Compassion fatigue or secondary traumatic stress was first recognised in the 1950s in nursing and other frontline professionals, such as first responders and fire fighters, who dealt with traumatic incidents (Beck, 2011). Subsequently, ‘compassionate care’ has featured in a number of key policy documents, in particular those published following the Francis report (Francis, 2013; NHS England, 2016). Debate continues about the definition of compassionate care and how its delivery is measured. In a systematic review to define compassion, Perez-Bret et al (2016) came to the conclusion that:
‘Compassion originates as an empathic response to suffering, as a rational process which pursues patients' wellbeing, through specific, ethical actions directed at finding a solution to their suffering’.
Compassion can be seen in the sensitive understanding of another's suffering by those delivering care, and who then seek solutions to relieve that suffering and promote wellbeing. Sinclair et al (2016) identified six themes of ‘perceptions of compassionate care’. These consisted of the following:
These themes appear to give a broader, descriptive definition that could be applied in practice. They show the multifaceted nature of compassionate care and the factors that impact on care delivery. Within these themes there is a clearly identifiable emergent relationship between compassion fatigue, interpersonal factors and barriers and enablers.
Compassion fatigue, or a gradual lessening of compassion with associated symptoms of anxiety, stress and negative attitudes, has the potential to impact significantly on productivity (Hooper et al, 2010; Lilius et al; 2011). It is hypothesised that individuals who have a greater capacity for feeling or empathy are more at risk of experiencing symptoms. The presence of compassion fatigue may be identified through assessing nurses' personal feelings relating to death and dying, or death anxiety (Coetzee and Klopper, 2010).
Mitchell et al (2006) found that nurses working in high-skill-level care environments often experienced higher levels of stress than nurses working in areas where they are less exposed to death and dying. Stress levels were found to be significantly higher in nurses working in clinical areas, where they were frequently exposed to sudden and traumatic deaths in patients. Nurses working in critical care and emergency departments have a greater exposure to sudden and unplanned death than those who work in palliative care or community services (Mitchell et al, 2006). However, it could be argued that, although death and dying may be more expected within these areas, the relationships that exist between nurses caring for patients over long periods of time may also result in significant emotional distress when death occurs (Mitchell et al, 2006).
Other factors that cause additional stress, specifically for nurses working in acute environments, include time constraints and balancing decisions on care delivery for the end-of-life patient and that of the ‘rescuable patient’. Results from a number of studies provide evidence of lower anxiety in nurses where these issues are less prevalent due to the clinical condition of patients within these areas (Mitchell et al, 2006; Moola et al, 2008) Culture, age and working experience may also influence coping mechanisms within this cohort (Peters et al, 2013). Anecdotal evidence points to a ‘culture of silence’ in many organisations where stressful events within critical care environments are not discussed, and where debriefing is random and rare.
It is suggested that because of frequent exposure to death and dying during their work ‘life’, many nurses experience compassion fatigue in addition to developing a greater awareness of their own mortality. The latter may result in experiencing feelings related to ‘death anxiety’, which impacts on the delivery of care at the end of life. As the focus is lost, and feelings of self-doubt, in association with entrenched feelings of negativity, prevail, the ability to deliver compassionate care diminishes or is lost (Black, 2007; Hooper et al, 2010).
In critical care units, the expectations of family are often unrealistic in relation to positive outcomes due to the technology available and expertise that is synonymous with these environments. These factors contribute to high levels of anxiety and stress among teams. Ciccarello (2003) suggested that nurses have the skills to deal with critically ill patients, but are less skilled in caring for the dying patient. This would support the concept of advanced care planning needs being promoted in critical care units; including end-of-life discussions to ‘humanise’ the environment (Angus et al, 2016).
Empirical evidence from studies looking at death anxiety in nurses have predominantly taken place in Western societies, with a small number in Asia. There is very little evidence from Middle Eastern or African countries. This may be due to cultural and religious beliefs associated with specific population groups and a reluctance to explore the concept (Najjir et al, 2009). Other factors that may have impacted on available data are recruitment of subjects and the ranking of importance of the subject matter within healthcare organisations (Najjir et al, 2009). One study in a hospital in Israel examined the relationship between nurses working in oncology and their attitudes towards death and caring for dying patients. The Frommelt Attitude towards Caring of the Dying and the Death Attitude Profile–Revised Scale was completed by 147 nurses and the conclusions were that personal attitudes towards death were associated with attitudes towards care of the dying patient. Some nurses used avoidance to cope with their own personal fears; secondary findings suggested a link between culture and religion in the development of death attitudes (Braun et al, 2010). The conclusions from the literature review were that education programmes on death and dying may reduce anxiety and improve end-of-life care. Greater insight by nurses into their own beliefs or attitudes in relation to death may also translate into patient care benefits.
The Death Attitude Profile Revised (DAP-R) questionnaire is similar in concept to the Frommelt Attitude towards Caring of the Dying and the Death Attitude Profile that can be used to assess attitudes and beliefs (Table 1). It is a multidimensional questionnaire with a seven-point Likert scale that allows for a broader scope of response from ‘strongly disagree’ (SD) to ‘strongly agree’ (SA) and asks questions in five identified dimensions. The tool is cost-effective and both easily implemented and validated in adult populations. The response dimension ranges from fear avoidance to approach, escape and neutral acceptance (Wong et al, 2004).
| Measurable dimension | Explanation |
|---|---|
| Fear of death | Confronting death and the feelings evoked |
| Death avoidance | Avoiding all thoughts or references to death to reduce death anxiety |
| Approach/acceptance | Death is viewed as a gateway to a happy afterlife |
| Escape acceptance | Death is viewed as an escape from a painful existence |
| Neutral acceptance | Death is viewed as a reality, which is neither feared nor welcomed |
Background
Historically there has been little education provided to nurses working within critical care units on death or care of the dying within critical environments (Curtis et al, 2010). This may have resulted from the view of critical care environments as being curative rather than palliative (Truog et al, 2008). There is an anecdotal perception that, with advances in technology, death is an unlikely outcome in intensive care units (Truog et al, 2008).
Research has suggested that between 16 and 85% of healthcare workers across the clinical spectrum will develop compassion fatigue (Hooper et al, 2010). There has also been little in the way of providing debriefing opportunities or recognition of the need for a more formal structured approach within this highly charged atmosphere to support staff post-traumatic incidents.
There is much debate around the subject of whether compassion can be taught (Bray et al, 2014; Christiansen et al, 2015; Winch et al, 2015). Is it an inherent trait? How do we measure compassionate care? Other themes related to delivery of compassionate care have included whether we can enable compassion care delivery through role modelling and what barriers to delivery exist. It is suggested that there are several barriers to implementation including individual and organisational factors, such as high workload, low staffing levels and lack of autonomy and feeling valued (Firth-Couzins et al, 2009).
Study aims and methodology
Thematic qualitative analysis was considered as a research methodology; however, anecdotal and observational evidence suggested that this may not have been the most appropriate approach owing to the subject matter. This has been identified as a factor in the lack of data in relation to this topic, with regional differences apart (Najjir et al, 2009). Following ethical approval, a mixed-methods study was undertaken. The quantitative arm used non-random purposive sampling of qualified nurses working across three subspecialty critical care units in a hospital in Qatar (n=255 full time equivalent). The research questionnaire was provided to the units along with secure boxes for completed data forms. DAP-R was used in conjunction with demographic profile questions (Wong et al, 2004).
Primary outcomes
The primary outcomes of the study were to assess attitudes towards death and dying in a multicultural cohort of critical care nurses. The hypothesis was that barriers and enablers to the delivery of compassionate care are multifactoral and include interpersonal causes. Independent variables of age, gender, professional qualifications designation, length of time qualified and length of time working in critical care units were measured against response to the five dimensions of the DAP-R questionnaire (Table 2).
| Variables | SICU |
TICU |
MICU |
| Gender | |||
|---|---|---|---|
| Male | 11(36.7) | 0 | 12(50) |
| Female | 19(63.3) | 27(100) | 12(50) |
| Age group (years) | |||
| 20–27 years | 3 (10.0) | — | — |
| 28–35 years | 2 (6.7) | 5(18.5) | 1(4.2) |
| 36–43 years | 18(60.0) | 14(51.9) | 15(62.5) |
| 44–51 years | 5 (16.7) | 4(14.8) | 6(25.0) |
| 52–59 years | 2(6.7) | 3(11.1) | — |
| 60 years | — | 1(3.7) | 2(8.3) |
| Profession | |||
| Head nurse | 1(3.3) | 1(3.7) | 2(8.3) |
| Charge nurse | 1(3.3) | — | 22(91.7) |
| Staff nurse | 27(90) | 26(86.7) | — |
| Case manager | 1 (3.3) | — | — |
| Educational level | |||
| Associate degree in Nursing | 6(20.0) | — | 1(4.2) |
| Diploma in Nursing | 3(10.0) | 8(29.6) | 3(12.5) |
| Bachelor's in nursing | 21(70.0) | 19(70.3) | 20(83.3) |
| HMC experience (years) | |||
| <2 | 3(10.0) | 10(37.0) | 4(16.7) |
| 2–6 | 14(46.7) | 9(33.3) | 7(29.2) |
| 7–11 | 10(33.3) | 6(22.2) | 6(25.0) |
| 12–16 | 3 (10.0) | 2(7.4) | 7(29.2) |
| ICU experience (years) | |||
| <2 | 7(23.3) | 11(40.7) | 9(37.5) |
| 2–6 | 12(40.0) | 9(33.3) | 5(20.8) |
| 7–11 | 10(33.3) | 1(3.7) | 7(29.2) |
| 12–16 | 1(3.3) | 2(7.4) | 2(8.3) |
| 17–21 | — | 4(14.8) | 1(4.2) |
Results
Data were analysed using the Pearson chi-squared test owing to small numbers (relative risk=81) (Tables 3–5).
| Dimension | SD | D | MD | U | MA | A | SA | |
|---|---|---|---|---|---|---|---|---|
| Fear of death | Sum | 30 | 61 | 11 | 35 | 11 | 27 | 29 |
| Mean | 4.3 | 8.7 | 1.6 | 5.0 | 1.6 | 3.9 | 4.1 | |
| Death avoidance | Sum | 23 | 29 | 7 | 18 | 14 | 39 | 13 |
| Mean | 4.6 | 5.8 | 1.4 | 3.6 | 2.8 | 7.8 | 2.6 | |
| Neutral acceptance | Sum | 10 | 10 | 4 | 19 | 2 | 46 | 53 |
| Mean | 2 | 2 | 0.8 | 3.8 | 0.4 | 9.2 | 10.6 | |
| Approach acceptance | Sum | 25 | 23 | 8 | 53 | 12 | 101 | 59 |
| Mean | 2.5 | 2.3 | 0.8 | 5.3 | 1.2 | 10.1 | 5.9 | |
| Escape acceptance | Sum | 32 | 40 | 2 | 15 | 8 | 21 | 25 |
| Mean | 6.4 | 8.0 | 0.4 | 3 | 1.6 | 4.2 | 5.0 | |
SD=strongly disagree; D=disagree; MD=moderately disagreed; U=undecided; MA=moderately agree; A=agree; SA=strongly agree
| Dimension | SD | D | MD | U | MA | A | SA | |
|---|---|---|---|---|---|---|---|---|
| Fear of death | Sum | 17 | 30 | 6 | 42 | 20 | 56 | 13 |
| Mean | 2.4 | 4.3 | 0.9 | 6.0 | 2.9 | 8.0 | 1.9 | |
| Death avoidance | Sum | 4 | 27 | 10 | 26 | 9 | 50 | 8 |
| Mean | 0.8 | 5.4 | 2.0 | 5.2 | 1.8 | 10 | 1.6 | |
| Neutral acceptance | Sum | 5 | 5 | 8 | 24 | 6 | 52 | 35 |
| Mean | 1.0 | 1.0 | 1.6 | 4.8 | 1.2 | 10.4 | 7 | |
| Approach acceptance | Sum | 8 | 16 | 13 | 67 | 10 | 106 | 44 |
| Mean | 0.8 | 1.6 | 1.3 | 6.7 | 1.0 | 10.6 | 4.4 | |
| Escape acceptance | Sum | 20 | 20 | 7 | 29 | 17 | 36 | 6 |
| Mean | 4 | 4 | 1.4 | 5.8 | 3.4 | 7.2 | 1.2 | |
SD=strongly disagree; D=disagree; MD=moderately disagreed; U=undecided; MA=moderately agree; A=agree; SA=strongly agree
| Dimension | SD | D | MD | U | MA | A | SA | |
|---|---|---|---|---|---|---|---|---|
| Fear of death | Sum | 22 | 30 | 12 | 19 | 12 | 55 | 13 |
| Mean | 3.1 | 4.3 | 1.7 | 2.7 | 1.7 | 7.9 | 1.9 | |
| Death avoidance | Sum | 16 | 31 | 10 | 15 | 6 | 30 | 8 |
| Mean | 3.2 | 6.2 | 2.0 | 3.0 | 1.2 | 6.0 | 1.6 | |
| Neutral acceptance | Sum | 4 | 4 | 4 | 10 | 6 | 38 | 50 |
| Mean | 0.8 | 0.8 | 0.8 | 2.0 | 1.2 | 7.6 | 10 | |
| Approach acceptance | Sum | 10 | 15 | 8 | 39 | 13 | 69 | 76 |
| Mean | 1.0 | 1.5 | 0.5 | 3.9 | 1.3 | 6.9 | 7.6 | |
| Escape acceptance | Sum | 22 | 27 | 7 | 14 | 8 | 27 | 11 |
| Mean | 4.4 | 5.4 | 1.4 | 2.8 | 1.6 | 5.4 | 2.2 | |
SD=strongly disagree; D=disagree; MD=moderately disagreed; U=undecided; MA=moderately agree; A=agree; SA=strongly agree
Demographic profile of sample
Participants were predominantly female, aged between 28 and 32 years, educated to Bachelor of Science in nursing (BSN) level, employed in the organisation from 2 to 11 years with 7-11 years' specialty experience.
Outcome data
The nurses surveyed predominantly took a neutral or acceptance approach to death and dying. In neutral acceptance, death is viewed as a reality, which is neither feared nor welcomed. Death is viewed as ‘a gateway to a happy afterlife. Fear of death positively correlated, or was associated, with qualification, gender and years in ICU and negatively correlated, or was associated, with age and time employed in the organisation. Death avoidance positively correlated with qualifications, gender, age group and years in ICU and negatively correlated to time employed in the organisation. Neutral acceptance positively correlated, or was associated, with qualification, age and years employed in the organisation, and negatively correlated, or associated with, gender and years worked in ICU. Approach acceptance positively correlated, or was associated, with qualifications, time employed in the organisation, years worked in ICU and negatively correlated to age and gender. Escape acceptance positively correlated, or was associated, with age and years employed in the organisation. Qualifications, age and years employed in the organisation were positively associated with a ‘neutral response’.
All of the above variables and years spent in ICU were associated with an ‘approach acceptance’. Female gender and longer working lives spent in critical care were clearly identifiable risk factors for presence of compassion fatigue as evidenced through approach and neutral acceptance response. The most significant predictor would appear to be length of time working in critical care. Additionally, anecdotal evidence would suggest that less than 20% of the nursing workforce in Hamad Medical Corporation consists of the indigenous Qatari population group, with the nursing workforce consisting largely of expatriate migrant workers. Nursing personal are predominantly recruited from other Arab countries in the region, India and the Philippines. Interpersonal factors in migrant populations that may impact on performance include extended family financial support responsibilities, long absences from countries of origin and cultural and religious beliefs. Contributory factors influencing participant responses may be the migration status of the nursing workforce. Ensuring the health and wellbeing of all employees is an organisational responsibility; however, it may be that additional supportive measures are required for this workforce group.
Qualitative outcome data
A small number of nurses (n=8) were asked to discuss the following questions:
The narrative highlighted two themes: the first being predominantly that compassionate care was not delivered in critical cares units, the focus was on delivering competent care. The second theme was that there was a desire to deliver competent care; however, barriers were identified that included staff shortages and high turnover of patients. Professional behaviour was associated with competency rather than compassion as a result of having to prioritise clinical care.
Discussion
In the nurses working in critical care who were surveyed, there appeared to be an acceptance of death as either an escape from pain and suffering, or a reality. Qualifications, age and years employed in the organisation were positively associated with a ‘neutral response’. These variables, including years spent in ICU were associated with an ‘approach’ acceptance. This would suggest that there is an element of compassion fatigue within the nursing cohort working in critical care environments reflected in their response to death and dying. What is also apparent from the literature is the influence of culture and beliefs on attitudes to death and dying. Compassion fatigue in nursing populations has been underexplored specifically within Middle Eastern and African populations. Additionally, migrant workforce populations have many significant issues that impact on coping abilities or compassion fatigue. These relate to personal life circumstances, such as financial responsibilities for families and extended families and the reasons for migration. Within the bigger picture of these pressing concerns, the result may be a distinct separation between professional and personal lives; with neither impacting on each other. Thus, the nursing role and actions may be functional and skilful but less empathetic. The job or task is completed to the best of their ability, but without the unnecessary emotional expenditure. Multiple variables are suggested as factors that impede or hinder the delivery of compassionate care. These include traditional biomedical rather than patient-centred care models (Firth-Cozens and Cornwell, 2009); an overemphasis on target-driven initiatives, cost measures including adjustment of staff to patient number ratios, resulting in higher levels of stress and staff burn-out (Bradshaw, 2009). Within this, nurse attitudes towards death and dying may be a manifestation of the presence of compassion fatigue. It may become the easier option to adopt a neutral acceptance or approach to death and dying to reduce the emotional labour of caring. Coping mechanisms develop to allow nurses to function within a clinical role without the burden of caring, this is necessary for survival. Certainly in societies or organisations where anxiety or stress impact on staff, emotional wellbeing and performance are not mandatory policies for evaluation or data collection. Neutral acceptance or approaches to death and dying may well be the preferred option for staff (World Health Organization, 2012).
Study limitations
The small sample size in the present study impacts on generalisability to the wider population and statistical significance, with approximately 30% of total critical care workforce participating in the survey. Annual leave and sickness absence are not included in analysis of available respondents. Despite the small sample size, there is a positive correlation seen between variables of age, gender duration of employment, length of time working in ICU and responses. The nature of the subject being researched, the ethnicity and a largely migrant workforce may also have impacted on response rate.
In Qatar (Arab) nationals account for less than 15% of the total population, with other Arabs at 13%. Asia provides the largest proportion of the expatriate workforce labour: 24% are Indian, 16% Nepali, 11% Filipino and 5% each from Bangladesh and Sri Lanka. Anecdotal evidence that less than 20% of the nursing workforce in HMC is Qatari; therefore, the nursing workforce consists largely of expatriate migrant workers. Cultural, societal and religious beliefs may also influence views on death and dying (Qatar World Population Review, 2018).
Pilot studies looking at implementing a process in which recommendations for care and treatment in emergency situations are provided may be useful for future work in this area. A substantial amount of work is being undertaken as part of the ReSPECT campaign endorsed by the Resuscitation Council UK (RCUK) in promoting a process that guides decision-making in emergency or life-threatening situations (RCUK, 2019).
Conclusions
There is a clear and pressing need for more research on compassion fatigue, attitudes towards death and dying and the delivery of compassionate care. The ability to view death in the neutral or acceptance domains may indicate the development of coping or self-preservation strategies to ensure care continues to be delivered. However, no emotional attachment exists, and empathy is often absent. Death and dying may be a frequent traumatic occurrence with which nurses working in critical care environments are faced. Without structured supportive measures, such as staff wellbeing or counselling services, or simple debrief and reflection exercises, the emotional and physical impact that occurs due to repeated exposure can only be dealt with on a personal level by detachment from the event that is causing the stress. Variables of extended family financial support responsibilities, long absences from countries of origin and cultural and religious beliefs in a largely migrant nursing workforce may be contributory factors influencing participant responses (Lancet, 2017). Klimecki et al (2012) propose, that there exists a state of empathetic fatigue, rather than compassion fatigue, where compassion remains but empathy as an emotion is blurred or diminished. The state of empathetic fatigue is more difficult to measure, but clearly, predisposing factors are similar to those of compassionate fatigue. The terms empathy and compassion are interlinked and to separate or silo may be counterproductive. What is clear is that greater measures of recognition of distress and emotional fatigue in healthcare professionals across the board is crucial.
Correlations with morbidity and mortality outcomes of patients in critical care units measured against evidence of compassion fatigue and patient/family satisfaction would be useful indicators of a relationship existing between attitudes to death and dying, and care delivery. The body of research needs to be expanded with comparative analysis made across clinical and geographical spectrums. This is particularly true when considering global employment migration as capacity demands increase within healthcare organisations worldwide (International Organization for Migration, 2018). Ageing populations, increasing industrialisation and a shift in population demographics in emerging nations have resulted in a capacity and demand gap that can only be filled through migration of people. However, culture, past and current life experiences may impact on the interpretation of compassionate care. Organisations and leaders are key stakeholders in promoting the delivery of compassionate care through putting in place mechanisms for recognising compassion fatigue and promoting supportive workplace environments.
The debate on whether compassion can be taught should also be on the agenda of education and healthcare research groups. This phenomenon occurs across geographical areas and sectors with multiple variables impacting on delivery of compassionate care. The debate continues about how to address this in a meaningful and effective way. Healthcare has become a global commodity. Punitive individual blame cultures do little to address the wider concerns for patients, families and health professionals.
Nurses working in high-skill-level care environments often experience higher levels of stress than nurses working in areas where they are less exposed to death and dying. Recognition that compassion fatigue is a factor that impedes the delivery of holistic and compassionate care is the first step. It is then incumbent on organisations to put in place pathways that not only recognise the early stages of compassion fatigue, but provide support interventions to enable nurses to manage workplace stress and anxiety.