we also found that medical (compared to surgical) diagnosis was an important determinant of a DNR directive. However, unlike the first 24 hr of ICU admission, patients’ inability to participate in resuscitation discussions was not found to be a significant predictor of DNR directives following 24 hr of ICU admission. We did not test whether the presence of a legal power of attorney, or the day and time of ICU admission were predictors of DNR directives following 24 hr of ICU admission.

Our results build on the findings of the original SUPPORT study of 6,802 seriously ill hospitalized patients.¹² Multivariable regression analysis of that cohort¹³ found that older age, patient self-reported prior functional impairment and prior quality of life, the probability of survival for two months based on the SUPPORT prognostic model, and the patients’ stated preference for CPR were associated with early DNR decisions. In two prospective multicentre ICU studies performed more than a decade ago in the United States, severity of illness based on the APACHE II system was the factor most strongly associated with DNR decisions during the ICU admission; older age and prior poor functional status were also associated.^18,33 Therefore, although clinical practice and societal values may change over time, our study affirmed earlier findings that age and illness severity, as measured by organ dysfunction, remain important determinants of DNR directives established after ICU admission.

While our prior study of the likelihood of establishing DNR decisions during the first 24 hr of ICU admission found systematic differences between centres,²³ we found no significant between-centre differences in this study. These results also contrast with a recent large, multicentre, prospective observational, European study (ETHICUS; 29) in which regional religious and cultural differences were explored more fully, and proved to be important determinants of decisions regarding limitations of life-sustaining therapies. We did not assess the influence of cultural or religious diversity on DNR directives in our study. Whether the findings of the ETHICUS study²⁹ are generalizable to North America is a worthy consideration for future research.

Strengths of this study include a focus exclusively on a cohort of heterogeneous mechanically ventilated ICU patients. A prospective multicentre design, enrollment of 765 patients in 15 centres, comprehensive data collection, and complete follow-up increase the precision of our findings. We used a Cox proportional hazards regression analysis to examine the baseline and time dependent factors most strongly associated with DNR directives, and we described the life-support withdrawal and outcomes of these patients.

This study has several limitations. We did not include patient comorbid illness as a potential predictor of DNR directives. In addition, we did not use a validated instrument to record physicians’ perceptions of patients’ prior functional status or physicians’ predictions of patients’ future survival. We did not validate physician predictions of future functional status with an assessment of patients’ actual functional status following ICU discharge. The generalizability of these results is strongest for patients treated in similar university-affiliated centres.

In summary, we found that many DNR directives for mechanically ventilated patients occur within the first week of an ICU admission, suggesting a pattern of documenting important end-of-life treatment preferences earlier in the course of critical illness than previously reported. Among patients with DNR orders during their ICU stay who do not undergo withdrawal of life-support, as part of a terminal weaning plan, one third survive their hospitalization, underscoring how factors other than the probability of survival determine these decisions. We showed that the independent factors associated with DNR decisions were: patient age, a medical diagnosis, organ dysfunction, physician prediction of a low probability of ICU survival, and physician perception of patient preference to limit advanced life-support. This study suggests that there are many dimensions to establishing DNR decisions in the ICU, and highlights the need for clear communication between physicians and surrogate decision-makers about patient preferences for resuscitation prior to, or early in the course of an acute illness. Further research about patients’ and surrogates’ understanding and expectations of life-support technology, and increasing realistic, sensitive discussions about resuscitation preferences prior to critical illness are imperative to improve end-of-life care.

	Acknowledgments

 
We dedicate this study to the late Dr. Peter Sjokvist who contributed so much to the field of end-of-life care, and whose insights were seminal in conducting and interpreting this study. We appreciate the support of the Canadian Critical Care Trials Group, and thank the research nurses, bedside nurses, residents and attending physicians who participated in this study.

	Footnotes

 
This study was funded by the following: a) National and Provincial Agencies: Medical Research Council of Canada; Department of Health, Province of Nova Scotia; Physicians Services Incorporated of Ontario; British Columbia Medical Services Foundation; and the Research Committee of the Orebro County Council, Sweden. b) Regional Peer-Review Agencies: Father Sean O’Sullivan Research Centre, St. Joseph’s Hospital; Health Services Research Fund, London Health Sciences Centre; University Internal Medicine Research Fund, Dalhousie University; Faculty of Medicine Intramural Grant, Dalhousie University; Camphill Medical Centre Research Fund, Halifax; and the Queen Elizabeth II Health Sciences Research Fund, Halifax. c) Career Support: T. Sinuff is a Research Fellow of the Canadian Institutes for Health Research. D. J. Cook is a Research Chair of the Canadian Institutes for Health Research. S. D. Walter is a Senior Investigator of the Canadian Institutes for Health Research.

Accepted for publication March 22, 2004. Revision accepted September 10, 2004.

	References

TOP Abstract Introduction Methods Results Discussion References

 
1 Peterson MW, Geist LJ, Schwartz DA, Konicek S, Moseley PL. Outcome after cardiopulmonary resuscitation in a medical intensive care unit. Chest 1991; 100: 168–74.[Abstract/Free Full Text]

2 Landry F, Parker JM, Phillips YY. Outcome of cardiopulmonary resuscitation in the intensive care setting. Arch Intern Med 1992; 152: 2305–8.[Abstract]

3 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: 394–401.[Medline]

4 Karetzky M, Zubair M, Parikh J. Cardiopulmonary resuscitation in intensive care unit and non-intensive care unit patients. Immediate and long-term survival. Arch Intern Med 1995; 155: 1277–80.[Abstract]

5 Wallace SK, Ewer MS, Price KJ, Feeley TW. Outcome and cost implications of cardiopulmonary resuscitation in the medical intensive care unit of a comprehensive cancer centre. Support Care Cancer 2002; 10: 425–9.[Medline]

6 Myrianthefs P, Kalafati M, Lemonidou C, et al. Efficacy of CPR in a general, adult ICU. Resuscitation 2003; 57: 43–8.[Medline]

7 Di Bari M, Chiarlone M, Fumagalli S, et al. Cardiopulmonary resuscitation of older, inhospital patients: immediate efficacy and long-term outcome. Crit Care Med 2000; 28: 2320–5.[Medline]

8 Bruce-Jones P, Roberts H, Bowker L, Cooney V. Resuscitating the elderly: what do the patients want? J Med Ethics 1996; 22: 154–9.[Abstract]

9 Heffner JE, Fahy B, Hilling L, Barbieri C. Attitudes regarding advance directives among patients in pulmonary rehabilitation. Am J Respir Crit Care Med 1996; 154: 1735–40.[Abstract]

10 Heffner JE, Barbieri C. End-of-life care preferences of patients enrolled in cardiovascular rehabilitation programs. Chest 2000; 117: 1474–81.[Abstract/Free Full Text]

11 Agard A, Hermeren G, Herlitz J. Should cardiopulmonary resuscitation be performed on patients with heart failure? The role of the patient in the decision-making process. J Intern Med 2000; 248: 279–86.[Medline]

12 The SUPPORT Principal Investigators. A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks of treatment (SUPPORT). JAMA 1995; 274: 1591–8.[Abstract]

13 Hakim RB, Teno JM, Harrell FE Jr, et al. Factors associated with do-not-resuscitate orders: patients’ preferences, prognoses, and physicians’ judgments. Study to understand prognoses and preferences for outcomes and risks of treatment. SUPPORT Investigators. Ann Intern Med 1996; 125: 284–93.[Abstract/Free Full Text]

14 Krumholz HM, Phillips RS, Hamel MB, et al. Resuscitation preferences among patients with severe congestive heart failure: results from the SUPPORT project. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments. Circulation 1998; 98: 648–55.[Abstract/Free Full Text]

15 Suhl J, Simons P, Reedy T, Garrick T. Myth of substituted judgment. Surrogate decision making regarding life support is unreliable. Arch Intern Med 1994; 154: 90–6.[Abstract]

16 Seckler AB, Meier DE, Mulvihill M, Cammer Paris BE. Substituted judgment: how accurate are proxy predictions? Ann Intern Med 1991; 115: 92–8.

17 Kernerman P, Cook DJ, Griffith LE. Documenting life-support preferences in hospitalized patients. J Crit Care 1997; 12: 155–60.[Medline]

18 Jayes RL, Zimmerman JE, Wagner DP, Knaus WA. Variations in the use of do-not-resuscitate orders in ICUs. Findings from a national study. Chest 1996; 110: 1332–9.[Abstract/Free Full Text]

19 Rapoport J, Teres D, Lemeshow S. Resource use implications of do not resuscitate orders for intensive care unit patients. Am J Respir Crit Care Med 1996; 153: 185–90.[Abstract]

20 Eidelman LA, Jakobson DJ, Pizov R, Geber D, Leibovitz L, Sprung CL. Foregoing life-sustaining treatment in an Israeli ICU. Intensive Care Med 1998; 24: 162–6.[Medline]

21 Prendergast TJ, Luce JM. Increasing incidence of withholding and withdrawing of life support from the critically ill. Am J Respir Crit Care Med 1997; 155: 15–20.[Abstract]

22 Jayes RL, Zimmerman JE, Wagner DP, Draper EA, Knaus WA. Do-not-resuscitate orders in intensive care units. Current practices and recent changes. JAMA 1993; 270: 2213–7.[Abstract]

23 Cook DJ, Guyatt G, Rocker G, et al.; Canadian Critical Care Trials Group. Cardiopulmonary resuscitation directives on admission to intensive care unit: an international observational study. Lancet 2001; 358: 1941–5.[Medline]

24 Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13: 818–25.[Medline]

25 Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995; 23: 1638–52.[Medline]

26 Pettila V, Pettila M, Sarna S, Voutilainen P, TakkuneN O. Comparison of multiple organ dysfunction scores in the prediction of hospital mortality in the critically ill. Crit Care Med 2002; 30: 1705–11.[Medline]

27 Collett D. Modelling Survival Data in Medical Research. New York: Chapman and Hall; 1994: 149–98.

28 Sprung CL, Eidelman LA, Pizov R, et al. Influence of alterations in forgoing life-sustaining treatment practices on a clinical sepsis trial. The HA-1A Sepsis Study Group. Crit Care Med 1997; 25: 383–7.[Medline]

29 Sprung CL, Cohen SL, Sjokvist P, et al.; Ethicus Study Group. End-of-life practices in European intensive care units. The Ethicus Study. JAMA 2003; 290: 790–7.[Abstract/Free Full Text]

30 Cook D, Rocker G, Marshall J, et al. for the Level of Care Study Investigators and The Canadian Critical Care Trials Group. Withdrawal of mechanical ventilation in anticipation of death in the intensive care unit. N Engl J Med 2003; 349: 1123–32.[Abstract/Free Full Text]

31 Guyatt G, Cook D, Weaver B, et al.; Level of Care Study Investigators and the Canadian Critical Care Trials Group. Influence of perceived functional and employment status on cardiopulmonary resuscitation directives. J Crit Care 2003; 18: 133–41.[Medline]

32 Goodlin SJ, Zhong Z, Lynn J, et al. Factors associated with use of cardiopulmonary resuscitation in seriously ill hospitalized adults. JAMA 1999; 282: 2333–9.[Abstract/Free Full Text]

33 Zimmerman JE, Knaus WA, Sharpe SM, Anderson AS, Draper EA, Wagner DP. The use and implications of do not resuscitate orders in intensive care units. JAMA 1986; 255: 351–6.[Abstract]

This article has been cited by other articles:

				D. Cook, G. Rocker, J. Marshall, L. Griffith, E. McDonald, G. Guyatt, and for the Level of Care Study Investigators and the Levels of Care in the Intensive Care Unit: A Research Program Am. J. Crit. Care., May 1, 2006; 15(3): 269 - 279. [Abstract] [Full Text] [PDF]

This Article Abstract Résumé de cet Article Full Text (PDF) Additional Material Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sinuff, T. Search for Related Content PubMed PubMed Citation Articles by Sinuff, T.