Original Manuscript
Psychosocial determinants of nurses’ intention to practise euthanasia in palliative care
Nursing Ethics 1–13 ª The Author(s) 2014 Reprints and permission: sagepub.co.uk/journalsPermissions.nav 10.1177/0969733014557117 nej.sagepub.com
Mireille Lavoie Laval University, Canada; Centre de recherche du CHU de Que´bec, Canada
Gaston Godin, Lydi-Anne Ve´zina-Im and Danielle Blondeauy Laval University, Canada
Isabelle Martineau Maison Michel-Sarrazin, Canada
Louis Roy Hoˆpital Enfant-Je´sus, Canada
Abstract Background: Most studies on euthanasia fail to explain the intentions of health professionals when faced with performing euthanasia and are atheoretical. Research objective: The purpose of this study was to identify the psychosocial determinants of nurses’ intention to practise euthanasia in palliative care if it were legalised. Research design: A cross-sectional study using a validated anonymous questionnaire based on an extended version of the Theory of Planned Behaviour. Participants and research context: A random sample of 445 nurses from the province of Quebec, Canada, was selected for participation in the study. Ethical considerations: The study was reviewed and approved by the Ethics Committee of the Centre hospitalier universitaire de Que´bec. Findings: The response rate was 44.2% and the mean score for intention was 4.61 + 1.90 (range: 1–7). The determinants of intention were the subjective (odds ratio ¼ 3.08; 95% confidence interval: 1.50– 6.35) and moral (odds ratio ¼ 2.95; 95% confidence interval: 1.58–5.49) norms. Specific beliefs which could discriminate nurses according to their level of intention were identified. Discussion: Overall, nurses have a slightly positive intention to practise euthanasia. Their family approval seems particularly important and also the approval of their medical colleagues. Nurses’ moral norm was related to beneficence, an ethical principle. Conclusion: To our knowledge, this is the first study to identify nurses’ motivations to practise euthanasia in palliative care using a validated psychosocial theory. It also has the distinction of identifying the ethical principles underlying nurses’ moral norm and intention.
y
D.B. was co-researcher but died in the process of revising the article.
Corresponding author: Mireille Lavoie, Faculty of Nursing, Laval University, 1050 avenue de Me´decine, Que´bec City, QC G1V 0A6, Canada. Email: [email protected]
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
2
Nursing Ethics
Keywords Determinant, euthanasia, intention, nurse, palliative care
Background Euthanasia (from the Greek eu (good) and thanatos (death)) is sometimes put forward as an alternative to palliative care.1 It must be recognised that despite great progress in palliative care, caregivers are still confronted with the physical, psychological or spiritual limitations of their power to alleviate suffering. In some cases, it is impossible to provide suitable treatment for patients whose suffering may be seen as ‘unnecessary’ or even ‘inhuman’. Yet, this practice is also held to be at odds with the philosophy of palliative care,2 which aims to relieve suffering while accompanying patients until the end of their lives. It is thus not surprising, even though euthanasia is illegal in Canada, that it remains a controversial topic and generates far from unanimous support among healthcare professionals. According to previous reviews of studies on attitude towards euthanasia among nurses, religion, age and nursing specialty significantly affect their views on euthanasia.3,4 Moreover, arguments for or against euthanasia among nurses would be predominantly guided by the following four ethical principles: respect for autonomy, non-maleficence, beneficence and justice.5 Another review also added quality of life, sanctity of life and the slippery slope as other important principles underlying nurses’ arguments for or against euthanasia.4 Nurses’ main arguments for supporting euthanasia appear to be the patient’s right to die with dignity, the wish to relieve the patient of unbearable pain (beneficence) and the patient’s right to decide about his/her life and death (autonomy).3 On the other hand, nurses’ main arguments for being against euthanasia are the following: that primary care should be about alleviating pain not ending life (non-maleficence, sanctity of life), their confidence in palliative care, religious and moral objections, and a concern for potential abuse such as performing euthanasia without patients’ consent.3 Most studies on euthanasia have explored the opinion of health professionals regarding its legalisation or their experience when this act is requested. However, they failed to explain the intentions of caregivers faced with performing an act of euthanasia, especially as most studies are also atheoretical.6 Therefore, the purpose of this study is to identify the psychosocial determinants of nurses’ intention to practise euthanasia in palliative care if it were legal in order to enlighten decision-makers regarding the impact of this practice on care practices and on the health professionals themselves.
Theoretical framework The study was guided by an extended version of the Theory of Planned Behaviour (TPB).7 Its efficacy in predicting intentions to adopt various health behaviours, including among health professionals, and the key role of intentions to predict behaviours have already been clearly established in a number of metaanalyses.8–13 According to the TPB, behaviour is predicted by intention and by perceived behavioural control (PBC) when the context is less volitional. Intention, in return, is formed of the following three constructs: attitude, subjective norm and PBC. Attitude is an evaluation, either positive or negative, of the adoption of a given behaviour. In this study, attitudes were evaluated by using its two components, cognitive and affective attitude as suggested by Triandis.14 Cognitive attitude refers to factual consequences (e.g. useless/useful) of the adoption of a given behaviour, while affective attitude is rather concerned with emotional consequences (e.g. unpleasant/pleasant). Subjective norm represents the perceived social pressure to adopt a given behaviour. PBC refers to people’s evaluation of their ability to adopt a given behaviour. Each construct is moreover related to a specific set of beliefs: attitude is related to behavioural beliefs, subjective norm to normative beliefs and PBC to control beliefs. Behavioural beliefs refer 2
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
3
to the perceived positive and negative consequences of the adoption of a given behaviour. Normative beliefs represent how individuals believe people who are important to them would react if they adopted a given behaviour (i.e. approve or disapprove). Control beliefs refer to elements that can hinder (barriers) or facilitate (facilitating factors) behavioural adoption. External factors such as socio-demographic variables (e.g. age, gender) can also influence the intention to adopt a given behaviour through the other constructs. New variables can be added to the TPB as long as they improve its predictive ability.7 In this study, two variables were added, professional norm and moral norm, given that there is evidence that they are determinants of health professionals’ intention to adopt various behaviours.15–19 Professional norm refers to the appropriateness of adopting a behaviour given one’s profession. Moral norm is a variable originating from the Theory of Interpersonal Behaviour.14 It is related to the appropriateness of adopting a given behaviour according to one’s personal and moral values. In another study on consent to organ donation, Blondeau et al.20 measured moral norm by using the following three ethical principles: autonomy, beneficence and justice. The results indicated that moral norm is mainly predicted by beneficence. In this study, moral norm, as well as beneficence, justice and autonomy, were then included. Concerning the concept of autonomy, the design of the study allowed us to assess its effect on intention directly since we used two versions of the questionnaire: in one version, the patient’s wishes concerning euthanasia were known, while in the other version, the patient’s wishes on this matter were unknown. In terms of external factors, the following socio-demographic and contextual variables were assessed: job position, main domain of practice, number of end-of-life patients nurses cared for in the past year and the percentage of their practice they represented, whether they have relatives who received palliative care before their death, years of experience, worksite, age, gender, level of education, religious affiliation and attitude towards the legalisation of euthanasia in Canada.
Methods Population and sample The population under study consisted of nurses from the province of Quebec, Canada. Head nurses were excluded from the study since they do not work at the bedside of patients. Nurses with underage patients (e.g. paediatrics) or patients with mental diseases (psychiatry) or whose job makes them unlikely to care for end-of-life patients (e.g. rehabilitation, plastic surgery) were also excluded. To obtain our sample, the provincial nurses’ professional Order of Quebec provided us with a list of their active members. A random sample of 445 nurses was obtained using random digit tables. The sample was weighted according to the domains of practice included in the study to reflect as closely as possible their distribution in the province. The study was reviewed and approved by the Ethics Committee of the Centre hospitalier universitaire (CHU) de Que´bec (reference number: HCS11-06-047).
Questionnaire development and validation The questionnaire was developed in accordance with the methodology recommended by the author of the TPB. A questionnaire based on this theory has to reflect the salient beliefs of a population. It is worth mentioning that in the original study, physicians’ psychosocial determinants were also identified (to be published), but only the results pertaining to nurses are reported in this article. To develop the belief items of the questionnaire, 29 health professionals (21 nurses and 8 physicians) sharing the same sociodemographic characteristics as those of the study sample completed a short open-ended questionnaire containing the following questions: (1) advantages and disadvantages of practicing euthanasia in palliative care 3
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
4
Nursing Ethics
(behavioural beliefs for cognitive attitude), (2) emotions that could encourage and hinder practicing euthanasia in palliative care (behavioural beliefs for affective attitude), (3) who would approve or disapprove of this practice (normative beliefs), (4) elements that could facilitate and hinder behavioural adoption (facilitating factors and barriers) and (5) elements that could facilitate and hinder behavioural adoption given one’s professional role (professional norm). The clinical vignettes used in the questionnaires were developed with the assistance of a nurse and a physician with many years of experience in caring for end-of-life patients. Nurses and physicians working in palliative care were also recruited to (1) ensure that the clinical vignettes adhered to clinical reality (two nurses and two physicians), (2) ensure that the different clinical vignettes were well counterbalanced (four nurses and four physicians) and (3) approve the preliminary version of the questionnaire (five nurses and four physicians). The psychometric qualities of the questionnaire were verified by means of a test–retest study. A total of 17 nurses completed the entire questionnaire two times at a 2-week interval. The questionnaire had good internal consistency with all alpha coefficients above 0.70 (range: 0.84–0.93).21 It also had good temporal stability with all intra-class coefficients above 0.71, except for facilitating factors whose coefficient was moderate (0.57).22
Data collection Data were collected by means of an anonymous self-administered questionnaire sent by mail (see Appendix 1 in the online version of the journal). All questionnaires were sent in mid-November 2012 with a personalised letter presenting the project, a fact sheet and with a preaddressed prepaid envelope. A first reminder was sent 1 week after the questionnaire was mailed, and a second reminder the following week (i.e. 2 weeks after questionnaire mailing). Two versions of questionnaires were used. Half of the sample got one version and the other half, the other. They only differed by the clinical vignette presenting a fictional patient before the items assessing the psychosocial variables (see Table 1). In the first version (A), the patient had made several explicit requests for euthanasia to the healthcare team (patient’s wishes known), while in the second version (B), the patient never clearly expressed his wishes concerning the practice of an act of euthanasia (patient’s wishes unknown). This allowed us to verify whether knowing patients’ position regarding euthanasia contributes to the prediction of nurses’ intention to practise euthanasia. This also represents an indirect assessment of the inclination of nurses to respect patients’ autonomy. The questionnaire contained 71 items, and its completion required between 15 and 20 min. The following definition of euthanasia was provided on the cover of the questionnaire: ‘an act which consists in intentionally causing the death of a person with an incurable disease’. The following definition of palliative care was also provided on the cover of the questionnaire: an approach to care for people who are living with a life-threatening illness, regardless of their age. The focus of care is on achieving comfort and ensuring dignity for the person and maximising quality of life for the patient, family and loved ones.
Participants were instructed to answer the questions by referring to the clinical vignette as if they were responsible for a case similar to the one described (i.e. Mr Brown). They were also reminded every other page that the questions refer to a context in which the practice of euthanasia would be legally accepted. All cognitive items were measured with a 7-point Likert-type scale (strongly/somewhat/slightly disagree, neither disagree nor agree, slightly/somewhat/strongly agree), except cognitive and affective attitude which were measured with 7-point semantic differential scales (e.g. very/somewhat/slightly inappropriate, neither one, slightly/somewhat/very appropriate). 4
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
5
Table 1. Clinical vignette used in the nurses’ questionnaires. Mary is Mr Brown’s nurse. He is 70 years old, married and a father of two. He suffers from cancer that is now generalised. Chemotherapy and radiotherapy treatments failed to stop the progression of the disease. He willingly accepted 3/3 level comfort care, which means the cessation of any curative or life-prolonging care, with the provision of palliative care. Mr Brown has many pulmonary, ganglionic and bone lesions that are very painful and partially responsive to analgesic treatment. He can barely hydrate and feed himself with protein shakes and his general state is very poor. He is now severely cachectic. His life expectancy is probably less than 10 days. While discussing with the attending physician, Mary notices that Mr Brown moans constantly, despite all efforts to relieve him. All possible therapeutic trials to control his pain have proven ineffective or caused intolerable side effects. The case of Mr Brown has already been discussed among the multidisciplinary team and with the family. Another physician confirmed the seriousness and irreversibility of his health status and the unappeasable state of his suffering. The option of sedation was also discussed, but Mr Brown rejected this alternative. All this information is recorded in his medical file. His condition thus raises the possibility of practising an act of euthanasia. It is important to note that this would be a legal act since the practice of euthanasia in an end-of-life context would have been legalised recently in Canada. At Mr Brown’s bedside, Mary and the physician realise that his speech is incoherent and he can no longer assume an active role in decisions concerning his care. However, it was clearly established that Mr Brown was apt during previous discussions concerning the possibility of cutting short his life by an act of euthanasia. Version A (patient’s wishes known): During those meetings, Mr Brown made several explicit requests for euthanasia to the healthcare team. Version B (patient’s wishes unknown): During those meetings, Mr Brown never clearly expressed his wishes concerning the practice of an act of euthanasia.
Statistical analyses Determinants of intention were identified by means of multiple logistic regressions. This statistical approach was chosen given that intention was non-normally distributed (U-shape). Intention was dichotomised at the median value (5.33). In order to choose which variables to enter into the prediction model, univariate analyses were carried out on all the cognitive, socio-demographic and contextual variables; only the variables with a p < 0.15 were included into the regression models. Of all the variables entered, only those reaching statistical significance (p < 0.05) were kept in the prediction model. All variables were always entered into the regression models in the following order: (1) direct variables of the TPB (cognitive attitude, affective attitude, subjective norm and PBC), (2) variables added to the TPB (moral norm, beneficence, justice and professional norm) and (3) socio-demographic and contextual variables. Multiple logistic regressions were also used to identify beliefs discriminating nurses according to their level of intention. All statistical analyses were performed using Statistical Analysis System (SAS) version 9.3 (SAS Institute, Inc., Cary, NC, USA).
Results Sample characteristics The overall response rate was 44.2%, which is comparable to a similar study among nurses.16 The respondents were similar to all the nurses of the province of Quebec, Canada, in terms of gender distribution 5
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
6
Nursing Ethics
Questionnaires mailed (n = 445)
Wrong address, now working in mental health (n = 6)
Nurses (n = 439)
Questionnaires completed (n = 194)
Nurses not meeting eligibility criteria (n = 40)
Questionnaires from eligible nurses (n = 154)
Questionnaire with > 50% of data missing (n = 1)
Questionnaires used in the analyses (n = 153) Version A (n = 80)
Version B (n = 73)
Figure 1. Flow of participants.
(% female: 89.5 vs 90.0; w2 ¼ 0.05, p ¼ 0.83), but they were slightly younger (40.8 + 11.4 vs 43.1 + 11.8; t ¼ 2.36, p ¼ 0.0181). The main reason why nurses did not meet eligibility criteria (40 out of 194) was because they did not care for end-of-life patients in the previous year. One nurse was excluded from the analyses since more than 50% of the questionnaire’s items were left empty. So, a total of 153 questionnaires were analysed. The flow of participants throughout the study is presented in Figure 1. Nurses’ most common domains of practice were geriatrics (30.5%), emergency (13.9%), intensive care (9.9%), housekeeping and home care (9.9%), internal medicine (7.3%), oncology 6
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
7
Table 2. Socio-demographic characteristics of nurses (n ¼ 153). Variables
Mean (SD)/%
Job position Nurse Other nurses (nurse clinician, etc.) Cared for end-of-life patients Number Percentage of practice Relatives received palliative care before death Yes No Years of experience 10 years Workplace Hospital Other (long-term care home, family medicine, etc.) Age Gender Male Female Level of education Collegial studies diploma University diploma Religious affiliation Yes No
55.92% 44.08% 22.32 (34.10) 21.22% 51.97% 48.03% 52.63% 47.37% 51.97% 48.03% 40.84 (11.35) 10.53% 89.47% 63.82% 36.18% 65.13% 34.87%
SD: standard deviation.
(4.6%) and surgery (4.6%). Complete socio-demographic characteristics of the participants are presented in Table 2.
Determinants of intention to practise euthanasia in palliative care Nurses were favourably predisposed to practise euthanasia in palliative care. The mean score for intention was 4.61 + 1.90 (possible scores ranging from 1 to 7). A total of 103 nurses had a positive intention (score >4), 44 nurses a negative intention (score 4), except for PBC (3.96 + 1.81) and justice (3.14 + 1.71). Univariate analyses indicated that all cognitions were significantly related to intention (all ps < 0.001). Among the socio-demographic and contextual variables, job position (odds ratio (OR) ¼ 0.68; 95% confidence interval (CI): 0.49–0.93, p ¼ 0.0164), number of end-of-life patients (OR ¼ 0.97; 95% CI: 0.95–0.99, p ¼ 0.0062), percentage of practice concerned with caring for end-of-life patients (OR ¼ 0.98; 95% CI: 0.96–0.99, p ¼ 0.0060), age (OR ¼ 0.97; 95% CI: 0.94–1.00, p ¼ 0.0410) and attitude towards the legalisation of euthanasia in Canada (OR ¼ 2.83; 95% CI: 1.90–4.21, p < 0.0001) were significantly related to intention, whereas years of experience (OR ¼ 0.83; 95% CI: 0.68–1.02, p ¼ 0.0712) and religious affiliation (OR ¼ 1.90; 95% CI: 0.96–3.78, p ¼ 0.0664) were not significant. All the other socio7
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
8
Nursing Ethics
Table 3. Logistic regression models for intention to practise euthanasia in palliative care among nurses (n ¼ 153). Models’ odds ratio (95% confidence interval) 1 Cognitive attitude Affective attitude Subjective norm Perceived behavioural control Moral norm Beneficence Justice Professional norm Job position Number of end-of-life patients Percentage of end-of-life patients Years of experience Age Religious affiliation Attitude towards legalisation Index of concordance (%)
0.90 (0.28–2.87) 1.10 (0.51–2.35) 5.66 (1.57–20.45) 1.06 (0.47–2.40) 3.45 (1.14–10.44) 0.97 (0.38–2.48) 1.27 (0.82–1.96) 0.90 (0.47–1.71) 0.87 (0.48–1.58) 0.97 (0.92–1.02) 0.98 (0.95–1.01) 0.70 (0.31–1.60) 1.07 (0.95–1.19) 0.62 (0.13–2.92) 1.27 (0.57–2.81) 95.1
2
3.08 (1.50–6.35) 2.95 (1.58–5.49)
92.6
demographic and contextual variables were clearly unrelated to intention (all ps > 0.15). The version of questionnaire (A vs B) was also assessed in univariate analyses to test the potential effect of knowing or not the patient’s wishes, and it was unrelated to intention (p ¼ 0.4908). In fact, the mean scores for intention did not significantly differ between the questionnaires (version A: 4.81 + 1.80 vs version B: 4.39 + 1.98, t ¼ 1.36, p ¼ 0.1758), meaning similar percentages of nurses had a positive intention (version A: 71.25%, version B: 63.01%), a negative intention (version A: 25.00%, version B: 32.88%) and a neutral intention (version A: 3.75%, version B: 4.11%). The final model (see Model 2 in Table 3).i shows that subjective norm (i.e. the perceived social pressure to practise euthanasia) (p ¼ 0.0023) and moral norm (i.e. the appropriateness of practising euthanasia according to one’s personal and moral values) (p ¼ 0.0007) were significant determinants of intention to practise euthanasia in palliative care among nurses. The model correctly classified 92.6% of nurses, which is considered excellent.
Structure of beliefs associated with level of intention The following two normative beliefs discriminated nurses according to their level of intention: My family (spouse, father, mother, etc.) (OR ¼ 3.28; 95% CI: 2.00–5.39)/The physicians with whom I work (OR ¼ 1.73; 95% CI: 1.29–2.34) would accept that I practise an act of euthanasia in a case similar to Mr Brown’s.
Nurses with a low level of intention had significantly lower scores on both beliefs compared to those with a high level of intention (my family: 3.49 + 1.96 vs 6.04 + 0.76, t ¼ 10.84, p < 0.0001; the physicians with whom I work: 3.03 + 1.76 vs 5.29 + 1.41, t ¼ 8.84, p < 0.0001). i
The results were similar when controlling for age (data not shown). In fact, age was poorly correlated to intention (r ¼ 0.21).
8
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
9
Given that moral norm was a determinant of intention, its underlying moral principles (beneficence and justice) were investigated. Two beneficence items discriminated nurses according to their level of intention. The beliefs were I would consider the following elements before practising an act of euthanasia in a case similar to Mr Brown’s. . . . (1) ‘the fact that this would help someone (OR ¼ 1.35; 95% CI: 1.02–1.77)’ and (2) ‘the fact that this would be acting in the best interest of an end-of-life person (OR ¼ 2.26; 95% CI: 1.44–3.53)’.
Nurses with a low intention had significantly lower scores on both items compared to those with a high level of intention (belief (1): 3.77 + 1.95 vs 5.77 + 1.44, t ¼ 7.24, p < 0.0001; belief (2): 4.49 + 2.01 vs 6.44 + 0.92, t ¼ 7.88, p < 0.0001). No justice items significantly discriminated nurses according to their level of intention (all ps > 0.05).
Discussion The results of this study suggest that most nurses have a positive intention to practise euthanasia in palliative care if such practice were legal. They also confirm the efficacy of using an extended version of the TPB to predict intention to adopt this behaviour. The prediction model included not only variables from the TPB but also variables added to the theory. Moreover, the results provide further evidence in the domain of health that moral norm is defined by the ethical principle of beneficence. Finally, it is rather surprising that nurses’ intention does not significantly vary according to the patient’s wishes concerning euthanasia. These findings are addressed below. Few studies have assessed nurses’ intention to practise euthanasia and not only their attitude towards the legalisation or whether they support this act. In this study, nurses had, overall, a positive intention to perform euthanasia, with a mean score slightly on the positive side (4.61 + 1.90). In fact, more than half of nurses (67.3%) had a positive intention (M ¼ 5.78 + 0.70) and only 28.8% had a negative intention (M ¼ 1.95 + 0.98), indicating polarised positions on the topic and that a majority of nurses had the intention to practise euthanasia. This result is similar to a study of Kuhse and Singer23 whereby 65% of Australian nurses indicated they would be willing to be involved in the process of euthanasia if it were legal and 35% indicated they would not. A first variable that significantly contributed to the prediction of nurses’ intention was the subjective norm. This result indicates that nurses would be inclined to practise an act of euthanasia because people who are the most important to them would accept, agree or think they should do so, specially their family and the physicians with whom they work. This might be related to the growing wind of support of the population towards the practice of euthanasia and the possible pressure this is exercising on healthcare providers. But more specifically, although a request for euthanasia remains the responsibility of physicians – as it is specified, for example, in the Belgian Act on Euthanasia (2002)24 – nurses’ sense of obligation with regard to the prescription of drugs cannot be disregarded.25,26 As Young and Ogden26 point out, ‘Ironically, although physicians prescribe many treatments, it is nurses who are often called upon to perform the act of administering intravenous and oral medications’ (p. 517). And in fact, it has been reported that cases of euthanasia are occasionally being performed by nurses in The Netherlands, although they do not have the legal authorisation to do so.25,27,28 For example, a recent study conducted in Belgium29 indicated that 12% of nurses do administer the lethal drugs prescribed by physicians in cases of euthanasia and even in cases where no explicit request was addressed by the patient. Nevertheless, since our larger study indicated that physicians were not so inclined to favour the practice of euthanasia (mean score for intention was 3.94 + 2.17), it calls for caution: healthcare providers should be cautious not to presume the intention of any caregivers. 9
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
10
Nursing Ethics
A second variable, added to the TPB, also significantly contributed to the prediction of nurses’ intention, which is the moral norm. While this variable has already been associated with healthcare workers’13,15 and nurses’18,19 motivation in previous studies, it is the first time that it is linked to euthanasia. Another new aspect is relating nurses’ moral norm to the ethical principle of beneficence. This result replicates those of a previous study on consent to organ donation among the general population.20 Three ethical principles (autonomy, beneficence and justice) were then tested as potential principles underlying moral norm and only beneficence was significantly related to this construct. Thus, this means that nurses will be motivated to practise euthanasia if they perceive that this act is in agreement with their personal values and principles, and also if they believe that this would help someone and be in the best interest of an end-of-life person. Surprisingly and contrary to previous reviews among nurses,3–5 the ethical principle of the patient’s autonomy did not have a significant impact on nurses’ intention to practise euthanasia. In fact, 63.01% of our respondents had even a positive intention to practise euthanasia although they did not know the patient’s wishes (or had his consent). This is rather troublesome but, in fact, not uncommon. Again, the study conducted by Inghelbrecht et al.29 showed that ‘administration of the life-ending drugs by the nurse occurred more frequently in the cases without an explicit request from the patient than in the cases of euthanasia’ (p. 908) where a request was explicit. How should this be understood since this is clearly against ethical code and present regulations where euthanasia is legalised? It has been reported that sometimes healthcare providers are faced with conflicts between competing ethical principles, such as autonomy and beneficence.30,31 In fact, according to Beauchamp and Childress,31 ‘whether respect for the autonomy of patients should have priority over professional beneficence directed at those patients is a central problem in biomedical ethics’ (p. 207). As such, in one vignette, these two ethical principles were well aligned, whereas in the other, there was a matter of debate. In this latter context, nurses were faced with making sure of respecting the patient’s autonomy, which was unknown, or relieving that person of refractory pain (beneficence). Moreover, given that nurses are often portrayed as patient advocates who want to help their patients,5 it is possible that for some of them the principle of beneficence might have overridden the absence of known wishes. Unlike previous reviews on nurses’ attitudes towards euthanasia, no socio-demographic and contextual variables had an impact on nurses’ intention in this study. In particular, religious affiliation did not contribute to nurses’ intention, while in the past, it has systematically been associated with a negative attitude towards euthanasia.3–5,32 Contrary to reviews among nurses,3,4 age and nursing specialty were also not determinants of intention. Finally, in this study, nurses’ attitude towards the legalisation of euthanasia in Canada was another variable unrelated to intention. While these results may at first seem to run counter to common sense, they support the assumptions of the TPB. According to this theory, sociodemographic variables are external factors whose influence on intention should be filtered (or mediated) through the psychosocial constructs, such as subjective norm, attitude and PBC; they should not be direct determinants of intention. These results thus provide additional support for the use of the TPB to identify nurses’ intention to practise euthanasia in palliative care. The results of this study further add to the increasing evidence on the efficacy of using an extended version of the TPB to study health professionals’ intention and behaviours. Already in 2008, a systematic review had identified the TPB as a promising theory to predict their behaviours.13 The novelty is the application of this theory to study nurses’ intention to practise euthanasia. In fact, most research in the field of palliative care is atheoretical. Yet, increasingly more authors are calling for more theory-based studies, given that the use of theories presents many advantages.33–36 For instance, theory can guide the selection of determinants to be tested as potential predictors of intention or behaviour.37 The use of theory can also facilitate replication of previous findings which is crucial to increasing scientific knowledge about certain behaviours.38 Additionally, this can provide a basis for refining and developing better theories.37 10
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
11
Strengths of this study are the novelty of using a psychosocial theory to study nurses’ intention to practise euthanasia in palliative care and the rigorous methodology used to develop and validate the survey instrument. In fact, the main methodological issues identified in two reviews on euthanasia,6,39 such as sample representativeness, reporting of the psychometric qualities of the items used in the questionnaires and use of multivariate statistical analyses, have been addressed. One limitation of the study is the low response rate. Yet, according to some authors, low response rates are more acceptable when the topic is controversial, such as in the case of euthanasia.40,41 Respondents were also compared to the population under study to ascertain they were similar in terms of gender distribution and age. Even though respondents were younger compared to nurses of the province of Quebec, Canada, the results were similar when controlling for age in the analyses. In conclusion, to our knowledge, this study is the first to report the application of a validated psychosocial theory to predict nurses’ intention to practise euthanasia in palliative care. It also has the distinction of identifying the ethical principles underlying nurses’ moral norm and intention. It is hoped that this will encourage more theoretical research on nurses’ motivation to practise euthanasia and on the ethical principles related to this practice. Conflict of interest The authors declare that there is no conflict of interest. Funding This study was supported by a grant from the Ethics Office of the Canadian Institutes of Health Research [EOG - 11392]. References 1. Gordijn B and Janssens R. The prevention of euthanasia through palliative care: new developments in The Netherlands. Patient Educ Couns 2000; 41(1): 35–46 (PMID: 10900365). 2. Pereira J, Anwar D, Pralong G, et al. Assisted suicide and euthanasia should not be practiced in palliative care units. J Palliat Med 2008; 11(8): 1074–1076 (PMID: 18980445). 3. Verpoort C, Gastmans C, De Bal N, et al. Nurses’ attitudes to euthanasia: a review of the literature. Nurs Ethics 2004; 11(4): 349–365 (PMID: 15253571). 4. Berghs M, Dierckx de Casterle B and Gastmans C. The complexity of nurses’ attitudes toward euthanasia: a review of the literature. J Med Ethics 2005; 31(8): 441–446 (PMID: 16076966). 5. Quaghebeur T, Dierckx de Casterle B and Gastmans C. Nursing and euthanasia: a review of argument-based ethics literature. Nurs Ethics 2009; 16(4): 466–486 (PMID: 19528103). 6. Vezina-Im LA, Lavoie M, Krol P, et al. Motivations of physicians and nurses to practice voluntary euthanasia: a systematic review. BMC Palliat Care 2014; 13(1): 20 (PMID: 24716567; PMCID: 4021095). 7. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process 1991; 50: 179–211. 8. Armitage CJ and Conner M. Efficacy of the theory of planned behaviour: a meta-analytic review. Br J Soc Psychol 2001; 40(4): 471–479. 9. Conner M and Sparks P. The theory of planned behaviour and health behaviour. In: Conner M and Norman P (eds) Predicting health behaviour. 2nd ed. Maidenhead: Open University Press, 2005, pp. 170–222. 10. Godin G and Kok G. The theory of planned behavior: a review of its applications to health-related behaviors. Am J Health Promot 1996; 11(2): 87–98. 11. Webb TL and Sheeran P. Does changing behavioral intentions engender behavior changes? A meta-analysis of the experimental evidence. Psychol Bull 2006; 132(2): 249–268. 12. McEachan R, Conner M, Taylor NJ, et al. Prospective prediction of health-related behaviours with the theory of planned behaviour: a meta-analysis. Health Psychol Rev 2011; 5(2): 97–144. 11
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
12
Nursing Ethics
13. Godin G, Belanger-Gravel A, Eccles M, et al. Healthcare professionals’ intentions and behaviours: a systematic review of studies based on social cognitive theories. Implement Sci 2008; 3: 36 (PMID: 18631386). 14. Triandis HC. Values, attitudes and interpersonal behavior. In: Page MM (ed.) Nebraska symposium on motivation: beliefs, attitudes and values, Vol. 1. Lincoln, NE: University of Nebraska Press, 1980, pp. 195–259. 15. Godin G, Vezina-Im LA and Naccache H. Determinants of influenza vaccination among healthcare workers. Infect Control Hosp Epidemiol 2010; 31(7): 689–693 (PMID: 20482373). 16. Daneault S, Beaudry M and Godin G. Psychosocial determinants of the intention of nurses and dietitians to recommend breastfeeding. Can J Public Health 2004; 95(2): 151–154 (PMID: 15074909). 17. Godin G, Beaulieu D, Touchette JS, et al. Intention to encourage complementary and alternative medicine among general practitioners and medical students. Behav Med 2007; 33(2): 67–77 (PMID: 17711808). 18. Cote F, Gagnon J, Houme PK, et al. Using the theory of planned behaviour to predict nurses’ intention to integrate research evidence into clinical decision-making. J Adv Nurs 2012; 68(10): 2289–2298 (PMID: 22229522). 19. Chabot G, Godin G and Gagnon MP. Determinants of the intention of elementary school nurses to adopt a redefined role in health promotion at school. Implement Sci 2010; 5: 93 (PMID: 21110872). 20. Blondeau D, Godin G, Gagne´ C, et al. Do ethical principles explain moral norm? A test for consent to organ donation. J Appl Biobehav Res 2004; 9(4): 230–243. 21. Nunnally JC. Psychometric theory. 2nd ed. New York: McGraw-Hill, 1978. 22. Fermanian J. Mesure de l’accord entre deux juges: cas quantitatif [Measuring agreement between 2 observers: A quantitative case]. Rev E´pide´miol Sante´ Publique 1984; 32(6): 408–413 (PMID: 6531504). 23. Kuhse H and Singer P. Voluntary euthanasia and the nurse: an Australian survey. Int J Nurs Stud 1993; 30(4): 311–322 (PMID: 8375974). 24. Ministry of Justice. Wet betreffende de euthanasie [Law on euthanasia] (The Belgian Act on Euthanasia of 28 May 2002) 24 August 2010, www.ejustice.just.fgov.be/loi/loi.htm 25. Bilsen JJ, Vander Stichele RH, Mortier F, et al. Involvement of nurses in physician-assisted dying. J Adv Nurs 2004; 47(6): 583–591 (PMID: 15324424). 26. Young MG and Ogden RD. The role of nurses in AIDS care regarding voluntary euthanasia and assisted suicide: a call for further dialogue. J Adv Nurs 2000; 31(3): 513–519 (PMID: 10718869). 27. Muller MT, Pijnenborg L, Onwuteaka-Philipsen BD, et al. The role of the nurse in active euthanasia and physician-assisted suicide. J Adv Nurs 1997; 26(2): 424–430 (PMID: 9292379). 28. Van Bruchem-van de Scheur A, van der Arend A, van Wijmen F, et al. Dutch nurses’ attitudes towards euthanasia and physician-assisted suicide. Nurs Ethics 2008; 15(2): 186–198 (PMID: 18272609). 29. Inghelbrecht E, Bilsen J, Mortier F, et al. The role of nurses in physician-assisted deaths in Belgium. CMAJ 2010; 182(9): 905–910 (PMID: 20479043; PMCID: 2882450). 30. Badger JM, Ladd RE and Adler P. Respecting patient autonomy versus protecting the patient’s health: a dilemma for healthcare providers. JONA’S Healthc Law Ethics Regul 2009; 11(4): 120–124; quiz 5–6 (PMID: 19935216). 31. Beauchamp TL and Childress JF. Principles of biomedical ethics. 6th ed. New York: Oxford University Press, 2009. 32. Emanuel EJ. Euthanasia and physician-assisted suicide: a review of the empirical data from the United States. Arch Intern Med 2002; 162(2): 142–152 (PMID: 11802747). 33. Baranowski T, Lin LS, Wetter DW, et al. Theory as mediating variables: why aren’t community interventions working as desired? Ann Epidemiol 1997; 7: S89–S95. 34. Fishbein M and Yzer MC. Using theory to design effective health behavior interventions. Commun Theor 2003; 13: 164–183. 35. Michie S and Abraham C. Interventions to change health behaviours: evidence-based or evidence-inspired? Psychol Health 2004; 19: 29–49. 36. Glanz K and Bishop DB. The role of behavioral science theory in development and implementation of public health interventions. Annu Rev Public Health 2010; 31: 399–418 (PMID: 20070207). 12
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
13
37. Michie S and Prestwich A. Are interventions theory-based? Development of a theory coding scheme. Health Psychol 2010; 29(1): 1–8 (PMID: 20063930). 38. Michie S, Fixsen D, Grimshaw JM, et al. Specifying and reporting complex behaviour change interventions: the need for a scientific method. Implement Sci 2009; 4: 40 (PMID: 19607700). 39. Rosenfeld B. Methodological issues in assisted suicide and euthanasia research. Psychol Public Policy Law 2000; 6(2): 559–574 (PMID: 12661541). 40. Sierles FS. How to do research with self-administered surveys. Acad Psychiatry 2003; 27(2): 104–113 (PMID: 12824111). 41. Rubenfeld GD. Surveys: an introduction. Respir Care 2004; 49(10): 1181–1185 (PMID: 15447800).
13
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Psychosocial determinants of nurses’ intention to practise euthanasia in palliative care
Nursing Ethics 1–13 ª The Author(s) 2014 Reprints and permission: sagepub.co.uk/journalsPermissions.nav 10.1177/0969733014557117 nej.sagepub.com
Mireille Lavoie Laval University, Canada; Centre de recherche du CHU de Que´bec, Canada
Gaston Godin, Lydi-Anne Ve´zina-Im and Danielle Blondeauy Laval University, Canada
Isabelle Martineau Maison Michel-Sarrazin, Canada
Louis Roy Hoˆpital Enfant-Je´sus, Canada
Abstract Background: Most studies on euthanasia fail to explain the intentions of health professionals when faced with performing euthanasia and are atheoretical. Research objective: The purpose of this study was to identify the psychosocial determinants of nurses’ intention to practise euthanasia in palliative care if it were legalised. Research design: A cross-sectional study using a validated anonymous questionnaire based on an extended version of the Theory of Planned Behaviour. Participants and research context: A random sample of 445 nurses from the province of Quebec, Canada, was selected for participation in the study. Ethical considerations: The study was reviewed and approved by the Ethics Committee of the Centre hospitalier universitaire de Que´bec. Findings: The response rate was 44.2% and the mean score for intention was 4.61 + 1.90 (range: 1–7). The determinants of intention were the subjective (odds ratio ¼ 3.08; 95% confidence interval: 1.50– 6.35) and moral (odds ratio ¼ 2.95; 95% confidence interval: 1.58–5.49) norms. Specific beliefs which could discriminate nurses according to their level of intention were identified. Discussion: Overall, nurses have a slightly positive intention to practise euthanasia. Their family approval seems particularly important and also the approval of their medical colleagues. Nurses’ moral norm was related to beneficence, an ethical principle. Conclusion: To our knowledge, this is the first study to identify nurses’ motivations to practise euthanasia in palliative care using a validated psychosocial theory. It also has the distinction of identifying the ethical principles underlying nurses’ moral norm and intention.
y
D.B. was co-researcher but died in the process of revising the article.
Corresponding author: Mireille Lavoie, Faculty of Nursing, Laval University, 1050 avenue de Me´decine, Que´bec City, QC G1V 0A6, Canada. Email: [email protected]
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
2
Nursing Ethics
Keywords Determinant, euthanasia, intention, nurse, palliative care
Background Euthanasia (from the Greek eu (good) and thanatos (death)) is sometimes put forward as an alternative to palliative care.1 It must be recognised that despite great progress in palliative care, caregivers are still confronted with the physical, psychological or spiritual limitations of their power to alleviate suffering. In some cases, it is impossible to provide suitable treatment for patients whose suffering may be seen as ‘unnecessary’ or even ‘inhuman’. Yet, this practice is also held to be at odds with the philosophy of palliative care,2 which aims to relieve suffering while accompanying patients until the end of their lives. It is thus not surprising, even though euthanasia is illegal in Canada, that it remains a controversial topic and generates far from unanimous support among healthcare professionals. According to previous reviews of studies on attitude towards euthanasia among nurses, religion, age and nursing specialty significantly affect their views on euthanasia.3,4 Moreover, arguments for or against euthanasia among nurses would be predominantly guided by the following four ethical principles: respect for autonomy, non-maleficence, beneficence and justice.5 Another review also added quality of life, sanctity of life and the slippery slope as other important principles underlying nurses’ arguments for or against euthanasia.4 Nurses’ main arguments for supporting euthanasia appear to be the patient’s right to die with dignity, the wish to relieve the patient of unbearable pain (beneficence) and the patient’s right to decide about his/her life and death (autonomy).3 On the other hand, nurses’ main arguments for being against euthanasia are the following: that primary care should be about alleviating pain not ending life (non-maleficence, sanctity of life), their confidence in palliative care, religious and moral objections, and a concern for potential abuse such as performing euthanasia without patients’ consent.3 Most studies on euthanasia have explored the opinion of health professionals regarding its legalisation or their experience when this act is requested. However, they failed to explain the intentions of caregivers faced with performing an act of euthanasia, especially as most studies are also atheoretical.6 Therefore, the purpose of this study is to identify the psychosocial determinants of nurses’ intention to practise euthanasia in palliative care if it were legal in order to enlighten decision-makers regarding the impact of this practice on care practices and on the health professionals themselves.
Theoretical framework The study was guided by an extended version of the Theory of Planned Behaviour (TPB).7 Its efficacy in predicting intentions to adopt various health behaviours, including among health professionals, and the key role of intentions to predict behaviours have already been clearly established in a number of metaanalyses.8–13 According to the TPB, behaviour is predicted by intention and by perceived behavioural control (PBC) when the context is less volitional. Intention, in return, is formed of the following three constructs: attitude, subjective norm and PBC. Attitude is an evaluation, either positive or negative, of the adoption of a given behaviour. In this study, attitudes were evaluated by using its two components, cognitive and affective attitude as suggested by Triandis.14 Cognitive attitude refers to factual consequences (e.g. useless/useful) of the adoption of a given behaviour, while affective attitude is rather concerned with emotional consequences (e.g. unpleasant/pleasant). Subjective norm represents the perceived social pressure to adopt a given behaviour. PBC refers to people’s evaluation of their ability to adopt a given behaviour. Each construct is moreover related to a specific set of beliefs: attitude is related to behavioural beliefs, subjective norm to normative beliefs and PBC to control beliefs. Behavioural beliefs refer 2
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
3
to the perceived positive and negative consequences of the adoption of a given behaviour. Normative beliefs represent how individuals believe people who are important to them would react if they adopted a given behaviour (i.e. approve or disapprove). Control beliefs refer to elements that can hinder (barriers) or facilitate (facilitating factors) behavioural adoption. External factors such as socio-demographic variables (e.g. age, gender) can also influence the intention to adopt a given behaviour through the other constructs. New variables can be added to the TPB as long as they improve its predictive ability.7 In this study, two variables were added, professional norm and moral norm, given that there is evidence that they are determinants of health professionals’ intention to adopt various behaviours.15–19 Professional norm refers to the appropriateness of adopting a behaviour given one’s profession. Moral norm is a variable originating from the Theory of Interpersonal Behaviour.14 It is related to the appropriateness of adopting a given behaviour according to one’s personal and moral values. In another study on consent to organ donation, Blondeau et al.20 measured moral norm by using the following three ethical principles: autonomy, beneficence and justice. The results indicated that moral norm is mainly predicted by beneficence. In this study, moral norm, as well as beneficence, justice and autonomy, were then included. Concerning the concept of autonomy, the design of the study allowed us to assess its effect on intention directly since we used two versions of the questionnaire: in one version, the patient’s wishes concerning euthanasia were known, while in the other version, the patient’s wishes on this matter were unknown. In terms of external factors, the following socio-demographic and contextual variables were assessed: job position, main domain of practice, number of end-of-life patients nurses cared for in the past year and the percentage of their practice they represented, whether they have relatives who received palliative care before their death, years of experience, worksite, age, gender, level of education, religious affiliation and attitude towards the legalisation of euthanasia in Canada.
Methods Population and sample The population under study consisted of nurses from the province of Quebec, Canada. Head nurses were excluded from the study since they do not work at the bedside of patients. Nurses with underage patients (e.g. paediatrics) or patients with mental diseases (psychiatry) or whose job makes them unlikely to care for end-of-life patients (e.g. rehabilitation, plastic surgery) were also excluded. To obtain our sample, the provincial nurses’ professional Order of Quebec provided us with a list of their active members. A random sample of 445 nurses was obtained using random digit tables. The sample was weighted according to the domains of practice included in the study to reflect as closely as possible their distribution in the province. The study was reviewed and approved by the Ethics Committee of the Centre hospitalier universitaire (CHU) de Que´bec (reference number: HCS11-06-047).
Questionnaire development and validation The questionnaire was developed in accordance with the methodology recommended by the author of the TPB. A questionnaire based on this theory has to reflect the salient beliefs of a population. It is worth mentioning that in the original study, physicians’ psychosocial determinants were also identified (to be published), but only the results pertaining to nurses are reported in this article. To develop the belief items of the questionnaire, 29 health professionals (21 nurses and 8 physicians) sharing the same sociodemographic characteristics as those of the study sample completed a short open-ended questionnaire containing the following questions: (1) advantages and disadvantages of practicing euthanasia in palliative care 3
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
4
Nursing Ethics
(behavioural beliefs for cognitive attitude), (2) emotions that could encourage and hinder practicing euthanasia in palliative care (behavioural beliefs for affective attitude), (3) who would approve or disapprove of this practice (normative beliefs), (4) elements that could facilitate and hinder behavioural adoption (facilitating factors and barriers) and (5) elements that could facilitate and hinder behavioural adoption given one’s professional role (professional norm). The clinical vignettes used in the questionnaires were developed with the assistance of a nurse and a physician with many years of experience in caring for end-of-life patients. Nurses and physicians working in palliative care were also recruited to (1) ensure that the clinical vignettes adhered to clinical reality (two nurses and two physicians), (2) ensure that the different clinical vignettes were well counterbalanced (four nurses and four physicians) and (3) approve the preliminary version of the questionnaire (five nurses and four physicians). The psychometric qualities of the questionnaire were verified by means of a test–retest study. A total of 17 nurses completed the entire questionnaire two times at a 2-week interval. The questionnaire had good internal consistency with all alpha coefficients above 0.70 (range: 0.84–0.93).21 It also had good temporal stability with all intra-class coefficients above 0.71, except for facilitating factors whose coefficient was moderate (0.57).22
Data collection Data were collected by means of an anonymous self-administered questionnaire sent by mail (see Appendix 1 in the online version of the journal). All questionnaires were sent in mid-November 2012 with a personalised letter presenting the project, a fact sheet and with a preaddressed prepaid envelope. A first reminder was sent 1 week after the questionnaire was mailed, and a second reminder the following week (i.e. 2 weeks after questionnaire mailing). Two versions of questionnaires were used. Half of the sample got one version and the other half, the other. They only differed by the clinical vignette presenting a fictional patient before the items assessing the psychosocial variables (see Table 1). In the first version (A), the patient had made several explicit requests for euthanasia to the healthcare team (patient’s wishes known), while in the second version (B), the patient never clearly expressed his wishes concerning the practice of an act of euthanasia (patient’s wishes unknown). This allowed us to verify whether knowing patients’ position regarding euthanasia contributes to the prediction of nurses’ intention to practise euthanasia. This also represents an indirect assessment of the inclination of nurses to respect patients’ autonomy. The questionnaire contained 71 items, and its completion required between 15 and 20 min. The following definition of euthanasia was provided on the cover of the questionnaire: ‘an act which consists in intentionally causing the death of a person with an incurable disease’. The following definition of palliative care was also provided on the cover of the questionnaire: an approach to care for people who are living with a life-threatening illness, regardless of their age. The focus of care is on achieving comfort and ensuring dignity for the person and maximising quality of life for the patient, family and loved ones.
Participants were instructed to answer the questions by referring to the clinical vignette as if they were responsible for a case similar to the one described (i.e. Mr Brown). They were also reminded every other page that the questions refer to a context in which the practice of euthanasia would be legally accepted. All cognitive items were measured with a 7-point Likert-type scale (strongly/somewhat/slightly disagree, neither disagree nor agree, slightly/somewhat/strongly agree), except cognitive and affective attitude which were measured with 7-point semantic differential scales (e.g. very/somewhat/slightly inappropriate, neither one, slightly/somewhat/very appropriate). 4
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
5
Table 1. Clinical vignette used in the nurses’ questionnaires. Mary is Mr Brown’s nurse. He is 70 years old, married and a father of two. He suffers from cancer that is now generalised. Chemotherapy and radiotherapy treatments failed to stop the progression of the disease. He willingly accepted 3/3 level comfort care, which means the cessation of any curative or life-prolonging care, with the provision of palliative care. Mr Brown has many pulmonary, ganglionic and bone lesions that are very painful and partially responsive to analgesic treatment. He can barely hydrate and feed himself with protein shakes and his general state is very poor. He is now severely cachectic. His life expectancy is probably less than 10 days. While discussing with the attending physician, Mary notices that Mr Brown moans constantly, despite all efforts to relieve him. All possible therapeutic trials to control his pain have proven ineffective or caused intolerable side effects. The case of Mr Brown has already been discussed among the multidisciplinary team and with the family. Another physician confirmed the seriousness and irreversibility of his health status and the unappeasable state of his suffering. The option of sedation was also discussed, but Mr Brown rejected this alternative. All this information is recorded in his medical file. His condition thus raises the possibility of practising an act of euthanasia. It is important to note that this would be a legal act since the practice of euthanasia in an end-of-life context would have been legalised recently in Canada. At Mr Brown’s bedside, Mary and the physician realise that his speech is incoherent and he can no longer assume an active role in decisions concerning his care. However, it was clearly established that Mr Brown was apt during previous discussions concerning the possibility of cutting short his life by an act of euthanasia. Version A (patient’s wishes known): During those meetings, Mr Brown made several explicit requests for euthanasia to the healthcare team. Version B (patient’s wishes unknown): During those meetings, Mr Brown never clearly expressed his wishes concerning the practice of an act of euthanasia.
Statistical analyses Determinants of intention were identified by means of multiple logistic regressions. This statistical approach was chosen given that intention was non-normally distributed (U-shape). Intention was dichotomised at the median value (5.33). In order to choose which variables to enter into the prediction model, univariate analyses were carried out on all the cognitive, socio-demographic and contextual variables; only the variables with a p < 0.15 were included into the regression models. Of all the variables entered, only those reaching statistical significance (p < 0.05) were kept in the prediction model. All variables were always entered into the regression models in the following order: (1) direct variables of the TPB (cognitive attitude, affective attitude, subjective norm and PBC), (2) variables added to the TPB (moral norm, beneficence, justice and professional norm) and (3) socio-demographic and contextual variables. Multiple logistic regressions were also used to identify beliefs discriminating nurses according to their level of intention. All statistical analyses were performed using Statistical Analysis System (SAS) version 9.3 (SAS Institute, Inc., Cary, NC, USA).
Results Sample characteristics The overall response rate was 44.2%, which is comparable to a similar study among nurses.16 The respondents were similar to all the nurses of the province of Quebec, Canada, in terms of gender distribution 5
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
6
Nursing Ethics
Questionnaires mailed (n = 445)
Wrong address, now working in mental health (n = 6)
Nurses (n = 439)
Questionnaires completed (n = 194)
Nurses not meeting eligibility criteria (n = 40)
Questionnaires from eligible nurses (n = 154)
Questionnaire with > 50% of data missing (n = 1)
Questionnaires used in the analyses (n = 153) Version A (n = 80)
Version B (n = 73)
Figure 1. Flow of participants.
(% female: 89.5 vs 90.0; w2 ¼ 0.05, p ¼ 0.83), but they were slightly younger (40.8 + 11.4 vs 43.1 + 11.8; t ¼ 2.36, p ¼ 0.0181). The main reason why nurses did not meet eligibility criteria (40 out of 194) was because they did not care for end-of-life patients in the previous year. One nurse was excluded from the analyses since more than 50% of the questionnaire’s items were left empty. So, a total of 153 questionnaires were analysed. The flow of participants throughout the study is presented in Figure 1. Nurses’ most common domains of practice were geriatrics (30.5%), emergency (13.9%), intensive care (9.9%), housekeeping and home care (9.9%), internal medicine (7.3%), oncology 6
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
7
Table 2. Socio-demographic characteristics of nurses (n ¼ 153). Variables
Mean (SD)/%
Job position Nurse Other nurses (nurse clinician, etc.) Cared for end-of-life patients Number Percentage of practice Relatives received palliative care before death Yes No Years of experience 10 years Workplace Hospital Other (long-term care home, family medicine, etc.) Age Gender Male Female Level of education Collegial studies diploma University diploma Religious affiliation Yes No
55.92% 44.08% 22.32 (34.10) 21.22% 51.97% 48.03% 52.63% 47.37% 51.97% 48.03% 40.84 (11.35) 10.53% 89.47% 63.82% 36.18% 65.13% 34.87%
SD: standard deviation.
(4.6%) and surgery (4.6%). Complete socio-demographic characteristics of the participants are presented in Table 2.
Determinants of intention to practise euthanasia in palliative care Nurses were favourably predisposed to practise euthanasia in palliative care. The mean score for intention was 4.61 + 1.90 (possible scores ranging from 1 to 7). A total of 103 nurses had a positive intention (score >4), 44 nurses a negative intention (score 4), except for PBC (3.96 + 1.81) and justice (3.14 + 1.71). Univariate analyses indicated that all cognitions were significantly related to intention (all ps < 0.001). Among the socio-demographic and contextual variables, job position (odds ratio (OR) ¼ 0.68; 95% confidence interval (CI): 0.49–0.93, p ¼ 0.0164), number of end-of-life patients (OR ¼ 0.97; 95% CI: 0.95–0.99, p ¼ 0.0062), percentage of practice concerned with caring for end-of-life patients (OR ¼ 0.98; 95% CI: 0.96–0.99, p ¼ 0.0060), age (OR ¼ 0.97; 95% CI: 0.94–1.00, p ¼ 0.0410) and attitude towards the legalisation of euthanasia in Canada (OR ¼ 2.83; 95% CI: 1.90–4.21, p < 0.0001) were significantly related to intention, whereas years of experience (OR ¼ 0.83; 95% CI: 0.68–1.02, p ¼ 0.0712) and religious affiliation (OR ¼ 1.90; 95% CI: 0.96–3.78, p ¼ 0.0664) were not significant. All the other socio7
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
8
Nursing Ethics
Table 3. Logistic regression models for intention to practise euthanasia in palliative care among nurses (n ¼ 153). Models’ odds ratio (95% confidence interval) 1 Cognitive attitude Affective attitude Subjective norm Perceived behavioural control Moral norm Beneficence Justice Professional norm Job position Number of end-of-life patients Percentage of end-of-life patients Years of experience Age Religious affiliation Attitude towards legalisation Index of concordance (%)
0.90 (0.28–2.87) 1.10 (0.51–2.35) 5.66 (1.57–20.45) 1.06 (0.47–2.40) 3.45 (1.14–10.44) 0.97 (0.38–2.48) 1.27 (0.82–1.96) 0.90 (0.47–1.71) 0.87 (0.48–1.58) 0.97 (0.92–1.02) 0.98 (0.95–1.01) 0.70 (0.31–1.60) 1.07 (0.95–1.19) 0.62 (0.13–2.92) 1.27 (0.57–2.81) 95.1
2
3.08 (1.50–6.35) 2.95 (1.58–5.49)
92.6
demographic and contextual variables were clearly unrelated to intention (all ps > 0.15). The version of questionnaire (A vs B) was also assessed in univariate analyses to test the potential effect of knowing or not the patient’s wishes, and it was unrelated to intention (p ¼ 0.4908). In fact, the mean scores for intention did not significantly differ between the questionnaires (version A: 4.81 + 1.80 vs version B: 4.39 + 1.98, t ¼ 1.36, p ¼ 0.1758), meaning similar percentages of nurses had a positive intention (version A: 71.25%, version B: 63.01%), a negative intention (version A: 25.00%, version B: 32.88%) and a neutral intention (version A: 3.75%, version B: 4.11%). The final model (see Model 2 in Table 3).i shows that subjective norm (i.e. the perceived social pressure to practise euthanasia) (p ¼ 0.0023) and moral norm (i.e. the appropriateness of practising euthanasia according to one’s personal and moral values) (p ¼ 0.0007) were significant determinants of intention to practise euthanasia in palliative care among nurses. The model correctly classified 92.6% of nurses, which is considered excellent.
Structure of beliefs associated with level of intention The following two normative beliefs discriminated nurses according to their level of intention: My family (spouse, father, mother, etc.) (OR ¼ 3.28; 95% CI: 2.00–5.39)/The physicians with whom I work (OR ¼ 1.73; 95% CI: 1.29–2.34) would accept that I practise an act of euthanasia in a case similar to Mr Brown’s.
Nurses with a low level of intention had significantly lower scores on both beliefs compared to those with a high level of intention (my family: 3.49 + 1.96 vs 6.04 + 0.76, t ¼ 10.84, p < 0.0001; the physicians with whom I work: 3.03 + 1.76 vs 5.29 + 1.41, t ¼ 8.84, p < 0.0001). i
The results were similar when controlling for age (data not shown). In fact, age was poorly correlated to intention (r ¼ 0.21).
8
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
9
Given that moral norm was a determinant of intention, its underlying moral principles (beneficence and justice) were investigated. Two beneficence items discriminated nurses according to their level of intention. The beliefs were I would consider the following elements before practising an act of euthanasia in a case similar to Mr Brown’s. . . . (1) ‘the fact that this would help someone (OR ¼ 1.35; 95% CI: 1.02–1.77)’ and (2) ‘the fact that this would be acting in the best interest of an end-of-life person (OR ¼ 2.26; 95% CI: 1.44–3.53)’.
Nurses with a low intention had significantly lower scores on both items compared to those with a high level of intention (belief (1): 3.77 + 1.95 vs 5.77 + 1.44, t ¼ 7.24, p < 0.0001; belief (2): 4.49 + 2.01 vs 6.44 + 0.92, t ¼ 7.88, p < 0.0001). No justice items significantly discriminated nurses according to their level of intention (all ps > 0.05).
Discussion The results of this study suggest that most nurses have a positive intention to practise euthanasia in palliative care if such practice were legal. They also confirm the efficacy of using an extended version of the TPB to predict intention to adopt this behaviour. The prediction model included not only variables from the TPB but also variables added to the theory. Moreover, the results provide further evidence in the domain of health that moral norm is defined by the ethical principle of beneficence. Finally, it is rather surprising that nurses’ intention does not significantly vary according to the patient’s wishes concerning euthanasia. These findings are addressed below. Few studies have assessed nurses’ intention to practise euthanasia and not only their attitude towards the legalisation or whether they support this act. In this study, nurses had, overall, a positive intention to perform euthanasia, with a mean score slightly on the positive side (4.61 + 1.90). In fact, more than half of nurses (67.3%) had a positive intention (M ¼ 5.78 + 0.70) and only 28.8% had a negative intention (M ¼ 1.95 + 0.98), indicating polarised positions on the topic and that a majority of nurses had the intention to practise euthanasia. This result is similar to a study of Kuhse and Singer23 whereby 65% of Australian nurses indicated they would be willing to be involved in the process of euthanasia if it were legal and 35% indicated they would not. A first variable that significantly contributed to the prediction of nurses’ intention was the subjective norm. This result indicates that nurses would be inclined to practise an act of euthanasia because people who are the most important to them would accept, agree or think they should do so, specially their family and the physicians with whom they work. This might be related to the growing wind of support of the population towards the practice of euthanasia and the possible pressure this is exercising on healthcare providers. But more specifically, although a request for euthanasia remains the responsibility of physicians – as it is specified, for example, in the Belgian Act on Euthanasia (2002)24 – nurses’ sense of obligation with regard to the prescription of drugs cannot be disregarded.25,26 As Young and Ogden26 point out, ‘Ironically, although physicians prescribe many treatments, it is nurses who are often called upon to perform the act of administering intravenous and oral medications’ (p. 517). And in fact, it has been reported that cases of euthanasia are occasionally being performed by nurses in The Netherlands, although they do not have the legal authorisation to do so.25,27,28 For example, a recent study conducted in Belgium29 indicated that 12% of nurses do administer the lethal drugs prescribed by physicians in cases of euthanasia and even in cases where no explicit request was addressed by the patient. Nevertheless, since our larger study indicated that physicians were not so inclined to favour the practice of euthanasia (mean score for intention was 3.94 + 2.17), it calls for caution: healthcare providers should be cautious not to presume the intention of any caregivers. 9
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
10
Nursing Ethics
A second variable, added to the TPB, also significantly contributed to the prediction of nurses’ intention, which is the moral norm. While this variable has already been associated with healthcare workers’13,15 and nurses’18,19 motivation in previous studies, it is the first time that it is linked to euthanasia. Another new aspect is relating nurses’ moral norm to the ethical principle of beneficence. This result replicates those of a previous study on consent to organ donation among the general population.20 Three ethical principles (autonomy, beneficence and justice) were then tested as potential principles underlying moral norm and only beneficence was significantly related to this construct. Thus, this means that nurses will be motivated to practise euthanasia if they perceive that this act is in agreement with their personal values and principles, and also if they believe that this would help someone and be in the best interest of an end-of-life person. Surprisingly and contrary to previous reviews among nurses,3–5 the ethical principle of the patient’s autonomy did not have a significant impact on nurses’ intention to practise euthanasia. In fact, 63.01% of our respondents had even a positive intention to practise euthanasia although they did not know the patient’s wishes (or had his consent). This is rather troublesome but, in fact, not uncommon. Again, the study conducted by Inghelbrecht et al.29 showed that ‘administration of the life-ending drugs by the nurse occurred more frequently in the cases without an explicit request from the patient than in the cases of euthanasia’ (p. 908) where a request was explicit. How should this be understood since this is clearly against ethical code and present regulations where euthanasia is legalised? It has been reported that sometimes healthcare providers are faced with conflicts between competing ethical principles, such as autonomy and beneficence.30,31 In fact, according to Beauchamp and Childress,31 ‘whether respect for the autonomy of patients should have priority over professional beneficence directed at those patients is a central problem in biomedical ethics’ (p. 207). As such, in one vignette, these two ethical principles were well aligned, whereas in the other, there was a matter of debate. In this latter context, nurses were faced with making sure of respecting the patient’s autonomy, which was unknown, or relieving that person of refractory pain (beneficence). Moreover, given that nurses are often portrayed as patient advocates who want to help their patients,5 it is possible that for some of them the principle of beneficence might have overridden the absence of known wishes. Unlike previous reviews on nurses’ attitudes towards euthanasia, no socio-demographic and contextual variables had an impact on nurses’ intention in this study. In particular, religious affiliation did not contribute to nurses’ intention, while in the past, it has systematically been associated with a negative attitude towards euthanasia.3–5,32 Contrary to reviews among nurses,3,4 age and nursing specialty were also not determinants of intention. Finally, in this study, nurses’ attitude towards the legalisation of euthanasia in Canada was another variable unrelated to intention. While these results may at first seem to run counter to common sense, they support the assumptions of the TPB. According to this theory, sociodemographic variables are external factors whose influence on intention should be filtered (or mediated) through the psychosocial constructs, such as subjective norm, attitude and PBC; they should not be direct determinants of intention. These results thus provide additional support for the use of the TPB to identify nurses’ intention to practise euthanasia in palliative care. The results of this study further add to the increasing evidence on the efficacy of using an extended version of the TPB to study health professionals’ intention and behaviours. Already in 2008, a systematic review had identified the TPB as a promising theory to predict their behaviours.13 The novelty is the application of this theory to study nurses’ intention to practise euthanasia. In fact, most research in the field of palliative care is atheoretical. Yet, increasingly more authors are calling for more theory-based studies, given that the use of theories presents many advantages.33–36 For instance, theory can guide the selection of determinants to be tested as potential predictors of intention or behaviour.37 The use of theory can also facilitate replication of previous findings which is crucial to increasing scientific knowledge about certain behaviours.38 Additionally, this can provide a basis for refining and developing better theories.37 10
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
11
Strengths of this study are the novelty of using a psychosocial theory to study nurses’ intention to practise euthanasia in palliative care and the rigorous methodology used to develop and validate the survey instrument. In fact, the main methodological issues identified in two reviews on euthanasia,6,39 such as sample representativeness, reporting of the psychometric qualities of the items used in the questionnaires and use of multivariate statistical analyses, have been addressed. One limitation of the study is the low response rate. Yet, according to some authors, low response rates are more acceptable when the topic is controversial, such as in the case of euthanasia.40,41 Respondents were also compared to the population under study to ascertain they were similar in terms of gender distribution and age. Even though respondents were younger compared to nurses of the province of Quebec, Canada, the results were similar when controlling for age in the analyses. In conclusion, to our knowledge, this study is the first to report the application of a validated psychosocial theory to predict nurses’ intention to practise euthanasia in palliative care. It also has the distinction of identifying the ethical principles underlying nurses’ moral norm and intention. It is hoped that this will encourage more theoretical research on nurses’ motivation to practise euthanasia and on the ethical principles related to this practice. Conflict of interest The authors declare that there is no conflict of interest. Funding This study was supported by a grant from the Ethics Office of the Canadian Institutes of Health Research [EOG - 11392]. References 1. Gordijn B and Janssens R. The prevention of euthanasia through palliative care: new developments in The Netherlands. Patient Educ Couns 2000; 41(1): 35–46 (PMID: 10900365). 2. Pereira J, Anwar D, Pralong G, et al. Assisted suicide and euthanasia should not be practiced in palliative care units. J Palliat Med 2008; 11(8): 1074–1076 (PMID: 18980445). 3. Verpoort C, Gastmans C, De Bal N, et al. Nurses’ attitudes to euthanasia: a review of the literature. Nurs Ethics 2004; 11(4): 349–365 (PMID: 15253571). 4. Berghs M, Dierckx de Casterle B and Gastmans C. The complexity of nurses’ attitudes toward euthanasia: a review of the literature. J Med Ethics 2005; 31(8): 441–446 (PMID: 16076966). 5. Quaghebeur T, Dierckx de Casterle B and Gastmans C. Nursing and euthanasia: a review of argument-based ethics literature. Nurs Ethics 2009; 16(4): 466–486 (PMID: 19528103). 6. Vezina-Im LA, Lavoie M, Krol P, et al. Motivations of physicians and nurses to practice voluntary euthanasia: a systematic review. BMC Palliat Care 2014; 13(1): 20 (PMID: 24716567; PMCID: 4021095). 7. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process 1991; 50: 179–211. 8. Armitage CJ and Conner M. Efficacy of the theory of planned behaviour: a meta-analytic review. Br J Soc Psychol 2001; 40(4): 471–479. 9. Conner M and Sparks P. The theory of planned behaviour and health behaviour. In: Conner M and Norman P (eds) Predicting health behaviour. 2nd ed. Maidenhead: Open University Press, 2005, pp. 170–222. 10. Godin G and Kok G. The theory of planned behavior: a review of its applications to health-related behaviors. Am J Health Promot 1996; 11(2): 87–98. 11. Webb TL and Sheeran P. Does changing behavioral intentions engender behavior changes? A meta-analysis of the experimental evidence. Psychol Bull 2006; 132(2): 249–268. 12. McEachan R, Conner M, Taylor NJ, et al. Prospective prediction of health-related behaviours with the theory of planned behaviour: a meta-analysis. Health Psychol Rev 2011; 5(2): 97–144. 11
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
12
Nursing Ethics
13. Godin G, Belanger-Gravel A, Eccles M, et al. Healthcare professionals’ intentions and behaviours: a systematic review of studies based on social cognitive theories. Implement Sci 2008; 3: 36 (PMID: 18631386). 14. Triandis HC. Values, attitudes and interpersonal behavior. In: Page MM (ed.) Nebraska symposium on motivation: beliefs, attitudes and values, Vol. 1. Lincoln, NE: University of Nebraska Press, 1980, pp. 195–259. 15. Godin G, Vezina-Im LA and Naccache H. Determinants of influenza vaccination among healthcare workers. Infect Control Hosp Epidemiol 2010; 31(7): 689–693 (PMID: 20482373). 16. Daneault S, Beaudry M and Godin G. Psychosocial determinants of the intention of nurses and dietitians to recommend breastfeeding. Can J Public Health 2004; 95(2): 151–154 (PMID: 15074909). 17. Godin G, Beaulieu D, Touchette JS, et al. Intention to encourage complementary and alternative medicine among general practitioners and medical students. Behav Med 2007; 33(2): 67–77 (PMID: 17711808). 18. Cote F, Gagnon J, Houme PK, et al. Using the theory of planned behaviour to predict nurses’ intention to integrate research evidence into clinical decision-making. J Adv Nurs 2012; 68(10): 2289–2298 (PMID: 22229522). 19. Chabot G, Godin G and Gagnon MP. Determinants of the intention of elementary school nurses to adopt a redefined role in health promotion at school. Implement Sci 2010; 5: 93 (PMID: 21110872). 20. Blondeau D, Godin G, Gagne´ C, et al. Do ethical principles explain moral norm? A test for consent to organ donation. J Appl Biobehav Res 2004; 9(4): 230–243. 21. Nunnally JC. Psychometric theory. 2nd ed. New York: McGraw-Hill, 1978. 22. Fermanian J. Mesure de l’accord entre deux juges: cas quantitatif [Measuring agreement between 2 observers: A quantitative case]. Rev E´pide´miol Sante´ Publique 1984; 32(6): 408–413 (PMID: 6531504). 23. Kuhse H and Singer P. Voluntary euthanasia and the nurse: an Australian survey. Int J Nurs Stud 1993; 30(4): 311–322 (PMID: 8375974). 24. Ministry of Justice. Wet betreffende de euthanasie [Law on euthanasia] (The Belgian Act on Euthanasia of 28 May 2002) 24 August 2010, www.ejustice.just.fgov.be/loi/loi.htm 25. Bilsen JJ, Vander Stichele RH, Mortier F, et al. Involvement of nurses in physician-assisted dying. J Adv Nurs 2004; 47(6): 583–591 (PMID: 15324424). 26. Young MG and Ogden RD. The role of nurses in AIDS care regarding voluntary euthanasia and assisted suicide: a call for further dialogue. J Adv Nurs 2000; 31(3): 513–519 (PMID: 10718869). 27. Muller MT, Pijnenborg L, Onwuteaka-Philipsen BD, et al. The role of the nurse in active euthanasia and physician-assisted suicide. J Adv Nurs 1997; 26(2): 424–430 (PMID: 9292379). 28. Van Bruchem-van de Scheur A, van der Arend A, van Wijmen F, et al. Dutch nurses’ attitudes towards euthanasia and physician-assisted suicide. Nurs Ethics 2008; 15(2): 186–198 (PMID: 18272609). 29. Inghelbrecht E, Bilsen J, Mortier F, et al. The role of nurses in physician-assisted deaths in Belgium. CMAJ 2010; 182(9): 905–910 (PMID: 20479043; PMCID: 2882450). 30. Badger JM, Ladd RE and Adler P. Respecting patient autonomy versus protecting the patient’s health: a dilemma for healthcare providers. JONA’S Healthc Law Ethics Regul 2009; 11(4): 120–124; quiz 5–6 (PMID: 19935216). 31. Beauchamp TL and Childress JF. Principles of biomedical ethics. 6th ed. New York: Oxford University Press, 2009. 32. Emanuel EJ. Euthanasia and physician-assisted suicide: a review of the empirical data from the United States. Arch Intern Med 2002; 162(2): 142–152 (PMID: 11802747). 33. Baranowski T, Lin LS, Wetter DW, et al. Theory as mediating variables: why aren’t community interventions working as desired? Ann Epidemiol 1997; 7: S89–S95. 34. Fishbein M and Yzer MC. Using theory to design effective health behavior interventions. Commun Theor 2003; 13: 164–183. 35. Michie S and Abraham C. Interventions to change health behaviours: evidence-based or evidence-inspired? Psychol Health 2004; 19: 29–49. 36. Glanz K and Bishop DB. The role of behavioral science theory in development and implementation of public health interventions. Annu Rev Public Health 2010; 31: 399–418 (PMID: 20070207). 12
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
Lavoie et al.
13
37. Michie S and Prestwich A. Are interventions theory-based? Development of a theory coding scheme. Health Psychol 2010; 29(1): 1–8 (PMID: 20063930). 38. Michie S, Fixsen D, Grimshaw JM, et al. Specifying and reporting complex behaviour change interventions: the need for a scientific method. Implement Sci 2009; 4: 40 (PMID: 19607700). 39. Rosenfeld B. Methodological issues in assisted suicide and euthanasia research. Psychol Public Policy Law 2000; 6(2): 559–574 (PMID: 12661541). 40. Sierles FS. How to do research with self-administered surveys. Acad Psychiatry 2003; 27(2): 104–113 (PMID: 12824111). 41. Rubenfeld GD. Surveys: an introduction. Respir Care 2004; 49(10): 1181–1185 (PMID: 15447800).
13
Downloaded from nej.sagepub.com at UNIV OF WISCONSIN OSHKOSH on November 15, 2015
