The Validity of Advance Directives in Acute Situations
A survey of doctors‘ and relatives‘ perceptions from an intensive care unit
; ; ; ; ;
Background: Nearly every fourth person in Germany has an advance directive that is to be used in certain medical situations. It is questionable, however, whether advance directives truly influence important treatment decisions in the intensive care unit. We studied the extent to which doctors and patients’ relatives agree on the applicability of advance directives in the acute setting.
Methods: A prospective study was carried out by questionnaire among the physicians and relatives of 50 patients with advance directives who were hospitalized on four different multidisciplinary intensive care units. The answers of 25 residents in training, 14 senior physicians, and 19 relatives were analyzed both quantitatively and qualitatively. The extent of agreement was assessed by means of Gwet’s AC1 with linear weighting.
Results: In most of the advance directives, the conditions under which they were meant to apply were stated in broad, general terms in prewritten blocks of text. 23 of the 50 patients (46%) died. All relatives stated that they were very familiar with the patients’ wishes; 18 of 19 were legally responsible for decision-making. In assessing whether the advance directive was applicable to the situation at hand, the strength of agreement between physicians and relatives as well as between the two groups of physicians was only fair and non-significant (0.35; 95% confidence interval [CI]: –0.01 to 0.71; p = 0.059 and 0.24; 95% CI: –0.03 to 0.50; p = 0.079). The relatives found the advance directives more useful than the doctors did (median, 5 vs. 3 [p = 0.018] on a Likert scale ranging from 0 [not useful at all] to 5 [very useful]) and favored their literal application (median,
5 vs. 4 [p = 0.018] on a Likert scale ranging from 0 [favoring the doctor’s interpretation] to 5 [favoring literal application]). 30 days after the decision,
13 relatives (68%) felt that the patient’s wishes had been fully complied with.
Conclusion: These groups’ clearly differing assessments of the applicability of advance directives imply that the currently most common types of advance directive are not suitable for use in intensive care. In order to support patients’ relatives in their role as surrogate participants in decision-making, improved advance directives should be developed, and their implementation should be incorporated into the training and continuing medical education of intensive-care physicians.
Almost one in four persons in Germany report that they have prepared an advance directive (1). Due to demographic changes, older, sicker patients are increasingly treated in intensive care units and as a result are more and more likely to die there. Advance directives are intended to guarantee an individual’s autonomy when that individual can no longer express himself or herself as a result of disease severity or treatment. Retrospective studies, however, show that advance directives in intensive care units have only a small effect on the management of life-sustaining treatment (2, 3). A retrospective data analysis found that although cardiopulmonary resuscitation was performed less frequently in patients who had an advance directive at the time of death there was no effect on the length of time spent in the ICU or on end-of-life treatment (2).
In intensive care units advance directives for patients who are unable to consent are binding for all those involved, provided they are valid. The representative—or, if the representative is absent or unavailable, the treatment team—must therefore first examine the directive’s validity (4). ICU doctors complain that advance directives are often too generally worded and are therefore rarely applicable to an ICU patient’s situation (5). To investigate this problem, we asked treating physicians and relatives of intensive care patients who were unable to consent to rate the validity of available advance directives.
Study design and conduct
A mixed-method study was conducted in four intensive care units (mixed surgical, cardiological, and neurological) of a university hospital, with a total of 72 beds, between September 2013 and March 2014. All ICU patients were screened on weekdays. Inclusion criteria were ICU stay lasting longer than 48 hours, loss of ability to consent as assessed by the treating senior physician, and availability of an advance directive. Senior physicians, residents in training, and next of kin were questioned in a structured interview including both closed and open questions within 48 hours of the patient’s inclusion in the study. All the interviews were conducted by one individual (NL). Answers to open questions were recorded and transcribed. Thirty days after the patients died or were discharged from the ICU we questioned the relatives again, by telephone, in a standardized interview. This procedure had been approved by the university hospital’s ethics committee (approval no. 3732–03/13).
For the primary question of how often advance directives were rated as valid, directives had to be rated as meeting at least one validity criterion. We also investigated the following points:
- Were there differences between the assessments of those involved?
- How were individual validity criteria rated?
- What reasons were given for the assessments?
- Were advance directives rated as helpful?
- Should they be interpreted literally?
- How did relatives rate the implementation of patients’ wishes 30 days later?
The development of the guideline for conducting the interviews is described in the eBox. The questionnaires are also available as part of the supplementary material (eQuestionnaires).
Advance directives were analyzed with regard to type of text (template or free text), type of validity criteria, and treatment wished or refused. The patients’ data was taken from their electronic patient records. Cross-classified tables were drawn up to compare how validity was rated by doctors and relatives on the one hand, and by senior physicians and residents on the other. Gwet’s AC1 (6) with linear weighting (7) was calculated as a measure of agreement. Senior physicians’ assessments were used in comparison of doctors and relatives, and missing interviews were replaced by information on the same patient provided by residents. Doctors’ and relatives’ answers on a scale were compared using the Wilcoxon rank-sum test. Comments were transcribed and qualitatively evaluated (eBox). Relatives’ and doctors’ demographic characteristics were described using descriptive statistics, and the judgements of resident and senior physicians were compared using Fisher’s exact test or the chi-square test.
Patients, advance directives, and interview partners
Of 2112 patients treated in the ICU during the study period, 134 met the inclusion criteria. For 50 of these, advance directives were available and relatives could be interviewed (Figure 1). Patients were mostly male and a median of 71.5 years old. The most common grounds for admission were medical. At the time of questioning, approximately two-thirds of the patients were receiving mechanical ventilation, 29 (58%) through a tube and four (8%) via a tracheostomy. Life-sustaining measures were limited in 56% of patients (Table 1). The ICU mortality rate was 46%. A median of three days elapsed between ICU admission and the point at which an advance directive documented in the patient’s medical records became available. Without exception, all advance directives contained a form of rejection of life-sustaining measures; 39 (78%) consisted of prewritten sections of text or fields to be checked, while the remainder were written independently by their authors. However, these often also contained standard wordings. Most advance directives met three or four validity criteria; one met as many as eight (Table 2).
It proved impossible to contact 28 of the relatives within the time window, and three refused to participate. Interviews were conducted in person with 19 relatives and by telephone with eight, in line with their own wishes. Most relatives were female (84%), spouses or life partners (68%), with a median age of 62 years. Eighteen were authorized to act as proxies to arrange medical care. All relatives stated that they were very familiar with the patient’s wishes. All the relatives were successfully contacted after 30 days. Of those questioned, 13 (68%) declared that the patient’s wishes had been followed during their ICU treatment. However, only two advance directives were assessed as suited to ICU treatment by relatives and doctors (Table 3).
Senior physicians (n = 14) were older (p = 0.036) and more experienced with advance directives in their daily work (p = 0.003) than residents (n = 25). Only two senior physicians and three residents had prepared an advance directive or arranged a medical care proxy themselves (eTable 1).
Rating of advance directives
The assessments of senior physicians were available for 43 advance directives, those of residents for 46, and those of relatives for 19. The assessments of residents and senior physicians could be compared in 39 cases, and those of relatives and doctors in 19 cases. Doctors declared 17 of 39 advance directives valid when questioned, but only six were rated as valid by both senior physicians and residents (Figure 2). Assessment of advance directive validity by senior physicians and residents matched slightly, but insignificantly (AC1 = 0.24; 95% confidence interval [95% CI]: –0.03 to 0.50; p = 0.079). When individual validity criteria were rated, doctors showed substantial agreement (AC1 = 0.7; 95% CI: 0.61 to 0.79; p ≤0.001). Of the 153 validity criteria, 25 were rated satisfactory, but only five were rated so by both groups (Table 4). When the assessments of doctors and relatives were compared, it was found that although eight advance directives were declared valid only four were declared so by both parties (Figure 2, Table 4). Of a total of 82 individual validity criteria, 22 were described as having been met, but assessments coincided in only six of these cases (Table 4) (AC1 for comparison of doctors and relatives found 0.35; 95% CI: –0.01 to 0.71; p = 0.059 and 0.38; 95% CI: 0.20 to 0.56; p ≤0.001 respectively). This indicated a slight degree of agreement (7).
Doctors explained the reasons for their assessments of advance directive validity in 78 of the questionnaires. Of these comments, 60 concerned the patient’s state of health or likely prognosis, while 17 concerned unclear or excessively general wording in the advance directive. Turning to relatives, 13 gave reasons for their assessments, five could not state the patient’s wishes in their current situation on the basis of the advance directive, and three were guided by doctors’ opinions (eTable 2).
Handling advance directives, follow-up
The answers of 17 doctor–relative pairs to the question of how literally an advance directive should be followed were evaluated. Relatives gave a median answer of 5 (interquartile range [IQR]: 3 to 5), and doctors a median answer of 3 (IQR: 1 to 4; p = 0.018). The scale ranged from 0 (according to the doctor’s interpretation) to 5 (literally). On the question of how helpful the advance directive was for the treatment decisions being faced, again 17 doctor–relative pairs provided comments, and these were evaluated. This scale ranged from 0 (not at all helpful) to 5 (very helpful). Relatives gave a median answer of 5 (IQR: 4 to 5), and treating physicians 4 (IQR: 2 to 5; p = 0.018). Doctors made comments in 82 interviews. Of these, 29 comments rated the advance directives as having excessively general wording or contradictory wording, 21 rated them as a source of additional knowledge concerning the patient, and 13 as a binding guide to treatment. Of the 13 relatives who stated their reasons, eight saw the advance directive as binding, three as helpful for a dignified death, and three as a means to put the choice of treatment in the hands of relatives. Three relatives complained of the difficulty of the decision. Doctors who had rated an advance directive as valid or uncertain were asked whether they experienced personal conflict when implementing the patient’s wishes. This question was answered by 22 doctors; 19 answered negatively, and three positively. Statements concerning the applicability of patients’ written wishes were made by 23 doctors and four relatives. Five treating physicians and three patient representatives were unsure or considered the wishes only partly applicable (eTable 2).
For 38 of the advance directives, doctors gave an assessment of whether it would have been helpful for the advance directive to be available when the patient was admitted to the ICU: 64% answered negatively, and 11% positively. For the other cases this question was not applicable, as the advance directives had been available.
After 30 days 13 (68%) of relatives said yes when asked whether the patient’s wishes had been fully implemented in the ICU. Only one person stated that the patient’s wishes had not been followed (Table 3). Comments illustrate the following problems on the part of relatives (eTable 2):
- Responsibility for life and death (“I’d have to say now that [she] should have died.”)
- Unclear wording of advance directive (“what was written down wasn’t enough.”)
- Altered normality (“He’s […] changed hugely […] but he’s still alive.”)
- Change in the patient’s perspective (“[…] my mother now [says] she’d want to be alive still.”)
This study is the first to investigate how relatives and treating physicians rated the validity of an available advance directive in a particular situation. The judgements regarding an advance directive or individual validity criteria, both of different groups of doctors and of treating physicians and relatives, coincided in only half the cases. Previous observational studies have shown that an advance directive has only a small effect on end-of-life treatment. However, neither doctors’ nor relatives’ judgement of advance directives had been investigated (2, 3, 8, 9). In contrast, the authors of a large US cohort study concluded that treatment of patients who were unable to consent at the end of life was more likely to comply with the patients’ wishes if patients had an advance directive (10). However, this finding was subject to criticism, as relatives’ assessments were given after a delay: a median of approximately13 months later (11).
Reasons for the low level of agreement between assessments
ICU doctors tend to interpret the spirit of advance directives, whereas relatives understand them literally. Doctors often describe the wording used in advance directives as unclear, contradictory, and not applicable to the actual medical situation. The written content of advance directives usually consisted of prewritten sections of text that were imprecise or contradictory and in critical cases insufficient to determine their validity unambiguously in view of ICU patients’ complex clinical conditions and uncertain prognoses. A similar finding was reached in a small survey of ICU doctors (5). Advance directives tended to serve doctors as a guideline, and in particular they provided additional knowledge concerning the patient. For the treatment decisions to be taken, however, they were rated as of little use. It therefore comes as no surprise that only a minority of the doctors have prepared an advance directive themselves.
On the one hand ICU patients’ relatives have great respect for the signed directive and want to implement it literally, but on the other hand advance directives fail because the ways in which they are worded do not help relatives make decisions or provide them with instructions. Although relatives know patients’ wishes precisely from discussion and generally find an advance directive helpful, comments show that its applicability is often impossible to ascertain with certainty in the specific situation faced. Relatives find it very challenging to bring their own needs into line with the patient’s wishes (12). They often suffer considerable psychological after-effects for long periods (13).
Senior physicians, who are usually older and tend to have more experience with intensive care, rated advance directives differently from residents undergoing specialist training. In an earlier study involving a questionnaire, which was based on a hypothetical case with an available advance directive, doctors and nurses also rated applicability differently from each other (14). There are many reasons for the differences in assessments, including personality and sex. A lack of training in palliative care as part of initial training and continuing medical education was also mentioned. Issues relating to palliative care play too small a role in the training provided to students (15). Because such issues are becoming increasingly important in an aging society, they should also be part of specialist training in intensive care medicine (16).
Low levels of conflict
In this study, after 30 days 68% of relatives answered that ICU treatment had fully complied with the patient’s wishes. This suggests that a consensus between those involved may have developed subsequently. Determining and interpreting the patient’s wishes is a joint task for treating physicians and patient representatives. Conversations with relatives as part of a dialogue-based process are particularly helpful in this (17). Trusting dialogue can also provide emotional support for the relatives (17). Relatives’ retrospective satisfaction is therefore probably also affected by the quality of such conversations. Physicians expressed few potential conflicts when individual patient wishes were implemented. Concerns relating to medical law, such as those mentioned in earlier surveys (18), were not raised. This may be taken as a sign that advance directives have become part of daily practice in ICUs since those earlier surveys were conducted.
Appropriate advance directives for ICUs
Although the actual situation cannot usually be predicted precisely, the findings of this study suggest it would be desirable for individuals to prepare advance directives that are assessed in the same way by all those involved in the event of ICU treatment and are directly applicable. Advance care planning (ACP) may be suitable for developing informed preferences for specific scenarios, discussing them with relatives, and documenting them. However, this requires qualified and, where appropriate, longer-term consultation discussions (19). An ACP program has already been successfully implemented in German nursing homes for the elderly (20). French intensive care doctors presented films showing scenarios they had faced in the ICU to individuals aged over 80 years and then asked them what their wishes were. Up to two-thirds of the participants refused mechanical ventilation or renal replacement therapy following a complication during intensive care (21). The majority of intensive care doctors who were asked about these scenarios tailored their treatment recommendations to the participants’ wishes (22).
Study strengths and weaknesses
The extent to which the inquiries made in this study can be generalized is limited, as although they were conducted in multidisciplinary ICUs they were conducted in only one hospital. The documented opinions provide only a snapshot at a single point in time, and subsequent changes resulting from clinical alterations have not been taken into account. The strengths of the study are its use of qualitative and quantitative procedures and the summary of the wording of the clauses that were prepared and made available for the inquiries.
We would like to thank Prof. Randall J. Curtis (University of Washington, Seattle, USA), Dr. Birgit van Oorschot (University Hospital of Würzburg), Dr. Ulrike Skorsetz (Ethics Committee Chair, University Hospital of Jena), and Dr. Heike Hoyer (Institute of Medical Statistics, Computer Sciences and Documentation, University Hospital of Jena) for their valuable support in planning this study. We would also like to thank Julia Hase for collaborating on subject-related analysis of the comments.
Conflict of interest statement
Ms. Leder received a doctoral grant from the Interdisciplinary Center for Clinical Research (IZKF, Interdisziplinäres Zentrum für Klinische Forschung), University Hospital of Jena for this study.
The study was partly financed by the Center for Sepsis Control and Care, University Hospital of Jena, which is funded by Germany’s Federal Ministry of Education and Research: no. FKZ 01EO1002.
The other authors declare that no conflict of interest exists.
Manuscript received on 10 April 2015, revised version accepted on 24 June 2015.
Translated from the original German by Caroline Shimakawa-Devitt, M.A.
PD Dr. med. Christiane S. Hartog
Department of Anesthesiology and Intensive Care Medicine
Center for Sepsis Control and Care
University Hospital of Jena
Erlanger Allee 101
07747 Jena, Germany
For eReferences please refer to:
eBox, eTables, eQuestionaire:
Center for Sepsis Control and Care (CSCC), Jena University Hospital: Leder, Dipl.-psych. Schwarzkopf, Prof. Dr. med. Reinhart, PD Dr. med. Hartog
Department of Neurology, Jena University Hospital: Prof. Dr. med. Witte
Department of Internal Medicine I, Jena University Hospital: Dr. med. Pfeifer
|1.||Deutsche Schlaganfall-Hilfe: Umfrage: Immer mehr Deutsche erstellen Patientenverfügung. www.schlaganfall-hilfe.de/patientenverfuegung (last accessed on 15 June 2015).|
|2.||Hartog CS, Peschel I, Schwarzkopf D, et al.: Are written advance directives helpful to guide end-of-life therapy in the intensive care unit? A retrospective matched-cohort study. J Crit Care 2014; 29: 128–33.|
|3.||Halpern NA, Pastores SM, Chou JF, Chawla S, Thaler HT: Advance directives in an oncologic intensive care unit: a contemporary analysis of their frequency, type, and impact. J Palliat Med 2011; 14: 483–9 CrossRef MEDLINEPubMed Central|
|4.||Jannsens U, Burchardi H, Duttge G, et al.: Therapiezieländerung und Therapiebegrenzung in der Intensivmedizin: Positionspapier der Sektion Ethik der Deutschen Interdisziplinären Vereinigung für Intensiv- und Notfallmedizin (DIVI); 2012; 3: 103–7.|
|5.||Langer S, Knorr JU, Berg A: Umgang mit Patientenverfügungen: Probleme durch pauschale Formulierungen. Dtsch Arztebl 2013; 110: 2186–8.|
|6.||Gwet KL: Computing inter-rater reliability and its variance in the presence of high agreement. Br J Math Stat Psychol 2008; 61: 29–48 CrossRef MEDLINE|
|7.||Landis JR, Koch GG: The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159–74 CrossRef MEDLINE|
|8.||Brinkman-Stoppelenburg A, Rietjens JA, van der Heide A: The effects of advance care planning on end-of-life care: a systematic review. Palliat Med 2014; 28: 1000–25 CrossRef MEDLINE|
|9.||Kish Wallace S, Martin CG, Shaw AD, Price KJ: Influence of an advance directive on the initiation of life support technology in critically ill cancer patients. Crit Care Med 2001; 29: 2294–8 CrossRefMEDLINE|
|10.||Silveira MJ, Kim SY, Langa KM: Advance directives and outcomes of surrogate decision making before death. N Engl J Med 2010; 362: 1211–8 CrossRe MEDLINE PubMed Central|
|11.||Gillick MR: Reversing the code status of advance directives? N Engl J Med 2010; 362: 1239–40 CrossRefMEDLINE|
|12.||Schenker Y, Crowley-Matoka M, Dohan D, Tiver GA, Arnold RM, White DB: I don’t want to be the one saying ’we should just let him die’: intrapersonal tensions experienced by surrogate decision makers in the ICU. J Gen Intern Med 2012; 27: 1657–65 CrossRef MEDLINE PubMed Central|
|13.||Davidson JE, Jones C, Bienvenu OJ: Family response to critical illness: postintensive care syndrome-family. Crit Care Med 2012; 40: 618–24 CrossRe MEDLINE|
|14.||Thompson T, Barbour R, Schwartz L: Adherence to advance directives in critical care decision making: vignette study. BMJ 2003; 327: 1011 CrossRef MEDLINEPubMed Central|
|15.||Weber M, Schmiedel S, Nauck F, Alt-Epping B: Knowledge and attitude of final–year medical students in Germany towards palliative care—an interinstitutional questionnaire-based study. BMC Palliat Care 2011; 10: 19 CrossRef MEDLINEPubMed Central|
|16.||Wiese CH, Vagts DA, Kampa U, et al.: [Palliative care oriented therapy for all patients : recommendations of an expert circle]. Anaesthesist 2012; 61: 529–36.|
|17.||Borasio GD: Das Patientenverfügungsgesetz und die medizinische Praxis. In: Borasio GD, Hessler HJ, Jox RJ, Meier C (eds.): Patientenverfügung: Das neue Gesetz in der Praxis. Stuttgart: Kohlhammer 2012; 26–36.|
|18.||van Oorschot B, Lipp V, Tietze A, Nickel N, Simon A: [Attitudes on euthanasia and medical advance directives]. Dtsch Med Wochenschr 2005; 130: 261–5.|
|19.||Lund S, Richardson A, May C: Barriers to advance care planning at the end of life: an explanatory systematic review of implementation studies. PLoS One 2015; 10: e0116629 CrossRef MEDLINE PubMed Central|
|20.||In der Schmitten J, Lex K, Mellert C, Rotharmel S, Wegscheider K, Marckmann G: Implementing an advance care planning program in German nursing homes: results of an inter-regionally controlled intervention trial. Dtsch Arztebl Int 2014; 111: 50–7.|
|21.||Philippart F, Vesin A, Bruel C, et al.: The ETHICA study (part I): elderly’s thoughts about intensive care unit admission for life-sustaining treatments. Intensive Care Med 2013; 39: 1565–73.|
|22.||Garrouste-Orgeas M, Tabah A, Vesin A, et al.: The ETHICA study (part II): simulation study of determinants and variability of ICU physician decisions in patients aged 80 or over. Intensive Care Med 2013; 39: 1574–83.|
|e1.||Huang R: Rqda: R-based qualitative data analysis. R package version 0.2–2. 2011. rqda.r-forge.r-project.org (last accessed 20 September 2012).|
|e2.||Kuckartz U: Qualitative Inhaltsanalyse. Methoden, Praxis. Computerunterstützung. Weinheim, Basel: Beltz Juventa 2012.|