Self-compassion training in palliative care during COVID-19: A pilot study

Abstract

Objectives

This pilot project replicated a self-compassion program to support health-care professionals in palliative care settings. We anticipated that undertaking this program would enhance participants’ psychological well-being.

Methods

Participants were recruited by convenience sampling from palliative care services in an area of Melbourne, Australia. Because of the COVID-19 pandemic, the program was offered online and comprised six once-weekly gatherings. Three survey rounds with identical questions were conducted prior to, immediately after, and 3 months after the training. The surveys consisted of 6 scales: the Self-Compassion Scale, the Santa Clara Brief Compassion Scale, the Cognitive and Affective Mindfulness Scale, the Depression Anxiety Stress Scale, the Professional Quality of Life Scale, and the Interpersonal Reactivity Index Scale. Paired-sample t-test and repeated measures analysis of variance analyses were used to compare participant responses.

Results

Nine participants working in palliative care completed the training and 3 surveys and included qualitative responses. The experience was overwhelmingly positive. Self-compassion (F = 14.44; p < 0.05) and mindfulness (F = 18.44; p < 0.05) scores significantly increased post-training, were picked up in a short time and endured. The emotional state improved by compassion satisfaction, showing a positive improvement 3 months post-training, along with there being no changes in compassion satisfaction, burnout, and dispositional empathy.

Significance of results

Participants all worked in a supportive environment, which encouraged self-care. Even against the pandemic difficulties, a short online program was effective, assisting participants to cultivate their inner resources in mindfulness and self-compassion. While a small sample size, expansion of the training may benefit the wider palliative care workforce.

Introduction

Compassion may be an important underlying motivation for health-care professionals’ initial career choice to make a difference and then an ongoing component that enables them to continue their work for those in need of care; compassion is one of the 5 nursing professional values. Health-care literature and policy indicate that for those needing care, compassionate care in health-care professionals directly correlates with their improved satisfaction with care (Sinclair et al. 2017).

Compassion is a response to suffering and means being sensitive to both our own and others’ suffering, which may enable deep motivation to alleviate and prevent it (Feldman and Kuyken 2011). Sinclair et al. (2018) explored health-care professionals’ views on compassion, concluding that not all pain can be alleviated or solved. Moreover, while there is acknowledgment that suffering is approachable when considered through the landscape of compassion (Feldman and Kuyken 2011), the ability to provide a compassionate environment of care is challenged by experiences of burnout and occupational stress (Sinclair et al. 2017). However, holding a compassionate response in a culture where the emphasis is on the care of others may mean that health-care professionals place these needs above their own (Neff et al. 2020).

While compassion is required in any area of health-care practice, it is well discussed as an essential skill in the specialty area of palliative/end-of-life care (Bessen et al. 2019; Brito-Pons and Librada-Flores 2018; Mills et al. 2018). In palliative care settings, compassion specifically involves understanding the dying person and their family’s physical, emotional, social, and spiritual suffering, as well as sensitive conversations about losses, grief, death, and dying. While these issues may cause distress for some, those who choose to work in palliative care exhibit less stress than their colleagues in other health-care settings (Peters et al. 2013a). Additionally, those who maintain their own well-being and support networks are more likely to be satisfied and engaged (Cross, 2019) and more likely to continue working in palliative care, being committed to quality care for those in need (Peters et al. 2013a).

Compassionate care has assumed considerable significance with the dire impact of the COVID-19 pandemic over 2020–2022 on health-care systems and no less in palliative care settings. Institutional health-care professionals have been at the front line of a relentless procession of seriously ill people with COVID-19, who most often received care and/or died alone, in isolation from their significant others, because of a ban on visitors. While not yet studied, anecdotal evidence suggests a rise in demand for home-based palliative care, and subsequent deaths at home, because of reticence to enter institutional care. COVID-19 has required clinicians to manage difficult encounters in many different ways; for example, health-care professionals have supported a dying person’s farewell using online tools (Hossain and Clatty 2021).

Compassion is required not only for others but, ideally, for oneself. Health-care professionals also struggle with suffering and difficulties simply because they are human, and additionally, they may also experience difficulties arising from their work (Brito-Pons and Librada-Flores 2018). Dev et al. (2018) investigated the widely accepted link between burnout and compassion fatigue, finding that the greater the experience of burnout, the more barriers there were to experiences of compassion, thus inhibiting a clinician’s ability to provide compassionate care. A balance then is required in valuing the care of self as equally paramount to the care of others.

Self-compassion may be a way of managing the difficulties experienced in working with others, offering protection and resilience against caregiver fatigue and burnout (Franco and Christie 2021). It may be experienced as a positive emotion to assist in maintaining one’s own health and well-being (Mills et al. 2018). Neff et al. (2020) argue that learning to be compassionate toward oneself may contribute toward sustaining caregiver roles and off-setting exposure to the trauma of others. Being self-compassionate may mean one is more likely to care for oneself (Mills 2021) and, thus, a way of caregivers sustaining themselves without becoming drained (Neff et al. 2020). However, Dev et al. (2018) critiqued the presence of self-compassion as a predictor of burnout as a way to manage environmental barriers that inhibit compassion; they concluded that self-compassion is perhaps more relevant to an individual’s own conduct than to external workplace factors that contribute to burnout, which remains outside the individual’s control.

Neff et al. (2016) describe self-compassion as including 3 interacting aspects: being kind toward the self rather than self-judging (treating oneself with warmth, encouragement, and care through words and behaviors when experiencing struggle); holding a sense of humanity versus isolation (taking the wider perspective that we are more connected by our human condition, with its inherent struggles, as opposed to feeling isolated); and being mindful (being present to our changing experience in a balanced and accepting way), rather than over-identifying when facing painful self-relevant thoughts and emotions.

Self-compassion may provide a buffer against the constant stress evident in most health-care settings and is a skill that can be learned and developed (Neff et al. 2020). Several self-compassion programs are reported (Franco and Christie 2021; Neff and Germer 2013; Neff et al. 2020; Watts et al. 2021) with different modes of delivery and for about 2 to two and a half hours over 8 weeks and a full-day retreat, but they commonly involve reflection and some meditation components, practiced over several weeks between sessions. Neff et al. (2020) modified their 8-week Mindful Self-Compassion program with time-poor health-care professionals in mind, given that the time requirements of many programs may themselves cause stress. Thus, the shorter “Self-Compassion Training for Healthcare Communities” (SCHC) program offered 6 sessions of approximately 1 hour per week, with an invitation to practice self-compassion skills during times of challenge, “on the spot” within the context of their work.

After participation in a self-compassion program, health-care professionals reported greater work engagement, less emotional and physical exhaustion, as well as less stress and sleep disturbances (Franco and Christie 2021; Neff and Germer 2013; Neff et al. 2020; Watts et al. 2021). “Compassion literacy” (Mills 2021) is the ability to balance compassion for others with self-compassion. In light of these benefits, self-compassion may be a valuable resource for those in caring roles.

This pilot project aimed to replicate and test the SCHC (Neff et al. 2020) to support people in caring roles in palliative care settings in Melbourne, Australia. The project was especially relevant because, as noted, the care of those with COVID-19 was proving difficult and stressful for palliative care staff. We anticipated that undertaking this program would enhance participants’ psychological well-being and resilience, which are factors known to mitigate against burnout.

Methods

As noted, this pilot study is a modified replication of Neff et al. (2020) study in the choice and use of survey tools to measure the impact of the training program. The noticeable difference between that program and this study was using online rather than face-to-face delivery.

Data collection

Over several months in early 2021, participants were recruited by convenience sampling from members of a consortium of palliative care services in the northern and western areas of Melbourne, Australia. Electronic notification of the self-compassion training was circulated throughout the organizations; those interested were invited to contact the program coordinator.

The self-compassion training program was initially offered as “in-person” but subsequently was modified to an online offering because of restrictions caused by COVID-19. The training, led by an experienced palliative care nurse, who was also an experienced mindfulness-based stress-reduction and mindful self-compassion facilitator, comprised once-weekly gathering for 75 minutes in the evenings, for 6 weeks, and delivered online over Zoom. Alongside the training, 3 survey rounds with identical questions were conducted prior to, immediately after, and 3 months after the training. Having received ethical approval (Monash University, 26013), the explanatory statement and consent forms were then electronically sent, and when consent was received, the first survey was emailed. Subsequent surveys were also emailed, with a reminder sent a week afterward. Members of the research team were not involved in any of the sessions and were unknown to the participants.

In addition to specific survey questions, open-ended questions were included in the final survey at the conclusion of the program, and this content was included in the results. To de-identify the data, each participant was instructed to create a pseudonym and maintain the use of the pseudonym for each survey throughout the study.

Content of the program

During the course of the 6-week program, participants learned and practiced skills including embodied attentional practices to downregulate the nervous system in times of stress; befriending difficult emotions; cultivating kind and encouraging self-talk to counteract self-criticism; informal mindfulness; an inclusive balancing meditation offering kindness to self and others during difficult encounters; and connecting to one’s values and purpose at work. Participants were encouraged to bring awareness to the times that they experienced stress or distress at work and experiment by applying these skills in real time throughout the 6 weeks. Reflection time was built into the program in the form of small online breakout groups, encouraging connection and shared learning.

Instruments

The questionnaire was electronically developed using Qualtrics, comprising demographic information and 6 scales utilized in a previous study (Neff et al. 2020). Demographic information included pseudonyms, age, and job roles.

The Self-Compassion Scale—Short Form consists of 12 items rated on a 5-point Likert scale ranging from “Mostly never” to “Almost always” (score range: 12–60) (Raes et al. 2011). Higher scores indicate higher levels of self-compassion. The overall self-compassion level was interpreted according to Raes et al. (2011), with a score between 12 and 29.9 considered to be low, between 30 and 42 to be moderate, and between 42.1 and 60 to be high.

The Santa Clara Brief Compassion Scale comprises 5 items rated on a 7-point Likert scale ranging from “Not at all true of me” to “Very true of me” (score range: 5–35) (Hwang et al. 2008). Higher scores imply higher levels of compassion toward others.

The Cognitive and Affective Mindfulness Scale contains 12 items rated on a 4-point Likert scale ranging from “Rarely/Not at all” to “Almost always” (score range: 12–48) (Feldman et al. 2006). Higher scores reflect greater mindful qualities.

The Depression Anxiety Stress Scale (DASS) comprises 21 items rated on a 4-point Likert scale ranging from “Did not apply to me at all” to “Applied to me very much or most of the time” (score range: 0–126) (Lovibond and Lovibond 1996). Higher scores indicate a more negative emotional state.

The Professional Quality of Life Scale contains 30 items rated on a 5-point Likert scale ranging from “Never” to “Very often.” Each of the 3 subscales (compassion satisfaction, burnout, and secondary traumatic stress) included 10 items (score range: 10–50) (Stamm 2010). Higher scores represent greater feelings of work-related pleasure (compassion satisfaction subscale), exhaustion (burnout subscale), or fear (secondary traumatic stress subscale).

The Interpersonal Reactivity Index Scale consists of 28 items rated on a 5-point Likert scale ranging from “Does not describe me well” to “Describes me very well” (score range: 28–140) (Davis 1980). Higher scores reflect higher levels of dispositional empathy.

Statistical analysis

The score for each scale was determined as the sum of individual item responses. For the DASS, the summed item responses were multiplied by 2 to calculate the final score. Descriptive analysis (mean and standard deviation) and normality tests were conducted for all variables. Repeated measures analysis of variance (ANOVA) analyses were performed to compare participant survey responses before, immediately after, and 3 months after the training program, and paired-sample t-test was used to compare between-group differences. A 95% confidence interval was applied to determine the statistical significance of all results. Statistical analysis for quantitative data was performed using the SPSS software (IBM SPSS Statistics for Windows, version 26.0. Armonk, NY: IBM Corp.). This study had a statistical power of 18%, with a sample size of 9, an alpha level of 0.05 (2 tails), and an effect size of 0.39 (Faul et al. 2007). According to Cohen (1988), effect sizes below 0.50 were interpreted as small, between 0.50 and 0.80 as moderate, and above 0.80 as large.

Results

Participant demographics

Twelve participants were recruited, of which 1 did not complete the program for an unknown reason and 2 individuals did not complete the 2nd and 3rd survey rounds, thus leaving a final study sample of 9 individuals for analysis. The participants were all women, aged between 41 and 64 years, and all worked in a palliative care setting, either institutional or home-based care. Participants were from a variety of disciplines including nursing, education, social work, chaplaincy, speech pathology, and counselling.

Feedback from participants

Responses to the overall program were overwhelmingly positive, with participants describing the challenges of “initially, remembering to do the practice” (P1) and “letting go of years of habit of being self-critical” (P2). Gaining an understanding of how being compassionate impacts one’s personal life was noted, with a participant suggesting new awareness of “being conscious of noticing my inner critic and offering myself compassion” (P3), particularly in “dealing with raw feelings and emotions” (P3). In describing the personal impact, 1 participant said she noticed the difference in “not being alone in challenges and to show myself more self-compassion” (P6) and “treating myself as I would a dear friend” (P9).

Participants then described applying their learnings about being self-compassionate in one’s work, in “knowing I can be compassionate towards myself while listening to clients” (P1) and developing the “ability to listen and share (if wanted to) with others without judgment” (P3).

In relation to the online delivery mode, participants commented that “it worked better than expected” (P6) and that they had “a reasonable sense of other participants and enjoyed their company throughout the course” (P4). One participant (P1) said they would not have been able to attend if the training had been in person.

Quantitative results

The self-compassion levels started at a moderate score and significantly increased by 16% immediately post-training (mean changed from 35.0 to 44.4) and by a further 6% (mean changed from 44.4 to 48.2) in the subsequent 3 months (F(2,16) = 14.44; p < 0.05). There was a significant increase (11%, mean changed from 29.1 to 34.3) in mindfulness levels immediately after training and a further 3% increase (mean changed from 34.3 to 35.9) in the following 3-month period (F(2, 16) = 18.44; p < 0.05) (Table 1).

Table 1. Comparison of participant responses prior to, immediately after, and 3 months after training program (N = 9)

* p < 0.05.

** p ≤ 0.001.

ANOVA, analysis of variance.

The self-reported emotional state (as evidenced from the DAAS, measuring depression, anxiety, and stress) showed a positive improvement 3 months post-training (mean changed from 39.3 to 16.7) from the initial survey (F(2,16) = 9.85; p < 0.05). While there was no significant change in the participants’ work-related pressure and exhaustion before and after training, the work-related fear level (secondary traumatic stress) significantly decreased by 9% immediately after training (mean changed from 25.1 to 20.4) and by a further 3% in the following 3 months (mean changed from 20.4 to 18.7) (F(2,16) = 11.24; p < 0.05).

Participants reported a relatively high baseline level of compassion toward others, which slightly declined post-training and further dropped in the following 3 months (mean changed from 28.2 to 25.8). A similar finding was noted in dispositional empathy (mean changed from 95.8 to 91.0). However, statistically, no significant change was found in the participants’ dispositional empathy level and their compassion toward others level.

Responses to the Compassion Satisfaction subscale in the Professional Quality of Life Scale indicated moderate to high levels of work-related pleasure (mean changed from 39.4 to 39.8). The Burnout subscale in the Professional Quality of Life Scale showed no significant change across the 3 time periods (mean changed from 31.8 to 31.2; F(2,16) = 0.26; p > 0.05) and showed a relatively low level of work-related fear, which reduced further across the 3 survey rounds (mean changed from 25.1 to 18.7; F(2,16) = 11.24; p < 0.05) (Table 1).

Discussion and conclusion

This paper describes the pilot delivery of a self-compassion program, modified to an online delivery format, to accommodate the meeting limitations imposed by COVID-19 restrictions, but otherwise replicating a study by Neff et al. (2020). The COVID-19 pandemic additionally meant that engagement with those needing care was often online, creating a novel challenge for palliative care staff charged with managing not only the volume of client care remotely but also an increase in the number of clients choosing to remain at home for their end-stage care.

The efficacy of online delivery was unknown, so we were interested in the results from that perspective, especially given the contextual pressures of increased workloads and changes in work practices imposed by COVID-19. The attrition rate was very low, with only 1 participant not completing the program for an unknown reason and 2 others who completed the program but did not complete all surveys. While there were few comments about the delivery mode, all were positive, highlighting the convenience of online attendance, which did not impede their ongoing participation, and assisted with establishing connections with other participants. This then challenges the need for in-person participation to be effective and enables equal consideration of the option of online delivery for time-poor health professionals for whom self-care is often a secondary consideration to care of others. A systematic review supports this, finding practical benefits for individuals from a number of small and relatively short training programs, like reducing staff stress and sick leave as well as improving job satisfaction (Burton et al. 2017).

As has been noted (Feldman and Kuyken 2011), the focus of compassion tends to be external and directed toward others, with an intention to alleviate their suffering, whereas the focus of this program was on the internal self, identifying one’s responses to situations of struggle/suffering and attending to it with kindness and care. Such programs provide a vehicle for acknowledging and alleviating one’s own suffering, giving back to oneself, recognizing that all of us have struggles in life. Mills (2021) suggests that while self-compassion is the basis for a healthy and sustainable workforce, it is under-researched.

The results showed that self-compassion scores were at a moderate level at the commencement of the program and increased significantly after the training, indicating that the changes that did occur were picked up in a short time (post-training) and then increased again at the 3-month measure. Arguably, the self-compassion skills cultivated during the program endured over time. While Neff’s study (2020) suggests that those with lower baseline self-compassion scores gain the most benefit from cultivating ways to be kinder to themselves, this study supports the notion that regardless of where one sits on the spectrum of dispositional self-compassion, enhancements to one’s capacity for self-compassion may occur as a result of skills training, positively shifting the experience of well-being.

While we did not ask about the length of time working in palliative care, our participants were older, and Peters et al. (2013a) found that older clinicians experienced significantly less stress than younger nurses. A focus on self-compassion reduces the stress inherent in caring roles, providing a significant skill to guard against burnout (Neff et al. 2020).

It is unsurprising that compassion toward others’ scores altered little over time from the high initial baseline, since for these health professional participants, this characteristic may have been a motivator for their choice of specialization, as well as fundamental to being an effective carer. There seems to be an inbuilt protection against stress, arising from the choice to work in a chosen specialization (Peters et al. 2013b). Moreover, it has long been known that those who work in palliative care experience less stress (Vachon 1995). Additionally, dispositional empathy was moderately high in these participants, which may protect them against stress and burnout.

The Professional Quality of Life Scale was used to measure compassion satisfaction, burnout, and secondary traumatic stress. Secondary traumatic stress scores indicated a relatively low measure, meaning that this was not an issue for participants. But surprisingly, given the other scores discussed above, burnout scores were in the middle, indicating that participants neither had positive feelings about their work nor scored high enough to be at risk of burnout. Further analysis of the questions asked in this measure shows that many of the questions were related to experiences of trauma and thus may not be applicable to the setting of this study. It is noted that Neff et al. (2020) came to a similar conclusion and used Maslach Burnout Inventory (Maslach et al. 1996) to measure burnout and fatigue in subsequent studies.

Participants’ self-reported emotional state (depression, anxiety, and stress levels) showed a positive improvement 3 months post-training. This perhaps indicates the effectiveness of the program, in teaching regulation of one’s inner state, as well as responses to others, by developing and practicing self-compassion skills.

However, given the unexpected and unprecedented burden of care created by COVID-19 limitations, it was noted that compassion satisfaction scored highly throughout the program and was maintained over the 3-month period. Higher compassion satisfaction scores indicated more positive feelings related to work. Mindfulness was significantly higher post-training, and there were statistically insignificant changes in compassion satisfaction, burnout, and dispositional empathy. White (2018) suggests that palliative care routinely involves complex relational experiences and that activities like mindfulness is a valuable approach to use in such difficult encounters. The high compassion satisfaction score indicates that participants were positively disposed toward their work, even amidst the resulting stressful changes in practice because of COVID-19. Learning skills like mindfulness and kindness to self in the training might have assisted with a further decrease in levels of traumatic stress that were maintained over the time period.

As noted, the overt supportive team culture together with an active death education program, both often found in palliative care settings, may assist in protecting clinicians against the distressing impacts of caring (O’Connor and Peyton 2015; Peters et al. 2013a, b). Like Neff et al.’s (2020) study, the group represented a range of disciplines, forming a virtual community over the time of the program. Additionally, multidisciplinary decision-making and specialized palliative care training/education may also provide a buffer (O’Mahony et al. 2016), and Cross (2019) suggests that self-care resources, as well as organizational activities like staff bereavement programs, may contribute to decreased stressors. For congruence, it is important to promote an organizational culture of compassionate care of self and others that reflects individual self-compassion (Garcia et al. 2021).

The results in this paper support those of Neff’s study (2020), that while being unable to separate the sufferings of others from one’s own well-being, the development of a self-compassionate approach could mitigate against the stressors of caring roles. Additionally, such training may enhance individual well-being and reduce fatigue and/or burnout (O’Mahony et al. 2016). Self-compassion and staff support need to go hand in hand, a shared responsibility between the individual health-care professional and their workplace management.

This pilot project indicates that a short online self-compassion program is effective, assisting participants to cultivate inner resources including self-compassion mindfulness and emotional regulation, to balance outward care and compassion toward those receiving palliative care and against the difficulties caused by the COVID-19 pandemic. These results align with other research, in that such training may assist an individual’s sense of well-being and reduce their fatigue and/or burnout. Expansion of this training to a broad offering as part of ongoing organizational education may benefit the wider palliative care workforce.

Limitations

This study has several limitations. It reports on a pilot project with very few participants who all self-selected, meaning that they had a preexisting interest in exploring self-compassion. A small sample size means our study might have been under power to detect the impact of the intervention on dispositional empathy and compassion toward others. Demographic data related to ethnicity were not collected and may have provided insight into the importance (or not) of self-compassion in different cultures.

All participants worked in palliative care, a sector that encourages self-care, so this may have assisted individuals in knowing that even under pressure, they were able to make a difference for individuals facing the end of their life. Thus, perhaps, many of the scores remained statistically unchanged because the supportive clinical setting means less burnout and depersonalization, but the small sample size inhibits further comment. Moreover, all participants were middle-aged and female, which, as noted by Neff et al. (2020), is the typical population who express interest in self-compassion. There would be benefits in exploring how to widen the appeal of self-compassion training to attract more diverse participants, in age, ethnicity, experience, and gender and supported by a positive workplace culture.