560007 research-article2014
PMJ0010.1177/0269216314560007Palliative MedicineHartog et al.
Original Article
End-of-life care in the intensive care unit: A patient-based questionnaire of intensive care unit staff perception and relatives’ psychological response
Palliative Medicine 1–10 © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269216314560007 pmj.sagepub.com
Christiane S Hartog1,2, Daniel Schwarzkopf1, Niels C Riedemann2, Ruediger Pfeifer3, Albrecht Guenther4, Kati Egerland2, Charles L Sprung5, Heike Hoyer6, Jochen Gensichen1,7 and Konrad Reinhart2
Abstract Background: Communication is a hallmark of end-of-life care in the intensive care unit. It may influence the impact of end-of-life care on patients’ relatives. We aimed to assess end-of-life care and communication from the perspective of intensive care unit staff and relate it to relatives’ psychological symptoms. Design: Prospective observational study based on consecutive patients with severe sepsis receiving end-of-life care; trial registration NCT01247792. Setting/participants: Four interdisciplinary intensive care units of a German University hospital. Responsible health personnel (attendings, residents and nurses) were questioned on the day of the first end-of-life decision (to withdraw or withhold life-supporting therapies) and after patients had died or were discharged. Relatives were interviewed by phone after 90 days. Results: Overall, 145 patients, 610 caregiver responses (92% response) and 84 relative interviews (70% response) were analysed. Most (86%) end-of-life decisions were initiated by attendings and only 2% by nurses; 41% of nurses did not know enough about endof-life decisions to communicate with relatives. Discomfort with end-of-life decisions was low. Relatives reported high satisfaction with decision-making and care, 87% thought their degree of involvement had been just right. However, 51%, 48% or 33% of relatives had symptoms of post-traumatic stress disorder, anxiety or depression, respectively. Predictors for depression and post-traumatic stress disorder were patient age and relatives’ gender. Relatives’ satisfaction with medical care and communication predicted less anxiety (p = 0.025). Conclusion: Communication should be improved within the intensive care unit caregiver team to strengthen the involvement of nurses in end-of-life care. Improved communication between caregivers and the family might lessen relatives’ long-term anxiety.
Keywords Palliative care, intensive care units, resuscitation orders, sepsis, attitude of health personnel
1Integrated
Research and Treatment Center, Center for Sepsis Control and Care (CSCC), Jena University Hospital, Jena, Germany 2Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Jena, Germany 3Department of Internal Medicine I, Jena University Hospital, Jena, Germany 4Department of Neurology, Jena University Hospital, Jena, Germany 5Department of Anesthesiology and Critical Care Medicine, Hadassah Hebrew University Medical Center, Jerusalem, Israel
6Institute
of Medical Statistics, Information Sciences and Documentation, Jena University Hospital, Jena, Germany 7Department of General Medicine, Jena University Hospital, Jena, Germany Corresponding author: Konrad Reinhart, Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Erlanger Allee 101, 07747 Jena, Germany. Email: [email protected]
2
Palliative Medicine
What is already known about the topic? •• Communication about end-of-life (EOL) care is a hallmark of quality EOL care in the intensive care unit (ICU). •• Communication takes place among the ICU caregiver team, that is, between nurses and physicians and between the team and the family. •• This study assessed the quality of communication in German ICUs from the perspectives of ICU staff and the family, based on consecutive patients with severe sepsis who received EOL care. What this paper adds? •• Our study identifies strengths and shortcomings of EOL communication among the interdisciplinary ICU team in German ICUs: distress with EOL decisions is very low, but nurses are rarely involved in EOL decision-making, and over 40% of nurses lacked sufficient knowledge about EOL decisions to communicate with relatives. •• The study also found a correlation between satisfaction of relatives with ICU care and communication and their long-term psychological outcomes. Relatives’ satisfaction with medical care and communication was a predictor for less anxiety after 90 days (p = 0.025). Implications for practice, theory or policy •• In the context of EOL care, communication and collaboration between ICU nurses and physicians need to be improved to strengthen the nurses’ involvement in this process. •• Improved communication between ICU caregivers and the family might lessen the relatives’ long-term anxiety.
Introduction In the intensive care unit (ICU), the shift from curative to palliative care results in many challenges.1,2 Effective communication and collaboration regarding end-of-life (EOL) issues between ICU nurses and physicians are the bases for providing high-quality EOL care to patients and their family members.3 However, interdisciplinary collaboration is challenged by differences in values, beliefs and professional roles,3,4 and true collaboration and open communication still elude ICU professionals, especially in the area of EOL care.3 Key issues of EOL decision-making include when to withhold or withdraw full life support, who should be involved in the decision-making process,5 sense of discomfort once EOL decision have been reached,6 elusive knowledge of patient preferences7 and poor communication between staff and patient surrogates.2 Decisionmaking and care in the ICU have an impact on the psychological stress which relatives of critically ill patients may experience, including anxiety, depression and posttraumatic stress disorder (PTSD).5,8,9 Psychological stress may be associated with being present at the time of death,10 being involved in EOL decision5 and needing to make decisions in discordance with the preferred decision-making role.10 On the other hand, improved EOL communication reduced psychological stress of family members.11 To identify areas where the practice of EOL care in our institution might be improved, we studied the process and impact of EOL communication and care from the perspective of ICU caregivers and relatives. We focused on patients with sepsis because we received government funding for sepsis research. In non-cardiological ICUs, sepsis is a major cause of death;12 moreover, sepsis is associated with an increased likelihood of EOL decisions.13,14 Because we
were not aware of patient-based questionnaires which assessed EOL decision-making from the perspective of ICU physicians, nurses and relatives, we developed questionnaires based on published recommendations and interviews with local critical care physicians and nurses. The trial is part of a larger effort to improve EOL care in our institution (trial registration NCT01247792).
Method Setting This prospective, observational study was conducted at the Jena University Hospital (JUH) in four ICUs with two mixed-surgical units with 24 and 26 beds, respectively, one neurological unit with 6 stroke and 6 neurological beds and one medical unit with 10 beds. Data were collected from December 2010 until February 2012. The JUH is a tertiary care facility with 1300 beds. Each ICU is a closed unit with an attending physician specializing in intensive care and several residents. The physician– patient ratio is approximately 1:8–14 on the mixed-surgical wards and 1:8–12 on the non-surgical wards. The nurse–patient ratio is 1:2–3. EOL decisions are usually initiated during interdisciplinary ward rounds and discussed with relatives considering patients’ written or presumed wishes. In our ICUs, the ‘shared decision-making’ model is used aiming to reach consensus with all stakeholders. The main reasons for limiting life support are poor prognosis, perceived non-beneficial treatment and patient preference. This study was conducted in accordance with the amended Declaration of Helsinki. The JUH
3
Hartog et al.
145 patients with severe sepsis and EOLD were included
121 main relatives could be reached after 90 days
136 patients rated by attendings 134 patients rated by nurses
84 main relatives participated
115 patients rated by all care-giver groups
124 patients rated by residents
Figure 1. Flow of participants.
ICU: intensive care unit. Caregivers of 145 patients with severe sepsis and end-of-life (EOL) decision, that is, attending ICU physicians, residents and nurses involved in the care of this patient, were each invited to participate on the day the first EOL decision was made and again on the day of discharge or death. The main relative, that is, the family member acting as health care proxy or the family member who was present during EOL talks was invited to participate after 90 days.
Ethics Committee approved the protocol and waived the need for written informed consent (No. 2964-11/10).
Study participants Flow of participants is depicted in Figure 1. All patients admitted to one of the four ICUs were screened daily for severe sepsis.15 Among these patients, those with an EOL decision were identified by attending ICU physicians (first day) and observed until death or discharge (last day). All caregivers, that is, attending physicians, residents and nurses caring for the patient, were questioned on the first and the last day by a study nurse. The last-day questionnaire was waived if less than 24 h passed between the first EOL decision and death or discharge. Weekend EOL decisions were documented within 48 h. Closest family members or relatives with the main decision-making role were identified from the patient file. Approximately 90 days after death or discharge of the patient, the study psychologist contacted the main relative by telephone. Relatives were excluded if no interview could be arranged despite 10 calls or within 1 month. If the main relative was unwilling to participate, another close relative who had had contact with ICU caregivers was asked to participate. Only one relative per patient was included. Relatives were informed about the aims of the interview and confidential handling of data before the interview.
Questionnaire development The staff questionnaire (Supplement) aimed to elicit the perspective of the team about EOL decision-making and
care. Items were based on published recommendations2,16 and modified by repeated discussions with medical and nursing leaders from all ICUs. Questionnaires were pilottested in a sample of 15 physicians and nurses from all ICUs and adapted accordingly. The final questionnaire addressed initiation and consensus with EOL decisions, timeliness and perceived burden of EOL decisions, advance directives, knowledge of patient preferences and communication with relatives. The questionnaires for relatives (Supplement) aimed to assess their perception of EOL decision-making and measured their emotional burden by the German Impact of Event Scale (IES)17 and the German Hospital Anxiety and Depression Scale (HADS).18 Both scales are considered reliable and validated tools in the ICU setting.19 The questionnaire was based on published research5,10,20–23 and modified by discussion with experienced ICU physicians. The interview was pre-tested with five relatives. Duration of the interviews varied between 30 and 90 min. To compare responders and non-responders, all relatives who were reached by phone were initially asked to answer four questions on overall satisfaction with care and information-giving and on present emotional suffering before proceeding to the main questionnaire.
Collected data EOL decisions were defined a priori: (1) DNR: do-notresuscitate order, that is, to not perform cardiopulmonary resuscitation (CPR) or intubation; (2) withhold: to not escalate ongoing or begin new life-sustaining treatment (LST); and (3) withdraw: to stop ongoing LST. Patient-related data
4
Palliative Medicine
included demographics, medical or surgical admission type, co-treatment by surgeon, disease severity (Simplified Acute Physiology Score–II (SAPS-II)24 and Sepsis-related Organ Failure Assessment (SOFA) scores25), LST, ICU and hospital length of stay (LOS) and ICU and hospital mortality. Ninety-day mortality of patients without EOL decisions was assessed from a web-based sepsis registry.16
Data analysis Descriptive statistics and Fisher’s exact or Chi-square test for nominal data and Wilcoxon rank-sum test for continuous data were used. Comparisons across time used paired tests (McNemar test for dichotomous and paired Wilcoxon rank-sum test for continuous variables). Caregiver ratings were compared using paired tests if data were available for one patient from all three caregiver groups. We used previously reported cut-offs (>7 for HADS subscales; >32 for German IES)11 to identify relatives at risk for PTSD, anxiety disorder and depression. Six self-developed items on satisfaction with care and communication showed high bivariate correlations and exploratory factor analysis revealed a single common factor. They were therefore summed up as ‘Satisfaction with medical care and communication in general’ (Supplemental Table 2). Three multiple linear regression models were calculated using the statistical software R26 to predict anxiety, depression and post-traumatic stress each. The set of predictors was chosen based on previous research.5,10,20–23 Only those predictors that showed a p-value ⩽ 0.1 by simple regression analysis were included in the multiple regression models.
Qualitative analysis of caregiver burden Caregivers rated the burden of EOL care for a particular patient on a Likert scale from 1 (no strain) to 10 (maximal strain). If burden was rated ⩾4, caregivers were asked to give reasons. The answers were analysed qualitatively by thematic analysis.27 Codes were created and organized into potential themes by psychologist (D.S.), reviewed by an experienced physician (C.S.H.) and discussed by both until consensus was achieved. Occurrence of themes was counted and compared between the three caregiver groups.
Figure 2. End-of-life (EOL) decisions. (a) EOL decisions were defined as withdrawal of life-sustaining treatment (LST), withholding LST or do-not-resuscitate-order (DNR). Presented here are the proportions of patients according to their most active treatment limitation on the day of the first EOL decision (‘first day’) and the day of death or discharge (‘last day’). On the last day, the most active treatment limitation was a withdrawal decision in 70% of patients, a withholding decision in 21% and receiving only a DNR in 9% of patients. (b) The cumulative probability of death according to type of treatment decision.
Results Patients and outcomes Of 5162 admitted ICU patients, 423 (8%) were diagnosed with severe sepsis. An EOL decision was made for 145 of 423 patients (34%) and preceded 65% (118 of 181) of all deaths from sepsis in the ICU. In 70% of patients, the most active treatment decision was withdrawal, in 21% of patients withholding and 9% of patients had only a DNR (Figure 2(a)). Characteristics of patients, treatments and
outcomes are shown in Supplemental Table 3. Patients with EOL decision had a median ICU stays of 241 h. Median 144.67 h (interquartile range (IQR): 36.0–331.5 h) passed from admission to first EOL decision. Death occurred in median 21.4 h (5.1–96.0 h) after the first EOL decision. Ninety-day mortality of patients with EOL decision was 92%. Survival time according to the most active treatment limitation can be seen in Figure 2(b). Co-treatment by a surgeon occurred in 80.7% (117/145) patients.
5
Hartog et al.
ICU caregiver perspectives Of 663 questionnaires, 610 were answered (92% response). In detail, 394 of 435 first-day questionnaires were returned by ICU staff members (136 attending physician questionnaires, 124 resident questionnaires and 134 nurse questionnaires, 91% response) and 216 of 228 last-day questionnaires were answered (75 attending physician questionnaires, 69 resident questionnaires and 72 nurse questionnaires; 95% response). Missing responses result from not reaching caregivers on account of working shifts, leave of absence or too high workload. Overall, 18 attending physicians participated and in median answered questionnaires for 11 patients (IQR: 2–17) each over the whole study period. Participating residents (N = 66) answered 2 (1, 4) (median (IQR)) questionnaires each and participating nurses (N = 128) answered 1 (1, 2) questionnaires each. Considering first- and last-day questionnaires, one patient received ratings by, in median, 3 (3, 5) different caregivers. EOL decisions were initiated most frequently by attending ICU physicians (86%) and least often by relatives (14%) or nurses (2%). Discomfort with EOL decision was notably low among all groups (median 0 on a scale from 0 to 10). On the day the EOL decisions were made, patient wishes were known to attendings in 55% (73 of 132 patients). If patient wishes were known, no one perceived that they were not followed. Table 1 compares differing perspectives of involved caregiver groups. Nurses had less knowledge than physicians about EOL decision-making, for example, about initiation of decisions, understanding and consensus with relatives, patient wishes. Nurses also more often denied they had sufficient knowledge to talk with family. On the other hand, nurses more often than physicians rated symptom control to be insufficient and more often rated EOL care to be a burden (ratings ⩾ 4 on a 0–10 scale; p ⩽ 0.002). The perceived burden of EOL care was overall low among all groups (median 2–4). Table 2 shows the results of the qualitative analysis of reasons given by those 124 staff members (31% of total) who experienced an increased burden. The most frequent reasons were complex care requirements for 50% of nurses, and the unpredictable course of the disease for 31% attendings and 39% residents. In a subgroup of 30 patients (23%), attendings judged that EOL decisions should have been made earlier. Patients with delayed EOL decision were older than patients with non-delayed EOL decision (76.7 vs 70.9 years; p = 0.027); the interval between onset of sepsis-related organ dysfunction and the first EOL decision was not significantly different (median 121.5 (27.4–328.6 h) vs 109 h (28.8–301.7 h); p = 0.708).
Relatives’ psychological burden and perspectives on ICU care and decision-making After 90 days, relatives of 121 patients could be contacted. All contacted relatives answered initial questions about
overall satisfaction with ICU care, physician or nurse communication and own acute emotional suffering. Answers did not differ between those relatives who consented to participate in the full interview (84/121, 70%) and those who declined to participate (each of these items p ⩾ 0.15). All participants except three had suffered the death of their relative. Characteristics and responses of relatives are given in Table 3 and in Supplemental Table 1. Most relatives were female (74%), with a median age of 57 years. Most relatives (77%) were also acting as legal proxies. After 90 days, 41 relatives (51%) experienced symptoms of PTSD; 40 (48%) and 27 (33%) suffered from symptoms of anxiety and depression, respectively. Relatives felt very satisfied with medical care and communication (median 9 on a scale of 0–10), fully agreed with EOL decision and fully believed the patients’ will had been considered (both median 10). Still, 30% and 13%, respectively, would have considered talking with a psychologist or clergy helpful. In all, 43% of relatives believed the responsibility for EOL decisions was shared, 37% accorded sole responsibility to themselves and 20% to the physicians; 87% of relatives thought their degree of involvement had been just right and 12% had wanted less involvement. Almost all (98%) felt the decision had been timely. Only 9% perceived conflicts about EOL care.
Risk factors for relatives’ psychological burden In multiple regressions (Table 4), spouses showed higher values for depression and PTSD than children and other relatives (p 32) HADS Anxiety HADS Depression HADS Total Anxiety symptoms (HADS Anxiety > 7) Depression symptoms (HADS Depression > 7)
57 (50–66) 62 (73.8) 29 (34.5) 35 (41.7) 35 (41.7) 2 (2.4) 3 (3.6) 9 (10.7) 65 (77.4) 25 (31.6) 6 (9) 19 (30.2) 8 (13.1) 9 (7.92–9.67)
33 (20–49) 41 (50.6) 7 (5–10) 6 (2.25–9) 13 (8–18.75) 40 (48.2) 27 (32.9)
EOL: end-of-life; IES Impact of Event Scale; HADS: Hospital Anxiety and Depression Scale; IQR: interquartile range. N (%), Median (IQR): scale from 0 to 10 (0 = very little; 10 = very much). aItems only answered by relatives taking part in the longer version of the interview (N ⩽ 67). bSum score, including satisfaction with medical care, satisfaction with physician talks, satisfaction with information, having been given enough time, comprehensible information, completeness of information (see exploratory factor analysis in Supplemental Table 2).
Senior consultants initiated most EOL decisions and knew most about the decision-making process. Nurses initiated only 2% of EOL decision, similar to findings from a European survey.28 Fewer nurses than physicians knew patients’ wishes; over 40% of nurses lacked sufficient knowledge to talk with relatives and more nurses felt burdened by EOL care. European surveys repeatedly described a discrepancy between nurses’ preference and actual involvement in EOL decision-making.29–31 International researchers point out that ICU nurses could provide an essential role in the provision of effective EOL care, especially in the emotional support of patient and family at the EOL.32–34 However, nurses also need more encouragement, education and opportunities to develop the necessary skills for palliative care in the ICU.33
Half of the relatives suffered symptoms of PTSD and anxiety and one-third suffered from symptoms of depression after 90 days. These rates are higher than those found in family members of general ICU patients in France and the United States.5,35 However, 92% of our patients died, which may explain their spouses’ or children’s increased emotional symptoms. Azoulay et al.5 also reported that half the family members of patients who died in the ICU showed symptoms of PTSD after 90 days. Consistent with previous reports, we found that younger patient age and female gender of the relative were also associated with increased psychological strain.5,10,20 On the other hand, we found that satisfaction with medical care and communication was related to lower levels of anxiety and depression. It has been shown that the manner in which caregivers communicate with relatives can influence their psychological stress. For example, receiving incomplete information,36 not finding the patient’s physician comforting37 or discrepancy with the preferred decision-making role10 all increased stress. However, relatives in our study were highly satisfied with the manner of communication by medical staff. Only 9% perceived conflicting opinions, less than previously reported.38 Still, up to 30% of relatives would have liked to talk with a psychologist; unfortunately, a psychologist is not routinely available in our ICU. The potential association of relatives’ anxiety and depression with EOL decisions that were delayed is interesting but should be interpreted with care. The uncertainty estimates, that is, the standard error, might be underestimated because of non-independence in attendings’ ratings. Also, it was based on the subjective judgment by attending physicians that such decision could have been reached earlier, maybe due to uncertainty about patient wishes, reluctance by the family or by the co-treating surgeon to forgo LST.39 These patients were considerably older with a median age of about 77 years. The reduced strain for the family might be explained by the advanced age of the patient,20 or by having more time to come to terms with the inevitable outcome. This finding seemingly disagrees with the traditional view that it may be desirable to reduce the duration of ICU treatment in patients with a poor prognosis.40 However, it needs to be interpreted with caution and confirmed by suitable clinical investigation. This study has limitations and strengths. It is a singlehospital survey; therefore, findings may not be generalizable. Our ICUs are closed units; the ICU structure (closed versus open) may influence decision-making.41 We did not assess other possible influences on work management in the ICU, such as post graduate competency training or the influence of nurse–patient ratio.42,43 It was an exploratory study which limits the conclusions that can be drawn from the findings. Some of the questionnaire items were self-developed and not validated. However, questions reflected quality concerns of experienced caregivers and pilot-testing ensured their face validity. Strengths include
8
Palliative Medicine
Table 4. Risk factors for relatives’ psychological burden. Relatives’ psychological burden at 90 days
HADS anxiety
HADS depression
IES
Simple regression coefficients
Simple regression coefficients
Multiple regression coefficients (95% CI)
Simple regression coefficients
Multiple regression coefficients (95% CI)
−0.19′
−0.14 (−0.38 to 0.1)
−0.13
0.05 0.14 0 0.03 0.19
−0.16
−0.28*
log(patient age)
Multiple regression coefficients (95% CI)
0
SAPS-II at admission Patient died in ICU log(ICU LOS) Withdrawal of therapy Patient received surgical cotreatment Attending’s opinion that endof-life decision was delayed Relative, male
0.08 −0.11 0.02 0 0.3*
Relative, higher education Relative, child
−0.05 −0.15
−0.16 −0.52***
Relative, other relationship
−0.05
−0.28*
Relative perception that information was contradictory Relative satisfaction with medical care and communication in generala
0.08
0.06
−0.02
−0.24* (−0.45 to −0.03) −0.25* (−0.48 to −0.01) −0.19′ (−0.42 to 0.03)
−0.18
−0.11
−0.38** −0.21′
−0.25*
0.22′ (−0.04 to 0.48) −0.37** (−0.6 to −0.14) −0.2′ (−0.43 to 0.03)
−0.26* (−0.49 to −0.03)
0.04 0.04 0.1 0.02 0.09 −0.28*
−0.27* (−0.5 to −0.04)
−0.14 −0.48*** (−0.72 to −0.24) −0.33** (−0.54 to −0.11)
−0.12 −0.29* −0.21′
HADS: Hospital Anxiety and Depression Scale; IES: Impact of Event Scale; CI: confidence interval; SAPS-II: Simplified Acute Physiology Score–II; ICU: intensive care unit; LOS: length of stay. All coefficients are standardized. Predictors significant at level 0.1 in simple regression were considered for multiple regression. aSum score, including satisfaction with medical care, satisfaction with physician talks, satisfaction with information, having been given enough time, comprehensible information, completeness of information (see exploratory factor analysis in Supplemental Table 2). ′p ⩽ 0.1; *p ⩽ 0.05; **p ⩽ 0.01; ***p ⩽ 0.001.
the prospective design with patient-based surveys, a priori defined categories, identification of EOL decisions based on daily caregiver assessments not on documentation, surveying physicians and nurses and comparing their assessments, direct reporting of perceptions about each EOL decision rather than theoretical responses about EOL decision in general, the disease-specific focus whereby some of the variability reported in observations of EOL care is avoided, the relatively large sample of relatives compared to previous interview studies, linking patientrelated data with caregivers’ and relatives’ perspectives, performing a 90-day follow-up of relatives with validated survey tools that have been used previously5,11 and the combination of quantitative and qualitative data to gain more comprehensive understanding.
Conclusion Our study identifies some shortcomings of EOL communication and links satisfaction of relatives with ICU care and communication with long-term psychological outcomes.
In order to improve EOL care in our ICUs, communication should be improved within the ICU team to strengthen the involvement of nurses in EOL care. Improved communication between caregivers and the family might lessen the relatives’ long-term anxiety. Acknowledgements We thank study nurse Anja Haucke for her dedicated work. Note of prior abstract publication/presentation: Data were partly presented as abstract at the European Society of Intensive Care Medicine (ESICM) 2012 meeting, October 13–17, Lisbon.
Declaration of conflicting interests The authors have reported that no potential conflicts of interest exist. The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of this article.
Ethical approval Approval for this study was granted by the Ethics Committee of the Jena University Hospital (No. 2964-11/10).
Hartog et al. Funding The study was funded partially by the German Federal Ministry of Ministry of Education and Research (BMBF), Germany, FKZ: 01EO1002. Daniel Schwarzkopf is funded in full by the Center for Sepsis Control and Care (CSCC). The CSCC is funded by the German Federal Ministry of Ministry of Education and Research (BMBF). Konrad Reinhart, Christiane S. Hartog and Jochen Gensichen receive CSCC research grants.
References 1. Prendergast TJ and Puntillo KA. Withdrawal of life support: intensive caring at the end of life. JAMA 2002; 288: 2732–2740. 2. Carlet J, Thijs LG, Antonelli M, et al. Challenges in endof-life care in the ICU. Statement of the 5th International Consensus Conference in Critical Care: Brussels, Belgium, April 2003. Intensive Care Med 2004; 30: 770–784. 3. Puntillo KA and McAdam JL. Communication between physicians and nurses as a target for improving end-of-life care in the intensive care unit: challenges and opportunities for moving forward. Crit Care Med 2006; 34: S332– S340. 4. Jensen HI, Ammentorp J, Johannessen H, et al. Challenges in end-of-life decisions in the intensive care unit: an ethical perspective. J Bioeth Inq 2013; 10: 93–101. 5. Azoulay E, Pochard F, Kentish-Barnes N, et al. Risk of post-traumatic stress symptoms in family members of intensive care unit patients. Am J Respir Crit Care Med 2005; 171: 987–994. 6. Griffith L, Cook D, Hanna S, et al. Clinician discomfort with life support plans for mechanically ventilated patients. Intensive Care Med 2004; 30: 1783–1790. 7. A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks of treatments (SUPPORT). The SUPPORT Principal Investigators. JAMA 1995; 274: 1591–1598. 8. Davidson JE, Jones C and Bienvenu OJ. Family response to critical illness: postintensive care syndrome-family. Crit Care Med 2012; 40: 618–624. 9. McAdam JL, Dracup KA, White DB, et al. Symptom experiences of family members of intensive care unit patients at high risk for dying. Crit Care Med 2010; 38: 1078–1085. 10. Gries CJ, Engelberg RA, Kross EK, et al. Predictors of symptoms of posttraumatic stress and depression in family members after patient death in the ICU. Chest 2010; 137: 280–287. 11. Lautrette A, Darmon M, Megarbane B, et al. A communication strategy and brochure for relatives of patients dying in the ICU. N Engl J Med 2007; 356: 469–478. 12. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29: 1303–1310. 13. Sprung CL, Cohen SL, Sjokvist P, et al. End-of-life practices in European intensive care units: the Ethicus Study. JAMA 2003; 290: 790–797. 14. Meissner A, Genga KR, Studart FS, et al. Epidemiology of and factors associated with end-of-life decisions in
9 a surgical intensive care unit. Crit Care Med 2010; 38: 1060–1068. 15. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 1992; 20: 864–874. 16. Schmidt K, Gensichen J, Thiel P, et al. A prospective observational web-based registry for assessment of long-term outcome in patients with severe sepsis. Infection 2013; 41(Suppl. 1): S84–S85. 17. Ferring D and Filipp S-H. Teststatistische Überprüfung der Impact of Event-Skala: Befunde zu Reliabilität und Stabilität. Diagnostica 1994; 40: 344–462. 18. Herrmann C and Buss U. Vorstellung und Validierung einer deutschen Version der ‘Hospital Anxiety and Depression Scale’ (HAD-Skala): Ein Fragebogen zur Erfassung des psychischen Befindens bei Patienten mit körperlichen Beschwerden. Diagnostica 1994; 40: 143–154. 19. Kentish-Barnes N, Lemiale V, Chaize M, et al. Assessing burden in families of critical care patients. Crit Care Med 2009; 37: S448–S456. 20. Kross EK, Engelberg RA, Gries CJ, et al. ICU care associated with symptoms of depression and posttraumatic stress disorder among family members of patients who die in the ICU. Chest 2011; 139: 795–801. 21. Pillai L, Aigalikar S, Vishwasrao SM, et al. Can we predict intensive care relatives at risk for posttraumatic stress disorder? Indian J Crit Care Med 2010; 14: 83–87. 22. Wall RJ, Curtis JR, Cooke CR, et al. Family satisfaction in the ICU: differences between families of survivors and nonsurvivors. Chest 2007; 132: 1425–1433. 23. Gries CJ, Curtis JR, Wall RJ, et al. Family member satisfaction with end-of-life decision making in the ICU. Chest 2008; 133: 704–712. 24. Le Gall JR, Lemeshow S and Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/ North American multicenter study. JAMA 1993; 270: 2957– 2963. 25. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsisrelated Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996; 22: 707–710. 26. Huang R. RQDA: R-based qualitative data analysis (R package version 0.2-2). http://rqda.r-forge.r-project.org/, 2011. 27. Braun V and Clarke V. Using thematic analysis in psychology. Qual Res Psychol 2006; 3: 77–101. 28. Benbenishty J, Ganz FD, Lippert A, et al. Nurse involvement in end-of-life decision making: the ETHICUS Study. Intensive Care Med 2006; 32: 129–132. 29. Ferrand E, Lemaire F, Regnier B, et al. Discrepancies between perceptions by physicians and nursing staff of intensive care unit end-of-life decisions. Am J Respir Crit Care Med 2003; 167: 1310–1315. 30. Jox RJ, Krebs M, Fegg M, et al. Limiting life-sustaining treatment in German intensive care units: a multiprofessional survey. J Crit Care 2010; 25: 413–419. 31. Jensen HI, Ammentorp J, Erlandsen M, et al. Withholding or withdrawing therapy in intensive care units: an analysis
10
32.
33.
34.
35.
36.
37.
Palliative Medicine of collaboration among healthcare professionals. Intensive Care Med 2011; 37: 1696–1705. Nelson JE, Cortez TB, Curtis JR, et al. Integrating palliative care in the ICU: the nurse in a leading role. J Hosp Palliat Nurs 2011; 13: 89–94. Efstathiou N and Clifford C. The critical care nurse’s role in end-of-life care: issues and challenges. Nurs Crit Care 2011; 16: 116–123. Ranse K, Yates P and Coyer F. End-of-life care in the intensive care setting: a descriptive exploratory qualitative study of nurses’ beliefs and practices. Aust Crit Care 2012; 25: 4–12. Anderson WG, Arnold RM, Angus DC, et al. Posttraumatic stress and complicated grief in family members of patients in the intensive care unit. J Gen Intern Med 2008; 23: 1871–1876. Azoulay E, Adrie C, De Lassence A, et al. Determinants of postintensive care unit mortality: a prospective multicenter study. Crit Care Med 2003; 31: 428–432. Siegel MD, Hayes E, Vanderwerker LC, et al. Psychiatric illness in the next of kin of patients who die in the intensive care unit. Crit Care Med 2008; 36: 1722–1728.
38. Abbott KH, Sago JG, Breen CM, et al. Families looking back: one year after discussion of withdrawal or withholding of life-sustaining support. Crit Care Med 2001; 29: 197–201. 39. Buchman TG. Surgeons and their patients near the end of life. Crit Care Med 2010; 38: 995–996. 40. Schneiderman LJ, Gilmer T, Teetzel HD, et al. Effect of ethics consultations on nonbeneficial life-sustaining treatments in the intensive care setting: a randomized controlled trial. JAMA 2003; 290: 1166–1172. 41. Cassell J, Buchman TG, Streat S, et al. Surgeons, intensivists, and the covenant of care: administrative models and values affecting care at the end of life – updated. Crit Care Med 2003; 31: 1551–1557; discussion 7–9. 42. Miranda DR, Rivera-Fernandez R and Nap RE. Critical care medicine in the hospital: lessons from the EURICUSstudies. Med Intensiva 2007; 31: 194–203. 43. Shortell SM, Zimmerman JE, Rousseau DM, et al. The performance of intensive care units: does good management make a difference? Med Care 1994; 32: 508–525.
PMJ0010.1177/0269216314560007Palliative MedicineHartog et al.
Original Article
End-of-life care in the intensive care unit: A patient-based questionnaire of intensive care unit staff perception and relatives’ psychological response
Palliative Medicine 1–10 © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269216314560007 pmj.sagepub.com
Christiane S Hartog1,2, Daniel Schwarzkopf1, Niels C Riedemann2, Ruediger Pfeifer3, Albrecht Guenther4, Kati Egerland2, Charles L Sprung5, Heike Hoyer6, Jochen Gensichen1,7 and Konrad Reinhart2
Abstract Background: Communication is a hallmark of end-of-life care in the intensive care unit. It may influence the impact of end-of-life care on patients’ relatives. We aimed to assess end-of-life care and communication from the perspective of intensive care unit staff and relate it to relatives’ psychological symptoms. Design: Prospective observational study based on consecutive patients with severe sepsis receiving end-of-life care; trial registration NCT01247792. Setting/participants: Four interdisciplinary intensive care units of a German University hospital. Responsible health personnel (attendings, residents and nurses) were questioned on the day of the first end-of-life decision (to withdraw or withhold life-supporting therapies) and after patients had died or were discharged. Relatives were interviewed by phone after 90 days. Results: Overall, 145 patients, 610 caregiver responses (92% response) and 84 relative interviews (70% response) were analysed. Most (86%) end-of-life decisions were initiated by attendings and only 2% by nurses; 41% of nurses did not know enough about endof-life decisions to communicate with relatives. Discomfort with end-of-life decisions was low. Relatives reported high satisfaction with decision-making and care, 87% thought their degree of involvement had been just right. However, 51%, 48% or 33% of relatives had symptoms of post-traumatic stress disorder, anxiety or depression, respectively. Predictors for depression and post-traumatic stress disorder were patient age and relatives’ gender. Relatives’ satisfaction with medical care and communication predicted less anxiety (p = 0.025). Conclusion: Communication should be improved within the intensive care unit caregiver team to strengthen the involvement of nurses in end-of-life care. Improved communication between caregivers and the family might lessen relatives’ long-term anxiety.
Keywords Palliative care, intensive care units, resuscitation orders, sepsis, attitude of health personnel
1Integrated
Research and Treatment Center, Center for Sepsis Control and Care (CSCC), Jena University Hospital, Jena, Germany 2Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Jena, Germany 3Department of Internal Medicine I, Jena University Hospital, Jena, Germany 4Department of Neurology, Jena University Hospital, Jena, Germany 5Department of Anesthesiology and Critical Care Medicine, Hadassah Hebrew University Medical Center, Jerusalem, Israel
6Institute
of Medical Statistics, Information Sciences and Documentation, Jena University Hospital, Jena, Germany 7Department of General Medicine, Jena University Hospital, Jena, Germany Corresponding author: Konrad Reinhart, Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Erlanger Allee 101, 07747 Jena, Germany. Email: [email protected]
2
Palliative Medicine
What is already known about the topic? •• Communication about end-of-life (EOL) care is a hallmark of quality EOL care in the intensive care unit (ICU). •• Communication takes place among the ICU caregiver team, that is, between nurses and physicians and between the team and the family. •• This study assessed the quality of communication in German ICUs from the perspectives of ICU staff and the family, based on consecutive patients with severe sepsis who received EOL care. What this paper adds? •• Our study identifies strengths and shortcomings of EOL communication among the interdisciplinary ICU team in German ICUs: distress with EOL decisions is very low, but nurses are rarely involved in EOL decision-making, and over 40% of nurses lacked sufficient knowledge about EOL decisions to communicate with relatives. •• The study also found a correlation between satisfaction of relatives with ICU care and communication and their long-term psychological outcomes. Relatives’ satisfaction with medical care and communication was a predictor for less anxiety after 90 days (p = 0.025). Implications for practice, theory or policy •• In the context of EOL care, communication and collaboration between ICU nurses and physicians need to be improved to strengthen the nurses’ involvement in this process. •• Improved communication between ICU caregivers and the family might lessen the relatives’ long-term anxiety.
Introduction In the intensive care unit (ICU), the shift from curative to palliative care results in many challenges.1,2 Effective communication and collaboration regarding end-of-life (EOL) issues between ICU nurses and physicians are the bases for providing high-quality EOL care to patients and their family members.3 However, interdisciplinary collaboration is challenged by differences in values, beliefs and professional roles,3,4 and true collaboration and open communication still elude ICU professionals, especially in the area of EOL care.3 Key issues of EOL decision-making include when to withhold or withdraw full life support, who should be involved in the decision-making process,5 sense of discomfort once EOL decision have been reached,6 elusive knowledge of patient preferences7 and poor communication between staff and patient surrogates.2 Decisionmaking and care in the ICU have an impact on the psychological stress which relatives of critically ill patients may experience, including anxiety, depression and posttraumatic stress disorder (PTSD).5,8,9 Psychological stress may be associated with being present at the time of death,10 being involved in EOL decision5 and needing to make decisions in discordance with the preferred decision-making role.10 On the other hand, improved EOL communication reduced psychological stress of family members.11 To identify areas where the practice of EOL care in our institution might be improved, we studied the process and impact of EOL communication and care from the perspective of ICU caregivers and relatives. We focused on patients with sepsis because we received government funding for sepsis research. In non-cardiological ICUs, sepsis is a major cause of death;12 moreover, sepsis is associated with an increased likelihood of EOL decisions.13,14 Because we
were not aware of patient-based questionnaires which assessed EOL decision-making from the perspective of ICU physicians, nurses and relatives, we developed questionnaires based on published recommendations and interviews with local critical care physicians and nurses. The trial is part of a larger effort to improve EOL care in our institution (trial registration NCT01247792).
Method Setting This prospective, observational study was conducted at the Jena University Hospital (JUH) in four ICUs with two mixed-surgical units with 24 and 26 beds, respectively, one neurological unit with 6 stroke and 6 neurological beds and one medical unit with 10 beds. Data were collected from December 2010 until February 2012. The JUH is a tertiary care facility with 1300 beds. Each ICU is a closed unit with an attending physician specializing in intensive care and several residents. The physician– patient ratio is approximately 1:8–14 on the mixed-surgical wards and 1:8–12 on the non-surgical wards. The nurse–patient ratio is 1:2–3. EOL decisions are usually initiated during interdisciplinary ward rounds and discussed with relatives considering patients’ written or presumed wishes. In our ICUs, the ‘shared decision-making’ model is used aiming to reach consensus with all stakeholders. The main reasons for limiting life support are poor prognosis, perceived non-beneficial treatment and patient preference. This study was conducted in accordance with the amended Declaration of Helsinki. The JUH
3
Hartog et al.
145 patients with severe sepsis and EOLD were included
121 main relatives could be reached after 90 days
136 patients rated by attendings 134 patients rated by nurses
84 main relatives participated
115 patients rated by all care-giver groups
124 patients rated by residents
Figure 1. Flow of participants.
ICU: intensive care unit. Caregivers of 145 patients with severe sepsis and end-of-life (EOL) decision, that is, attending ICU physicians, residents and nurses involved in the care of this patient, were each invited to participate on the day the first EOL decision was made and again on the day of discharge or death. The main relative, that is, the family member acting as health care proxy or the family member who was present during EOL talks was invited to participate after 90 days.
Ethics Committee approved the protocol and waived the need for written informed consent (No. 2964-11/10).
Study participants Flow of participants is depicted in Figure 1. All patients admitted to one of the four ICUs were screened daily for severe sepsis.15 Among these patients, those with an EOL decision were identified by attending ICU physicians (first day) and observed until death or discharge (last day). All caregivers, that is, attending physicians, residents and nurses caring for the patient, were questioned on the first and the last day by a study nurse. The last-day questionnaire was waived if less than 24 h passed between the first EOL decision and death or discharge. Weekend EOL decisions were documented within 48 h. Closest family members or relatives with the main decision-making role were identified from the patient file. Approximately 90 days after death or discharge of the patient, the study psychologist contacted the main relative by telephone. Relatives were excluded if no interview could be arranged despite 10 calls or within 1 month. If the main relative was unwilling to participate, another close relative who had had contact with ICU caregivers was asked to participate. Only one relative per patient was included. Relatives were informed about the aims of the interview and confidential handling of data before the interview.
Questionnaire development The staff questionnaire (Supplement) aimed to elicit the perspective of the team about EOL decision-making and
care. Items were based on published recommendations2,16 and modified by repeated discussions with medical and nursing leaders from all ICUs. Questionnaires were pilottested in a sample of 15 physicians and nurses from all ICUs and adapted accordingly. The final questionnaire addressed initiation and consensus with EOL decisions, timeliness and perceived burden of EOL decisions, advance directives, knowledge of patient preferences and communication with relatives. The questionnaires for relatives (Supplement) aimed to assess their perception of EOL decision-making and measured their emotional burden by the German Impact of Event Scale (IES)17 and the German Hospital Anxiety and Depression Scale (HADS).18 Both scales are considered reliable and validated tools in the ICU setting.19 The questionnaire was based on published research5,10,20–23 and modified by discussion with experienced ICU physicians. The interview was pre-tested with five relatives. Duration of the interviews varied between 30 and 90 min. To compare responders and non-responders, all relatives who were reached by phone were initially asked to answer four questions on overall satisfaction with care and information-giving and on present emotional suffering before proceeding to the main questionnaire.
Collected data EOL decisions were defined a priori: (1) DNR: do-notresuscitate order, that is, to not perform cardiopulmonary resuscitation (CPR) or intubation; (2) withhold: to not escalate ongoing or begin new life-sustaining treatment (LST); and (3) withdraw: to stop ongoing LST. Patient-related data
4
Palliative Medicine
included demographics, medical or surgical admission type, co-treatment by surgeon, disease severity (Simplified Acute Physiology Score–II (SAPS-II)24 and Sepsis-related Organ Failure Assessment (SOFA) scores25), LST, ICU and hospital length of stay (LOS) and ICU and hospital mortality. Ninety-day mortality of patients without EOL decisions was assessed from a web-based sepsis registry.16
Data analysis Descriptive statistics and Fisher’s exact or Chi-square test for nominal data and Wilcoxon rank-sum test for continuous data were used. Comparisons across time used paired tests (McNemar test for dichotomous and paired Wilcoxon rank-sum test for continuous variables). Caregiver ratings were compared using paired tests if data were available for one patient from all three caregiver groups. We used previously reported cut-offs (>7 for HADS subscales; >32 for German IES)11 to identify relatives at risk for PTSD, anxiety disorder and depression. Six self-developed items on satisfaction with care and communication showed high bivariate correlations and exploratory factor analysis revealed a single common factor. They were therefore summed up as ‘Satisfaction with medical care and communication in general’ (Supplemental Table 2). Three multiple linear regression models were calculated using the statistical software R26 to predict anxiety, depression and post-traumatic stress each. The set of predictors was chosen based on previous research.5,10,20–23 Only those predictors that showed a p-value ⩽ 0.1 by simple regression analysis were included in the multiple regression models.
Qualitative analysis of caregiver burden Caregivers rated the burden of EOL care for a particular patient on a Likert scale from 1 (no strain) to 10 (maximal strain). If burden was rated ⩾4, caregivers were asked to give reasons. The answers were analysed qualitatively by thematic analysis.27 Codes were created and organized into potential themes by psychologist (D.S.), reviewed by an experienced physician (C.S.H.) and discussed by both until consensus was achieved. Occurrence of themes was counted and compared between the three caregiver groups.
Figure 2. End-of-life (EOL) decisions. (a) EOL decisions were defined as withdrawal of life-sustaining treatment (LST), withholding LST or do-not-resuscitate-order (DNR). Presented here are the proportions of patients according to their most active treatment limitation on the day of the first EOL decision (‘first day’) and the day of death or discharge (‘last day’). On the last day, the most active treatment limitation was a withdrawal decision in 70% of patients, a withholding decision in 21% and receiving only a DNR in 9% of patients. (b) The cumulative probability of death according to type of treatment decision.
Results Patients and outcomes Of 5162 admitted ICU patients, 423 (8%) were diagnosed with severe sepsis. An EOL decision was made for 145 of 423 patients (34%) and preceded 65% (118 of 181) of all deaths from sepsis in the ICU. In 70% of patients, the most active treatment decision was withdrawal, in 21% of patients withholding and 9% of patients had only a DNR (Figure 2(a)). Characteristics of patients, treatments and
outcomes are shown in Supplemental Table 3. Patients with EOL decision had a median ICU stays of 241 h. Median 144.67 h (interquartile range (IQR): 36.0–331.5 h) passed from admission to first EOL decision. Death occurred in median 21.4 h (5.1–96.0 h) after the first EOL decision. Ninety-day mortality of patients with EOL decision was 92%. Survival time according to the most active treatment limitation can be seen in Figure 2(b). Co-treatment by a surgeon occurred in 80.7% (117/145) patients.
5
Hartog et al.
ICU caregiver perspectives Of 663 questionnaires, 610 were answered (92% response). In detail, 394 of 435 first-day questionnaires were returned by ICU staff members (136 attending physician questionnaires, 124 resident questionnaires and 134 nurse questionnaires, 91% response) and 216 of 228 last-day questionnaires were answered (75 attending physician questionnaires, 69 resident questionnaires and 72 nurse questionnaires; 95% response). Missing responses result from not reaching caregivers on account of working shifts, leave of absence or too high workload. Overall, 18 attending physicians participated and in median answered questionnaires for 11 patients (IQR: 2–17) each over the whole study period. Participating residents (N = 66) answered 2 (1, 4) (median (IQR)) questionnaires each and participating nurses (N = 128) answered 1 (1, 2) questionnaires each. Considering first- and last-day questionnaires, one patient received ratings by, in median, 3 (3, 5) different caregivers. EOL decisions were initiated most frequently by attending ICU physicians (86%) and least often by relatives (14%) or nurses (2%). Discomfort with EOL decision was notably low among all groups (median 0 on a scale from 0 to 10). On the day the EOL decisions were made, patient wishes were known to attendings in 55% (73 of 132 patients). If patient wishes were known, no one perceived that they were not followed. Table 1 compares differing perspectives of involved caregiver groups. Nurses had less knowledge than physicians about EOL decision-making, for example, about initiation of decisions, understanding and consensus with relatives, patient wishes. Nurses also more often denied they had sufficient knowledge to talk with family. On the other hand, nurses more often than physicians rated symptom control to be insufficient and more often rated EOL care to be a burden (ratings ⩾ 4 on a 0–10 scale; p ⩽ 0.002). The perceived burden of EOL care was overall low among all groups (median 2–4). Table 2 shows the results of the qualitative analysis of reasons given by those 124 staff members (31% of total) who experienced an increased burden. The most frequent reasons were complex care requirements for 50% of nurses, and the unpredictable course of the disease for 31% attendings and 39% residents. In a subgroup of 30 patients (23%), attendings judged that EOL decisions should have been made earlier. Patients with delayed EOL decision were older than patients with non-delayed EOL decision (76.7 vs 70.9 years; p = 0.027); the interval between onset of sepsis-related organ dysfunction and the first EOL decision was not significantly different (median 121.5 (27.4–328.6 h) vs 109 h (28.8–301.7 h); p = 0.708).
Relatives’ psychological burden and perspectives on ICU care and decision-making After 90 days, relatives of 121 patients could be contacted. All contacted relatives answered initial questions about
overall satisfaction with ICU care, physician or nurse communication and own acute emotional suffering. Answers did not differ between those relatives who consented to participate in the full interview (84/121, 70%) and those who declined to participate (each of these items p ⩾ 0.15). All participants except three had suffered the death of their relative. Characteristics and responses of relatives are given in Table 3 and in Supplemental Table 1. Most relatives were female (74%), with a median age of 57 years. Most relatives (77%) were also acting as legal proxies. After 90 days, 41 relatives (51%) experienced symptoms of PTSD; 40 (48%) and 27 (33%) suffered from symptoms of anxiety and depression, respectively. Relatives felt very satisfied with medical care and communication (median 9 on a scale of 0–10), fully agreed with EOL decision and fully believed the patients’ will had been considered (both median 10). Still, 30% and 13%, respectively, would have considered talking with a psychologist or clergy helpful. In all, 43% of relatives believed the responsibility for EOL decisions was shared, 37% accorded sole responsibility to themselves and 20% to the physicians; 87% of relatives thought their degree of involvement had been just right and 12% had wanted less involvement. Almost all (98%) felt the decision had been timely. Only 9% perceived conflicts about EOL care.
Risk factors for relatives’ psychological burden In multiple regressions (Table 4), spouses showed higher values for depression and PTSD than children and other relatives (p 32) HADS Anxiety HADS Depression HADS Total Anxiety symptoms (HADS Anxiety > 7) Depression symptoms (HADS Depression > 7)
57 (50–66) 62 (73.8) 29 (34.5) 35 (41.7) 35 (41.7) 2 (2.4) 3 (3.6) 9 (10.7) 65 (77.4) 25 (31.6) 6 (9) 19 (30.2) 8 (13.1) 9 (7.92–9.67)
33 (20–49) 41 (50.6) 7 (5–10) 6 (2.25–9) 13 (8–18.75) 40 (48.2) 27 (32.9)
EOL: end-of-life; IES Impact of Event Scale; HADS: Hospital Anxiety and Depression Scale; IQR: interquartile range. N (%), Median (IQR): scale from 0 to 10 (0 = very little; 10 = very much). aItems only answered by relatives taking part in the longer version of the interview (N ⩽ 67). bSum score, including satisfaction with medical care, satisfaction with physician talks, satisfaction with information, having been given enough time, comprehensible information, completeness of information (see exploratory factor analysis in Supplemental Table 2).
Senior consultants initiated most EOL decisions and knew most about the decision-making process. Nurses initiated only 2% of EOL decision, similar to findings from a European survey.28 Fewer nurses than physicians knew patients’ wishes; over 40% of nurses lacked sufficient knowledge to talk with relatives and more nurses felt burdened by EOL care. European surveys repeatedly described a discrepancy between nurses’ preference and actual involvement in EOL decision-making.29–31 International researchers point out that ICU nurses could provide an essential role in the provision of effective EOL care, especially in the emotional support of patient and family at the EOL.32–34 However, nurses also need more encouragement, education and opportunities to develop the necessary skills for palliative care in the ICU.33
Half of the relatives suffered symptoms of PTSD and anxiety and one-third suffered from symptoms of depression after 90 days. These rates are higher than those found in family members of general ICU patients in France and the United States.5,35 However, 92% of our patients died, which may explain their spouses’ or children’s increased emotional symptoms. Azoulay et al.5 also reported that half the family members of patients who died in the ICU showed symptoms of PTSD after 90 days. Consistent with previous reports, we found that younger patient age and female gender of the relative were also associated with increased psychological strain.5,10,20 On the other hand, we found that satisfaction with medical care and communication was related to lower levels of anxiety and depression. It has been shown that the manner in which caregivers communicate with relatives can influence their psychological stress. For example, receiving incomplete information,36 not finding the patient’s physician comforting37 or discrepancy with the preferred decision-making role10 all increased stress. However, relatives in our study were highly satisfied with the manner of communication by medical staff. Only 9% perceived conflicting opinions, less than previously reported.38 Still, up to 30% of relatives would have liked to talk with a psychologist; unfortunately, a psychologist is not routinely available in our ICU. The potential association of relatives’ anxiety and depression with EOL decisions that were delayed is interesting but should be interpreted with care. The uncertainty estimates, that is, the standard error, might be underestimated because of non-independence in attendings’ ratings. Also, it was based on the subjective judgment by attending physicians that such decision could have been reached earlier, maybe due to uncertainty about patient wishes, reluctance by the family or by the co-treating surgeon to forgo LST.39 These patients were considerably older with a median age of about 77 years. The reduced strain for the family might be explained by the advanced age of the patient,20 or by having more time to come to terms with the inevitable outcome. This finding seemingly disagrees with the traditional view that it may be desirable to reduce the duration of ICU treatment in patients with a poor prognosis.40 However, it needs to be interpreted with caution and confirmed by suitable clinical investigation. This study has limitations and strengths. It is a singlehospital survey; therefore, findings may not be generalizable. Our ICUs are closed units; the ICU structure (closed versus open) may influence decision-making.41 We did not assess other possible influences on work management in the ICU, such as post graduate competency training or the influence of nurse–patient ratio.42,43 It was an exploratory study which limits the conclusions that can be drawn from the findings. Some of the questionnaire items were self-developed and not validated. However, questions reflected quality concerns of experienced caregivers and pilot-testing ensured their face validity. Strengths include
8
Palliative Medicine
Table 4. Risk factors for relatives’ psychological burden. Relatives’ psychological burden at 90 days
HADS anxiety
HADS depression
IES
Simple regression coefficients
Simple regression coefficients
Multiple regression coefficients (95% CI)
Simple regression coefficients
Multiple regression coefficients (95% CI)
−0.19′
−0.14 (−0.38 to 0.1)
−0.13
0.05 0.14 0 0.03 0.19
−0.16
−0.28*
log(patient age)
Multiple regression coefficients (95% CI)
0
SAPS-II at admission Patient died in ICU log(ICU LOS) Withdrawal of therapy Patient received surgical cotreatment Attending’s opinion that endof-life decision was delayed Relative, male
0.08 −0.11 0.02 0 0.3*
Relative, higher education Relative, child
−0.05 −0.15
−0.16 −0.52***
Relative, other relationship
−0.05
−0.28*
Relative perception that information was contradictory Relative satisfaction with medical care and communication in generala
0.08
0.06
−0.02
−0.24* (−0.45 to −0.03) −0.25* (−0.48 to −0.01) −0.19′ (−0.42 to 0.03)
−0.18
−0.11
−0.38** −0.21′
−0.25*
0.22′ (−0.04 to 0.48) −0.37** (−0.6 to −0.14) −0.2′ (−0.43 to 0.03)
−0.26* (−0.49 to −0.03)
0.04 0.04 0.1 0.02 0.09 −0.28*
−0.27* (−0.5 to −0.04)
−0.14 −0.48*** (−0.72 to −0.24) −0.33** (−0.54 to −0.11)
−0.12 −0.29* −0.21′
HADS: Hospital Anxiety and Depression Scale; IES: Impact of Event Scale; CI: confidence interval; SAPS-II: Simplified Acute Physiology Score–II; ICU: intensive care unit; LOS: length of stay. All coefficients are standardized. Predictors significant at level 0.1 in simple regression were considered for multiple regression. aSum score, including satisfaction with medical care, satisfaction with physician talks, satisfaction with information, having been given enough time, comprehensible information, completeness of information (see exploratory factor analysis in Supplemental Table 2). ′p ⩽ 0.1; *p ⩽ 0.05; **p ⩽ 0.01; ***p ⩽ 0.001.
the prospective design with patient-based surveys, a priori defined categories, identification of EOL decisions based on daily caregiver assessments not on documentation, surveying physicians and nurses and comparing their assessments, direct reporting of perceptions about each EOL decision rather than theoretical responses about EOL decision in general, the disease-specific focus whereby some of the variability reported in observations of EOL care is avoided, the relatively large sample of relatives compared to previous interview studies, linking patientrelated data with caregivers’ and relatives’ perspectives, performing a 90-day follow-up of relatives with validated survey tools that have been used previously5,11 and the combination of quantitative and qualitative data to gain more comprehensive understanding.
Conclusion Our study identifies some shortcomings of EOL communication and links satisfaction of relatives with ICU care and communication with long-term psychological outcomes.
In order to improve EOL care in our ICUs, communication should be improved within the ICU team to strengthen the involvement of nurses in EOL care. Improved communication between caregivers and the family might lessen the relatives’ long-term anxiety. Acknowledgements We thank study nurse Anja Haucke for her dedicated work. Note of prior abstract publication/presentation: Data were partly presented as abstract at the European Society of Intensive Care Medicine (ESICM) 2012 meeting, October 13–17, Lisbon.
Declaration of conflicting interests The authors have reported that no potential conflicts of interest exist. The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of this article.
Ethical approval Approval for this study was granted by the Ethics Committee of the Jena University Hospital (No. 2964-11/10).
Hartog et al. Funding The study was funded partially by the German Federal Ministry of Ministry of Education and Research (BMBF), Germany, FKZ: 01EO1002. Daniel Schwarzkopf is funded in full by the Center for Sepsis Control and Care (CSCC). The CSCC is funded by the German Federal Ministry of Ministry of Education and Research (BMBF). Konrad Reinhart, Christiane S. Hartog and Jochen Gensichen receive CSCC research grants.
References 1. Prendergast TJ and Puntillo KA. Withdrawal of life support: intensive caring at the end of life. JAMA 2002; 288: 2732–2740. 2. Carlet J, Thijs LG, Antonelli M, et al. Challenges in endof-life care in the ICU. Statement of the 5th International Consensus Conference in Critical Care: Brussels, Belgium, April 2003. Intensive Care Med 2004; 30: 770–784. 3. Puntillo KA and McAdam JL. Communication between physicians and nurses as a target for improving end-of-life care in the intensive care unit: challenges and opportunities for moving forward. Crit Care Med 2006; 34: S332– S340. 4. Jensen HI, Ammentorp J, Johannessen H, et al. Challenges in end-of-life decisions in the intensive care unit: an ethical perspective. J Bioeth Inq 2013; 10: 93–101. 5. Azoulay E, Pochard F, Kentish-Barnes N, et al. Risk of post-traumatic stress symptoms in family members of intensive care unit patients. Am J Respir Crit Care Med 2005; 171: 987–994. 6. Griffith L, Cook D, Hanna S, et al. Clinician discomfort with life support plans for mechanically ventilated patients. Intensive Care Med 2004; 30: 1783–1790. 7. A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks of treatments (SUPPORT). The SUPPORT Principal Investigators. JAMA 1995; 274: 1591–1598. 8. Davidson JE, Jones C and Bienvenu OJ. Family response to critical illness: postintensive care syndrome-family. Crit Care Med 2012; 40: 618–624. 9. McAdam JL, Dracup KA, White DB, et al. Symptom experiences of family members of intensive care unit patients at high risk for dying. Crit Care Med 2010; 38: 1078–1085. 10. Gries CJ, Engelberg RA, Kross EK, et al. Predictors of symptoms of posttraumatic stress and depression in family members after patient death in the ICU. Chest 2010; 137: 280–287. 11. Lautrette A, Darmon M, Megarbane B, et al. A communication strategy and brochure for relatives of patients dying in the ICU. N Engl J Med 2007; 356: 469–478. 12. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29: 1303–1310. 13. Sprung CL, Cohen SL, Sjokvist P, et al. End-of-life practices in European intensive care units: the Ethicus Study. JAMA 2003; 290: 790–797. 14. Meissner A, Genga KR, Studart FS, et al. Epidemiology of and factors associated with end-of-life decisions in
9 a surgical intensive care unit. Crit Care Med 2010; 38: 1060–1068. 15. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 1992; 20: 864–874. 16. Schmidt K, Gensichen J, Thiel P, et al. A prospective observational web-based registry for assessment of long-term outcome in patients with severe sepsis. Infection 2013; 41(Suppl. 1): S84–S85. 17. Ferring D and Filipp S-H. Teststatistische Überprüfung der Impact of Event-Skala: Befunde zu Reliabilität und Stabilität. Diagnostica 1994; 40: 344–462. 18. Herrmann C and Buss U. Vorstellung und Validierung einer deutschen Version der ‘Hospital Anxiety and Depression Scale’ (HAD-Skala): Ein Fragebogen zur Erfassung des psychischen Befindens bei Patienten mit körperlichen Beschwerden. Diagnostica 1994; 40: 143–154. 19. Kentish-Barnes N, Lemiale V, Chaize M, et al. Assessing burden in families of critical care patients. Crit Care Med 2009; 37: S448–S456. 20. Kross EK, Engelberg RA, Gries CJ, et al. ICU care associated with symptoms of depression and posttraumatic stress disorder among family members of patients who die in the ICU. Chest 2011; 139: 795–801. 21. Pillai L, Aigalikar S, Vishwasrao SM, et al. Can we predict intensive care relatives at risk for posttraumatic stress disorder? Indian J Crit Care Med 2010; 14: 83–87. 22. Wall RJ, Curtis JR, Cooke CR, et al. Family satisfaction in the ICU: differences between families of survivors and nonsurvivors. Chest 2007; 132: 1425–1433. 23. Gries CJ, Curtis JR, Wall RJ, et al. Family member satisfaction with end-of-life decision making in the ICU. Chest 2008; 133: 704–712. 24. Le Gall JR, Lemeshow S and Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/ North American multicenter study. JAMA 1993; 270: 2957– 2963. 25. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsisrelated Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996; 22: 707–710. 26. Huang R. RQDA: R-based qualitative data analysis (R package version 0.2-2). http://rqda.r-forge.r-project.org/, 2011. 27. Braun V and Clarke V. Using thematic analysis in psychology. Qual Res Psychol 2006; 3: 77–101. 28. Benbenishty J, Ganz FD, Lippert A, et al. Nurse involvement in end-of-life decision making: the ETHICUS Study. Intensive Care Med 2006; 32: 129–132. 29. Ferrand E, Lemaire F, Regnier B, et al. Discrepancies between perceptions by physicians and nursing staff of intensive care unit end-of-life decisions. Am J Respir Crit Care Med 2003; 167: 1310–1315. 30. Jox RJ, Krebs M, Fegg M, et al. Limiting life-sustaining treatment in German intensive care units: a multiprofessional survey. J Crit Care 2010; 25: 413–419. 31. Jensen HI, Ammentorp J, Erlandsen M, et al. Withholding or withdrawing therapy in intensive care units: an analysis
10
32.
33.
34.
35.
36.
37.
Palliative Medicine of collaboration among healthcare professionals. Intensive Care Med 2011; 37: 1696–1705. Nelson JE, Cortez TB, Curtis JR, et al. Integrating palliative care in the ICU: the nurse in a leading role. J Hosp Palliat Nurs 2011; 13: 89–94. Efstathiou N and Clifford C. The critical care nurse’s role in end-of-life care: issues and challenges. Nurs Crit Care 2011; 16: 116–123. Ranse K, Yates P and Coyer F. End-of-life care in the intensive care setting: a descriptive exploratory qualitative study of nurses’ beliefs and practices. Aust Crit Care 2012; 25: 4–12. Anderson WG, Arnold RM, Angus DC, et al. Posttraumatic stress and complicated grief in family members of patients in the intensive care unit. J Gen Intern Med 2008; 23: 1871–1876. Azoulay E, Adrie C, De Lassence A, et al. Determinants of postintensive care unit mortality: a prospective multicenter study. Crit Care Med 2003; 31: 428–432. Siegel MD, Hayes E, Vanderwerker LC, et al. Psychiatric illness in the next of kin of patients who die in the intensive care unit. Crit Care Med 2008; 36: 1722–1728.
38. Abbott KH, Sago JG, Breen CM, et al. Families looking back: one year after discussion of withdrawal or withholding of life-sustaining support. Crit Care Med 2001; 29: 197–201. 39. Buchman TG. Surgeons and their patients near the end of life. Crit Care Med 2010; 38: 995–996. 40. Schneiderman LJ, Gilmer T, Teetzel HD, et al. Effect of ethics consultations on nonbeneficial life-sustaining treatments in the intensive care setting: a randomized controlled trial. JAMA 2003; 290: 1166–1172. 41. Cassell J, Buchman TG, Streat S, et al. Surgeons, intensivists, and the covenant of care: administrative models and values affecting care at the end of life – updated. Crit Care Med 2003; 31: 1551–1557; discussion 7–9. 42. Miranda DR, Rivera-Fernandez R and Nap RE. Critical care medicine in the hospital: lessons from the EURICUSstudies. Med Intensiva 2007; 31: 194–203. 43. Shortell SM, Zimmerman JE, Rousseau DM, et al. The performance of intensive care units: does good management make a difference? Med Care 1994; 32: 508–525.
