- Email: [email protected]
Interpretation and intent: A study of the (mis)understanding of DNAR orders in a teaching hospital
Resuscitation 81 (2010) 1138–1141
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Interpretation and intent: A study of the (mis)understanding of DNAR orders in a teaching hospital夽,夽夽 Zoë Fritz a,∗ , Jonathan Fuld a , Stephen Haydock a , Chris Palmer b a b
Department of Acute Medicine, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK Centre of Applied Medical Statistics, Cambridge University, Cambridge, UK
a r t i c l e
i n f o
Article history: Received 17 February 2010 Received in revised form 5 May 2010 Accepted 23 May 2010 Keywords: Resuscitation orders Quality of health care Attitude of health personnel Interdisciplinary communication
a b s t r a c t Do not attempt resuscitation (DNAR) orders have been shown to be subject to misinterpretation in the 1980s and 1990s. We investigated whether this was still the case, and examined what perceptions doctors and nurses had of what care patients with DNAR orders receive. Methods: Using an anonymous written questionnaire, we directly approached 50 doctors and 40 nurses from a range of medical specialities and grades in our teaching hospital. Results: All 50 physicians and 35/40 nurses took part. Using McNemar’s test, there were highly significant differences (p < 0.0001) in what doctors believed ‘should’ take place and what they perceived ‘in practice’ occurred on patients with DNAR orders in all areas questioned (e.g., frequency of nursing observations and contacting medical staff in the event of a patient’s deterioration). Using Fisher’s exact test, there were significant differences between what nursing staff thought occurred and what doctors thought should occur, for example, frequency of nursing observations (p < 0.001), contacting the medical team (p = 0.01) and giving fluids (p < 0.005). Conclusions: Despite widespread use of DNAR orders, they are still misunderstood. This article highlights the frequency with which DNAR orders are interpreted to mean that other care should be withheld. In addition, it shows that although some doctors know that this should not be the case, they believe that DNAR orders affect the care that their patients receive. We propose that options for more detailed care plans should be embedded within the resuscitation decision and documentation to improve communication and understanding. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Do not attempt resuscitation (DNAR) orders are likely to affect all people, directly or indirectly: A total of 68% of the population die in hospital,1 and of those, 80% die with DNAR order forms in place.2 DNAR order forms not only affect those who are dying: many patients with DNAR order forms are discharged home.
夽 A Spanish translated version of the abstract of this article appears as Appendix in the final online version at doi:10.1016/j.resuscitation.2010.05.014. 夽夽 The corresponding author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive licence (or non-exclusive for government employees) on a worldwide basis to the BMJ Publishing Group Ltd. and its Licensees to permit this article (if accepted) to be published in JME editions and any other BMJPGL products to exploit all subsidiary rights, as set out in our licence (http://group.bmj.com/products/journals/instructions-for-authors/licence-forms). ∗ Corresponding author at: Department of Acute Medicine, Box 275, Addenbrooke’s Hospital, Hills Road, Cambridge, Cambridgeshire, CB2 0QQ, UK. Tel.: +44 1223 348320; fax: +44 1223 348319. E-mail address: [email protected] (Z. Fritz). 0300-9572/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2010.05.014
DNAR order forms instruct the clinical team what to do in the event of a cardiac or respiratory arrest. They are not meant to affect other aspects of care.3 They are filled in either when a physician believes that a patient would not benefit from attempted resuscitation or at a patient’s request. They are not considered routinely: a DNAR order is an additional piece of paperwork inserted in the front of the notes. This is so that this important decision, when made, can be immediately accessible and instantly recognisable in the event of someone arresting. Anecdotally, we have observed that the DNAR order can be interpreted as a negative signal and have noticed that some physicians will not fill out such forms for fear that their patients may receive suboptimal therapy. In the 1980s, shortly after the concept of do not resuscitate (DNR) was first introduced, several studies explored what the staff had understood by the term. In 1984, Uhlmann et al.4 showed both variability and uncertainty in the interpretation of DNR amongst practising physicians. La Puma et al.,5 in 1988, showed that different physicians looking after the same patient understood the DNR to mean a wide variety of actions, including, in some cases, withholding dialysis (86%) and intravenous (IV) fluids (21%). A more
Z. Fritz et al. / Resuscitation 81 (2010) 1138–1141
1139
Table 1 Summary of doctors’ and nurses’ answers with statistical comparison. Doctors were asked:
Frequency of nursing observations Contacting outreach Contacting medical team Out-of-hours medical escalation Pain relief given Fluids given
Nurses
McNemar’s p-value: 1 vs. 2
Fisher’s exact p: 1 vs. 3
1. “Do you think that a DNAR decision should alter. . .”
2. “In practice, do you think that a DNAR decision alters. . .”
3. “Do you think a DNAR decision alters. . .”
3/50 (6%)
36/50 (72%)
15/35 (42.8%)
p < 0.001
p < 0.001
22/50 (44%)
45/50 (90%)
16/35 (45.7%)
p < 0.001
p = 0.055
5/50 (10%)
42/50 (84%)
7/35 (20%)
p < 0.001
p = 0.01
20/50 (40%)
43/50 (86%)
17/33 (51.5%)
p < 0.001
p = 0.17
5/50 (10%) 4/50 (8%)
27/50 (54%) 26/50 (52%)
14/34 (41%) 13/34 (38%)
p < 0.001 p < 0.001
p < 0.005 p < 0.005
Difference between doctors’ paired answers for (1) what should occur and (2) what occurs in practice assessed using McNemar’s test. Difference between (1) what doctors believe should be altered and (3) what nurses state does alter assessed using Fisher’s exact test.
recent questionnaire on staff awareness of DNAR orders (the ‘A’ for ‘attempt’ was added in the 1990s to improve the poor success rates) carried out by Smith et al.6 in a District General Hospital showed inconsistency about which information staff felt should be included in DNAR order documentation and what, if any, continuing care should be given to patients, who are not for resuscitation. We designed a questionnaire to examine what care nurses and doctors considered appropriate for patients with DNAR orders; in addition, we assessed what physician perceptions were of the care actually given. 2. Methods The Regional Ethics Committee reviewed the project and concluded that the work fell into the category of ‘audit’. A total of 50 physicians participating in the general medical take were approached at morning handover, over the course of 2 weeks. They were asked to complete a written questionnaire anonymously, returning it immediately after completion. They were told that it would take about 5 min to complete and that it was investigating DNAR decisions and subsequent practice. Along with other multiple-choice questions, doctors were asked, “Do you think that a DNAR decision should alter: (a) frequency of nursing observations; (b) nursing staff contacting outreach team when a patient deteriorates; (c) nursing staff contacting medical staff when a patient deteriorates; (d) out-of-hours medical escalation in the event of deterioration; (e) pain relief given; (f) fluids given; and (g) prognosis.” For each phrase, the doctors were asked to identify if a DNAR order made the specific action (a)–(g) more likely, less likely or had no effect. On the following page of the questionnaire (so that this would not bias the answers to the first question), the doctors were additionally asked: “In practice, do you think that a DNAR decision alters. . . (the phrase being completed using the actions (a)–(g) above).” A total of 40 ward nurses were asked to complete a slightly different written questionnaire about DNAR decisions, anonymously. Along with other multiple-choice questions, the nurses were asked, “Do you think that a DNAR decision alters.. (the phrase being completed using the actions (a)–(g) above).” For
each phrase, the doctors were asked to identify if a DNAR order made the specific action (a)–(g) more likely, less likely or had no effect. The data were analysed using SPSS software. McNemar’s test was used for the paired data (doctors’ answers) and Fisher’s exact test for the comparison between nurses’ and doctors’ answers. The accuracy of any estimated percentage, with 95% confidence interval, based on a sample of size 50 (or 35) individuals, is no more than ±14% (or ±17%). 3. Results All 50 physicians (15 consultants and 35 registrars) took part. Of the 40 nurses approached, five declined, saying that they were not confident in answering questions about DNAR orders. Of those who did take part, 25 nurses were from General Medicine and 10 nurses from Care of the Elderly wards. Five nurses did not identify which area they worked in. There were 15 Staff Nurses, nine Senior Staff, six Junior Sisters and one Senior Sister. Four did not identify their grade. A summary of all of the responses, with statistical comparison is provided in Table 1. When nurses were asked, “Do you think a DNAR decision alters..?” the results were as shown in Fig. 1. A total of 43% believed that DNAR orders altered the frequency of nursing observations. In the event of a patient deteriorating, many believed that DNAR orders altered how often nursing staff contacted Outreach teams (46%) and medical staff (20%) and whether out-of-hours medical escalation would occur (51%). In addition, 40% believed that DNAR orders altered pain relief given, while 38% also believed it altered fluids given.
Fig. 1. Nurses’ answers (expressed as percentage) to “Do you think a DNAR Decision alters. . .”.
1140
Z. Fritz et al. / Resuscitation 81 (2010) 1138–1141
Fig. 2. Doctors’ answers to “How frequently do you discuss other interventions (ventilation, direct current cardioversion (DCCV) and inotropes)?”.
Fig. 3. Doctors’ answers (expressed as percentage) to “Do you think a DNAR decision should alter. . .”.
Fig. 4. Doctors’ answers (expressed as percentage) to: “In practice, do you think that aDNARdecision alters. . .”.
Doctors were initially asked: “How frequently do you have a DNAR discussion with a patient or their family before filling out the DNAR form?” and “How frequently do you discuss other interventions (ventilation, direct current cardioversion (DCCV) and inotropes)?” and reported as depicted in Fig. 2. Results from the answers doctors gave in response to the questions: “Do you think a DNAR decision should alter. . .” and, “In practice, do you think a DNAR decision alters. . .” are shown in Figs. 3 and 4, respectively. A total of 70% of physicians believed that nursing observation frequency was reduced in the event of a patient deteriorating. As many as 90% of physicians believed that nursing staff contacted outreach less, while 84% believed medical teams were contacted
less and 86% that out-of-hours medical escalation was given to a lesser extent. A total of 54% believed that the amount of pain relief given was altered and 92% believed that different amounts of fluids were given. Using McNemar’s test for paired data, there are highly significant differences (p < 0.0001) between what doctors think should occur with a DNAR order, and what they believe occurs in practice for each of the categories. Doctors and nurses were also given an open question about how to improve the form. Nine nurses answered the question, of which seven were about format. The remaining two were: “more elegant appearance . . .–. . . more teaching on form” and “It needs to be clear whether active treatment is required or not due to implications for care.” Seventeen doctors provided comments, of which 10 included comments about the need for more options to be included, for example: “Include specific boxes forcing decisions on escalation of therapy (intensive care unit (ICU)/high dependency unit (HDU)/noninvasive ventilation (NIV))”; “In some circumstances, it would be useful to document the ceiling of treatment as discussed with the team, the patient and the family, that is, for NIV but not for intubation, etc.”; “More scope for levels of care – . . .DNAR but. . .”; “different types of ‘R’ (Resuscitation)’; “more information available regarding: other decisions, for example, for ICU/intubation and DCCV’. The remaining seven comments were concerned with visibility, consultant signature and time for review. Comparison was made between what nurses thought was altered by a DNAR order form, and what doctors thought should be altered. A significant difference was seen, using Fisher’s exact test, for frequency of nursing observations (p < 0.001), contacting a medical team (p = 0.01), contacting outreach (p = 0.055), fluids given (p < 0.005) and pain relief given (p < 0.005). No significant difference was seen for out-of-hours medical escalation (p = 0.17).
4. Discussion In this questionnaire, variability was revealed in both what effect doctors thought a DNAR order ‘should’ have, and what nurses thought occurred. A worrying proportion of nurses believed that it reduced out-of-hours medical escalation, contacting the outreach team and frequency of nursing observations. Similarly, an alarming proportion of doctors recorded that they believed that those with DNAR orders should have reduced referral to outreach and medical teams. Both groups showed more variability in their answers than one would hope for, given the joint guidelines stating, “DNAR order decisions apply only to cardiopulmonary resuscitation (CPR) and not to any other aspects of treatment”.8,3 Importantly, a significant difference was seen between what nurses and doctors believed occurred when a DNAR order was in place for the questions of frequency of nursing observations, contacting outreach, contacting a medical team and the amount of fluids and pain relief given. When doctors were asked whether they believed, ‘in practice’ that changes occurred in patient care, the vast majority believed that nursing observations, referral to outreach and out-of-hours medical escalation were reduced. There were highly significant differences between what doctors think should occur with a DNAR order, and what they believe occurs in practice for each of the categories. These findings have major implications. First, it is clearly common for the DNAR order form to be misinterpreted by both doctors and nurses to mean that care should be reduced; thus, the concern is that patients with DNAR orders may be receiving substandard care. In fact, several studies have suggested that this is the case: Beach and Morrison7 carried out a survey of doctors on hypotheti-
Z. Fritz et al. / Resuscitation 81 (2010) 1138–1141
cal patients, and demonstrated that the presence of a DNAR order had an effect on decision making. Other studies have shown that those with DNAR orders have less intensive care [8,9] and increased mortality [10,11]. Second, the perception that those with DNAR orders receive substandard care is pervasive: in every category, significantly more doctors believed that, in practice, DNAR orders led to less care than they believed should occur. This widespread belief may lead to doctors being less willing to fill out DNAR order forms on patients, even in circumstances where they do not think attempted CPR would be appropriate, for fear of diminishing the overall level of care that the patient receives. The questionnaire also revealed that 82% of all doctors usually or almost always discussed resuscitation decisions with patients, while 62% usually or almost always discussed additional interventions. Despite this being a common practice, there is no scope on most current DNAR order forms (including the improved national one currently being piloted by the Resuscitation Council) for documentation of such decisions. Many of the comments on the physician questionnaires related to the desire to have “more options”. The study has several limitations. The questionnaire was not validated, and no retest questions were done. In asking the nursing staff the question “Do you think a DNAR alters?” we were trying to capture how they interpreted the form, and might act on it. They were not asked the question in two parts because they are not required to fill in DNAR orders; however, it is possible that there was variability in the interpretation of the question. The nursing staff may have felt pressured in replying to the questionnaires in the ward setting. Nevertheless, these data represent the beliefs of 85 members of staff at a particular point in time and is therefore relevant to the kind of day-to-day interpretations of DNAR orders that we are trying to capture. 5. New directions? One approach to the problems expanded upon above would be to educate health-care personnel better, both in their initial training and as an ongoing part of, for example, the adult life support (ALS) course that many go on. There is, however, little in the literature to support this; educational attempts to improve DNAR or end-oflife discussions have proven ineffectual,12 and no studies, to our knowledge, have yet been carried out assessing the impact of education programmes on health-care professionals’ understanding of DNAR. At the end of our questionnaire, in the “How would you improve the current form?” section several doctors and one nurse suggested that, instead of the DNAR order, a more detailed form would improve communication. Some such forms have been used elsewhere, and their impact on documentation,13,14 rate of CPR attempts and ventilation15 assessed. In addition, advanced care plans have been developed to improve end-of-life care for palliative patients (physician orders for life-sustaining treatment (POLST)16 ) or for those specifically at risk of deterioration (TEP,17 ). Following this questionnaire, we have developed a universal form of treatment options (UFTO), to address some of the issues raised, and are currently in the process of assessing this approach for its effect on both quantitative and qualitative outcomes. Given new working patterns, including multiple shift changes and less continuity of care, the problems of miscommunication surrounding DNAR orders are intensified.
1141
New approaches to deal with the issues surrounding DNAR orders are required, while maintaining the clarity of instruction that the current DNAR provides in the event of a patient arresting. A new approach should address the ad hoc nature of completing the form, the stigma associated with a DNAR order and the ambiguity about what do when a patient deteriorates that are clearly still prevalent with the DNAR order. With an appropriately evaluated new form and process, perhaps we can improve the care that those who are not for resuscitation receive. Conflict of interest The authors do not have any conflicts of interest. Acknowledgements We would like to thank the audit department at Addenbrooke’s Hospital and, in particular, Amara Siddique for her initial work with the data, Lorna Mason for her valuable administrative help and Angela Fritz for her tireless reviewing of the article and for the useful suggestions she made. In addition, we thank the NIHR Research for Patient Benefit Programme, who are supporting related work that three of the authors (ZF, JF and CP) are currently undertaking, although they have not had any direct role in this article. References 1. N. Office of Statistics Series DH1 (no 38, 2005). Table 17, pgs 56–7, http://www. statistics.gov.uk/StatBase/Product.asp?vlnk=620&Pos=1&ColRank=2&Rank=272 (accessed 4th May). 2. Aune S, Herlitz J, Bang A. Characteristics of patients who die in hospital with no attempt at resuscitation. Resuscitation 2005;65:291–9. 3. Decisions relating to cardiopulmonary resuscitation. A joint statement from the British Medical Association, the Resuscitation Council (UK) and the Royal College of Nursing; 2007. 4. Uhlmann RF, Cassel CK, McDonald WJ. Some treatment-withholding implications of no-code orders in an academic hospital. Crit Care Med 1984;12: 879–81. 5. La Puma J, Silverstein MD, Stocking CB, Roland D, Siegler M. Life-sustaining treatment. A prospective study of patients with DNR orders in a teaching hospital. Arch Intern Med 1988;148:2193–8. 6. Smith GB, Poplett N, Williams D. Staff awareness of a ‘Do Not Attempt Resuscitation’ policy in a District General Hospital. Resuscitation 2005;65: 159–63. 7. Beach MC, Morrison RS. The effect of do-not-resuscitate orders on physician decision-making. J Am Geriatr Soc 2002;50:2057–61. 8. Chen JL, Sosnov J, Lessard D, Goldberg RJ. Impact of do-not-resuscitation orders on quality of care performance measures in patients hospitalized with acute heart failure. Am Heart J 2008;156:78–84. 9. Lipton HL. Do-not-resuscitate decisions in a community hospital, incidence, implications, and outcomes. JAMA 1986;256:1164–9. 10. Shepardson LB, Youngner SJ, Speroff T, Rosenthal GE. Increased risk of death in patients with do-not-resuscitate orders. Med Care 1999;37:727–37. 11. Wenger NS, Pearson ML, Desmond KA, Brook RH, Kahn KL. Outcomes of patients with do-not-resuscitate orders. Toward an understanding of what donot-resuscitate orders mean and how they affect patients. Arch Intern Med 1995;155:2063–8. 12. Shorr AF, Niven AS, Katz DE, Parker JM, Eliasson AH. Regulatory and educational initiatives fail to promote discussions regarding end-of-life care. J Pain Symptom Manage 2000;19:168–73. 13. Mittelberger JA, Lo B, Martin D, Uhlmann RF. Impact of a procedure-specific do not resuscitate order form on documentation of do not resuscitate orders. Arch Intern Med 1993;153:228–32. 14. O’Toole EE, Youngner SJ, Juknialis BW, Daly B, Bartlett ET, Landefeld CS. Evaluation of a treatment limitation policy with a specific treatment-limiting order page. Arch Intern Med 1994;154:425–32. 15. Davila F. The impact of do-not-resuscitate and patient care category policies on CPR and ventilator support rates. Arch Intern Med 1996;156:405–8. 16. Tolle SW, Tilden VP, Nelson CA, Dunn PM. A prospective study of the efficacy of the physician order form for life-sustaining treatment. J Am Geriatr Soc 1998;46:1097–102. 17. Mercer M. Re: DNAR or AND? In: BMJ online [eletter]; 2009, http://www.bmj. com/cgi/eletters/338/apr30 1/b1723#213249 (accessed 4th May).
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Interpretation and intent: A study of the (mis)understanding of DNAR orders in a teaching hospital夽,夽夽 Zoë Fritz a,∗ , Jonathan Fuld a , Stephen Haydock a , Chris Palmer b a b
Department of Acute Medicine, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK Centre of Applied Medical Statistics, Cambridge University, Cambridge, UK
a r t i c l e
i n f o
Article history: Received 17 February 2010 Received in revised form 5 May 2010 Accepted 23 May 2010 Keywords: Resuscitation orders Quality of health care Attitude of health personnel Interdisciplinary communication
a b s t r a c t Do not attempt resuscitation (DNAR) orders have been shown to be subject to misinterpretation in the 1980s and 1990s. We investigated whether this was still the case, and examined what perceptions doctors and nurses had of what care patients with DNAR orders receive. Methods: Using an anonymous written questionnaire, we directly approached 50 doctors and 40 nurses from a range of medical specialities and grades in our teaching hospital. Results: All 50 physicians and 35/40 nurses took part. Using McNemar’s test, there were highly significant differences (p < 0.0001) in what doctors believed ‘should’ take place and what they perceived ‘in practice’ occurred on patients with DNAR orders in all areas questioned (e.g., frequency of nursing observations and contacting medical staff in the event of a patient’s deterioration). Using Fisher’s exact test, there were significant differences between what nursing staff thought occurred and what doctors thought should occur, for example, frequency of nursing observations (p < 0.001), contacting the medical team (p = 0.01) and giving fluids (p < 0.005). Conclusions: Despite widespread use of DNAR orders, they are still misunderstood. This article highlights the frequency with which DNAR orders are interpreted to mean that other care should be withheld. In addition, it shows that although some doctors know that this should not be the case, they believe that DNAR orders affect the care that their patients receive. We propose that options for more detailed care plans should be embedded within the resuscitation decision and documentation to improve communication and understanding. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Do not attempt resuscitation (DNAR) orders are likely to affect all people, directly or indirectly: A total of 68% of the population die in hospital,1 and of those, 80% die with DNAR order forms in place.2 DNAR order forms not only affect those who are dying: many patients with DNAR order forms are discharged home.
夽 A Spanish translated version of the abstract of this article appears as Appendix in the final online version at doi:10.1016/j.resuscitation.2010.05.014. 夽夽 The corresponding author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive licence (or non-exclusive for government employees) on a worldwide basis to the BMJ Publishing Group Ltd. and its Licensees to permit this article (if accepted) to be published in JME editions and any other BMJPGL products to exploit all subsidiary rights, as set out in our licence (http://group.bmj.com/products/journals/instructions-for-authors/licence-forms). ∗ Corresponding author at: Department of Acute Medicine, Box 275, Addenbrooke’s Hospital, Hills Road, Cambridge, Cambridgeshire, CB2 0QQ, UK. Tel.: +44 1223 348320; fax: +44 1223 348319. E-mail address: [email protected] (Z. Fritz). 0300-9572/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2010.05.014
DNAR order forms instruct the clinical team what to do in the event of a cardiac or respiratory arrest. They are not meant to affect other aspects of care.3 They are filled in either when a physician believes that a patient would not benefit from attempted resuscitation or at a patient’s request. They are not considered routinely: a DNAR order is an additional piece of paperwork inserted in the front of the notes. This is so that this important decision, when made, can be immediately accessible and instantly recognisable in the event of someone arresting. Anecdotally, we have observed that the DNAR order can be interpreted as a negative signal and have noticed that some physicians will not fill out such forms for fear that their patients may receive suboptimal therapy. In the 1980s, shortly after the concept of do not resuscitate (DNR) was first introduced, several studies explored what the staff had understood by the term. In 1984, Uhlmann et al.4 showed both variability and uncertainty in the interpretation of DNR amongst practising physicians. La Puma et al.,5 in 1988, showed that different physicians looking after the same patient understood the DNR to mean a wide variety of actions, including, in some cases, withholding dialysis (86%) and intravenous (IV) fluids (21%). A more
Z. Fritz et al. / Resuscitation 81 (2010) 1138–1141
1139
Table 1 Summary of doctors’ and nurses’ answers with statistical comparison. Doctors were asked:
Frequency of nursing observations Contacting outreach Contacting medical team Out-of-hours medical escalation Pain relief given Fluids given
Nurses
McNemar’s p-value: 1 vs. 2
Fisher’s exact p: 1 vs. 3
1. “Do you think that a DNAR decision should alter. . .”
2. “In practice, do you think that a DNAR decision alters. . .”
3. “Do you think a DNAR decision alters. . .”
3/50 (6%)
36/50 (72%)
15/35 (42.8%)
p < 0.001
p < 0.001
22/50 (44%)
45/50 (90%)
16/35 (45.7%)
p < 0.001
p = 0.055
5/50 (10%)
42/50 (84%)
7/35 (20%)
p < 0.001
p = 0.01
20/50 (40%)
43/50 (86%)
17/33 (51.5%)
p < 0.001
p = 0.17
5/50 (10%) 4/50 (8%)
27/50 (54%) 26/50 (52%)
14/34 (41%) 13/34 (38%)
p < 0.001 p < 0.001
p < 0.005 p < 0.005
Difference between doctors’ paired answers for (1) what should occur and (2) what occurs in practice assessed using McNemar’s test. Difference between (1) what doctors believe should be altered and (3) what nurses state does alter assessed using Fisher’s exact test.
recent questionnaire on staff awareness of DNAR orders (the ‘A’ for ‘attempt’ was added in the 1990s to improve the poor success rates) carried out by Smith et al.6 in a District General Hospital showed inconsistency about which information staff felt should be included in DNAR order documentation and what, if any, continuing care should be given to patients, who are not for resuscitation. We designed a questionnaire to examine what care nurses and doctors considered appropriate for patients with DNAR orders; in addition, we assessed what physician perceptions were of the care actually given. 2. Methods The Regional Ethics Committee reviewed the project and concluded that the work fell into the category of ‘audit’. A total of 50 physicians participating in the general medical take were approached at morning handover, over the course of 2 weeks. They were asked to complete a written questionnaire anonymously, returning it immediately after completion. They were told that it would take about 5 min to complete and that it was investigating DNAR decisions and subsequent practice. Along with other multiple-choice questions, doctors were asked, “Do you think that a DNAR decision should alter: (a) frequency of nursing observations; (b) nursing staff contacting outreach team when a patient deteriorates; (c) nursing staff contacting medical staff when a patient deteriorates; (d) out-of-hours medical escalation in the event of deterioration; (e) pain relief given; (f) fluids given; and (g) prognosis.” For each phrase, the doctors were asked to identify if a DNAR order made the specific action (a)–(g) more likely, less likely or had no effect. On the following page of the questionnaire (so that this would not bias the answers to the first question), the doctors were additionally asked: “In practice, do you think that a DNAR decision alters. . . (the phrase being completed using the actions (a)–(g) above).” A total of 40 ward nurses were asked to complete a slightly different written questionnaire about DNAR decisions, anonymously. Along with other multiple-choice questions, the nurses were asked, “Do you think that a DNAR decision alters.. (the phrase being completed using the actions (a)–(g) above).” For
each phrase, the doctors were asked to identify if a DNAR order made the specific action (a)–(g) more likely, less likely or had no effect. The data were analysed using SPSS software. McNemar’s test was used for the paired data (doctors’ answers) and Fisher’s exact test for the comparison between nurses’ and doctors’ answers. The accuracy of any estimated percentage, with 95% confidence interval, based on a sample of size 50 (or 35) individuals, is no more than ±14% (or ±17%). 3. Results All 50 physicians (15 consultants and 35 registrars) took part. Of the 40 nurses approached, five declined, saying that they were not confident in answering questions about DNAR orders. Of those who did take part, 25 nurses were from General Medicine and 10 nurses from Care of the Elderly wards. Five nurses did not identify which area they worked in. There were 15 Staff Nurses, nine Senior Staff, six Junior Sisters and one Senior Sister. Four did not identify their grade. A summary of all of the responses, with statistical comparison is provided in Table 1. When nurses were asked, “Do you think a DNAR decision alters..?” the results were as shown in Fig. 1. A total of 43% believed that DNAR orders altered the frequency of nursing observations. In the event of a patient deteriorating, many believed that DNAR orders altered how often nursing staff contacted Outreach teams (46%) and medical staff (20%) and whether out-of-hours medical escalation would occur (51%). In addition, 40% believed that DNAR orders altered pain relief given, while 38% also believed it altered fluids given.
Fig. 1. Nurses’ answers (expressed as percentage) to “Do you think a DNAR Decision alters. . .”.
1140
Z. Fritz et al. / Resuscitation 81 (2010) 1138–1141
Fig. 2. Doctors’ answers to “How frequently do you discuss other interventions (ventilation, direct current cardioversion (DCCV) and inotropes)?”.
Fig. 3. Doctors’ answers (expressed as percentage) to “Do you think a DNAR decision should alter. . .”.
Fig. 4. Doctors’ answers (expressed as percentage) to: “In practice, do you think that aDNARdecision alters. . .”.
Doctors were initially asked: “How frequently do you have a DNAR discussion with a patient or their family before filling out the DNAR form?” and “How frequently do you discuss other interventions (ventilation, direct current cardioversion (DCCV) and inotropes)?” and reported as depicted in Fig. 2. Results from the answers doctors gave in response to the questions: “Do you think a DNAR decision should alter. . .” and, “In practice, do you think a DNAR decision alters. . .” are shown in Figs. 3 and 4, respectively. A total of 70% of physicians believed that nursing observation frequency was reduced in the event of a patient deteriorating. As many as 90% of physicians believed that nursing staff contacted outreach less, while 84% believed medical teams were contacted
less and 86% that out-of-hours medical escalation was given to a lesser extent. A total of 54% believed that the amount of pain relief given was altered and 92% believed that different amounts of fluids were given. Using McNemar’s test for paired data, there are highly significant differences (p < 0.0001) between what doctors think should occur with a DNAR order, and what they believe occurs in practice for each of the categories. Doctors and nurses were also given an open question about how to improve the form. Nine nurses answered the question, of which seven were about format. The remaining two were: “more elegant appearance . . .–. . . more teaching on form” and “It needs to be clear whether active treatment is required or not due to implications for care.” Seventeen doctors provided comments, of which 10 included comments about the need for more options to be included, for example: “Include specific boxes forcing decisions on escalation of therapy (intensive care unit (ICU)/high dependency unit (HDU)/noninvasive ventilation (NIV))”; “In some circumstances, it would be useful to document the ceiling of treatment as discussed with the team, the patient and the family, that is, for NIV but not for intubation, etc.”; “More scope for levels of care – . . .DNAR but. . .”; “different types of ‘R’ (Resuscitation)’; “more information available regarding: other decisions, for example, for ICU/intubation and DCCV’. The remaining seven comments were concerned with visibility, consultant signature and time for review. Comparison was made between what nurses thought was altered by a DNAR order form, and what doctors thought should be altered. A significant difference was seen, using Fisher’s exact test, for frequency of nursing observations (p < 0.001), contacting a medical team (p = 0.01), contacting outreach (p = 0.055), fluids given (p < 0.005) and pain relief given (p < 0.005). No significant difference was seen for out-of-hours medical escalation (p = 0.17).
4. Discussion In this questionnaire, variability was revealed in both what effect doctors thought a DNAR order ‘should’ have, and what nurses thought occurred. A worrying proportion of nurses believed that it reduced out-of-hours medical escalation, contacting the outreach team and frequency of nursing observations. Similarly, an alarming proportion of doctors recorded that they believed that those with DNAR orders should have reduced referral to outreach and medical teams. Both groups showed more variability in their answers than one would hope for, given the joint guidelines stating, “DNAR order decisions apply only to cardiopulmonary resuscitation (CPR) and not to any other aspects of treatment”.8,3 Importantly, a significant difference was seen between what nurses and doctors believed occurred when a DNAR order was in place for the questions of frequency of nursing observations, contacting outreach, contacting a medical team and the amount of fluids and pain relief given. When doctors were asked whether they believed, ‘in practice’ that changes occurred in patient care, the vast majority believed that nursing observations, referral to outreach and out-of-hours medical escalation were reduced. There were highly significant differences between what doctors think should occur with a DNAR order, and what they believe occurs in practice for each of the categories. These findings have major implications. First, it is clearly common for the DNAR order form to be misinterpreted by both doctors and nurses to mean that care should be reduced; thus, the concern is that patients with DNAR orders may be receiving substandard care. In fact, several studies have suggested that this is the case: Beach and Morrison7 carried out a survey of doctors on hypotheti-
Z. Fritz et al. / Resuscitation 81 (2010) 1138–1141
cal patients, and demonstrated that the presence of a DNAR order had an effect on decision making. Other studies have shown that those with DNAR orders have less intensive care [8,9] and increased mortality [10,11]. Second, the perception that those with DNAR orders receive substandard care is pervasive: in every category, significantly more doctors believed that, in practice, DNAR orders led to less care than they believed should occur. This widespread belief may lead to doctors being less willing to fill out DNAR order forms on patients, even in circumstances where they do not think attempted CPR would be appropriate, for fear of diminishing the overall level of care that the patient receives. The questionnaire also revealed that 82% of all doctors usually or almost always discussed resuscitation decisions with patients, while 62% usually or almost always discussed additional interventions. Despite this being a common practice, there is no scope on most current DNAR order forms (including the improved national one currently being piloted by the Resuscitation Council) for documentation of such decisions. Many of the comments on the physician questionnaires related to the desire to have “more options”. The study has several limitations. The questionnaire was not validated, and no retest questions were done. In asking the nursing staff the question “Do you think a DNAR alters?” we were trying to capture how they interpreted the form, and might act on it. They were not asked the question in two parts because they are not required to fill in DNAR orders; however, it is possible that there was variability in the interpretation of the question. The nursing staff may have felt pressured in replying to the questionnaires in the ward setting. Nevertheless, these data represent the beliefs of 85 members of staff at a particular point in time and is therefore relevant to the kind of day-to-day interpretations of DNAR orders that we are trying to capture. 5. New directions? One approach to the problems expanded upon above would be to educate health-care personnel better, both in their initial training and as an ongoing part of, for example, the adult life support (ALS) course that many go on. There is, however, little in the literature to support this; educational attempts to improve DNAR or end-oflife discussions have proven ineffectual,12 and no studies, to our knowledge, have yet been carried out assessing the impact of education programmes on health-care professionals’ understanding of DNAR. At the end of our questionnaire, in the “How would you improve the current form?” section several doctors and one nurse suggested that, instead of the DNAR order, a more detailed form would improve communication. Some such forms have been used elsewhere, and their impact on documentation,13,14 rate of CPR attempts and ventilation15 assessed. In addition, advanced care plans have been developed to improve end-of-life care for palliative patients (physician orders for life-sustaining treatment (POLST)16 ) or for those specifically at risk of deterioration (TEP,17 ). Following this questionnaire, we have developed a universal form of treatment options (UFTO), to address some of the issues raised, and are currently in the process of assessing this approach for its effect on both quantitative and qualitative outcomes. Given new working patterns, including multiple shift changes and less continuity of care, the problems of miscommunication surrounding DNAR orders are intensified.
1141
New approaches to deal with the issues surrounding DNAR orders are required, while maintaining the clarity of instruction that the current DNAR provides in the event of a patient arresting. A new approach should address the ad hoc nature of completing the form, the stigma associated with a DNAR order and the ambiguity about what do when a patient deteriorates that are clearly still prevalent with the DNAR order. With an appropriately evaluated new form and process, perhaps we can improve the care that those who are not for resuscitation receive. Conflict of interest The authors do not have any conflicts of interest. Acknowledgements We would like to thank the audit department at Addenbrooke’s Hospital and, in particular, Amara Siddique for her initial work with the data, Lorna Mason for her valuable administrative help and Angela Fritz for her tireless reviewing of the article and for the useful suggestions she made. In addition, we thank the NIHR Research for Patient Benefit Programme, who are supporting related work that three of the authors (ZF, JF and CP) are currently undertaking, although they have not had any direct role in this article. References 1. N. Office of Statistics Series DH1 (no 38, 2005). Table 17, pgs 56–7, http://www. statistics.gov.uk/StatBase/Product.asp?vlnk=620&Pos=1&ColRank=2&Rank=272 (accessed 4th May). 2. Aune S, Herlitz J, Bang A. Characteristics of patients who die in hospital with no attempt at resuscitation. Resuscitation 2005;65:291–9. 3. Decisions relating to cardiopulmonary resuscitation. A joint statement from the British Medical Association, the Resuscitation Council (UK) and the Royal College of Nursing; 2007. 4. Uhlmann RF, Cassel CK, McDonald WJ. Some treatment-withholding implications of no-code orders in an academic hospital. Crit Care Med 1984;12: 879–81. 5. La Puma J, Silverstein MD, Stocking CB, Roland D, Siegler M. Life-sustaining treatment. A prospective study of patients with DNR orders in a teaching hospital. Arch Intern Med 1988;148:2193–8. 6. Smith GB, Poplett N, Williams D. Staff awareness of a ‘Do Not Attempt Resuscitation’ policy in a District General Hospital. Resuscitation 2005;65: 159–63. 7. Beach MC, Morrison RS. The effect of do-not-resuscitate orders on physician decision-making. J Am Geriatr Soc 2002;50:2057–61. 8. Chen JL, Sosnov J, Lessard D, Goldberg RJ. Impact of do-not-resuscitation orders on quality of care performance measures in patients hospitalized with acute heart failure. Am Heart J 2008;156:78–84. 9. Lipton HL. Do-not-resuscitate decisions in a community hospital, incidence, implications, and outcomes. JAMA 1986;256:1164–9. 10. Shepardson LB, Youngner SJ, Speroff T, Rosenthal GE. Increased risk of death in patients with do-not-resuscitate orders. Med Care 1999;37:727–37. 11. Wenger NS, Pearson ML, Desmond KA, Brook RH, Kahn KL. Outcomes of patients with do-not-resuscitate orders. Toward an understanding of what donot-resuscitate orders mean and how they affect patients. Arch Intern Med 1995;155:2063–8. 12. Shorr AF, Niven AS, Katz DE, Parker JM, Eliasson AH. Regulatory and educational initiatives fail to promote discussions regarding end-of-life care. J Pain Symptom Manage 2000;19:168–73. 13. Mittelberger JA, Lo B, Martin D, Uhlmann RF. Impact of a procedure-specific do not resuscitate order form on documentation of do not resuscitate orders. Arch Intern Med 1993;153:228–32. 14. O’Toole EE, Youngner SJ, Juknialis BW, Daly B, Bartlett ET, Landefeld CS. Evaluation of a treatment limitation policy with a specific treatment-limiting order page. Arch Intern Med 1994;154:425–32. 15. Davila F. The impact of do-not-resuscitate and patient care category policies on CPR and ventilator support rates. Arch Intern Med 1996;156:405–8. 16. Tolle SW, Tilden VP, Nelson CA, Dunn PM. A prospective study of the efficacy of the physician order form for life-sustaining treatment. J Am Geriatr Soc 1998;46:1097–102. 17. Mercer M. Re: DNAR or AND? In: BMJ online [eletter]; 2009, http://www.bmj. com/cgi/eletters/338/apr30 1/b1723#213249 (accessed 4th May).