An online mindfulness intervention to enhance compassion in nursing practice: A feasibility and acceptability study with nursing students

Abstract

Background

Compassion is an important component of nursing care, but public enquiries into care failures have noted that it is not always evident. Mindfulness interventions have the potential to support compassion. However, the feasibility of delivering a mindfulness intervention at scale to nursing students has not been established.

Objectives

To develop and test the feasibility and acceptability of a tailored mindfulness based online intervention to foster compassion in nursing students within clinical practice.

Design

A randomised feasibility study with a waiting list control.

Setting

A UK nursing faculty within a large university.

Participants

Post-graduate, post-registration and pre-registration nursing students (N = 77).

Methods

An online five module mindfulness based intervention (Mindful Nursing Online) was developed and tailored to support compassion in clinical nursing. The feasibility study comprised 77 participants randomised in a 2:1 ratio into an immediate access group (intervention, n = 50) or a delayed access group (waiting list control, n = 27). Data on feasibility through completion, attrition and practice rates, were collected through follow-up questionnaires at post-intervention, and 14 and 20 weeks after baseline. Acceptability data was collected through semi-structured interviews with 12 participants.

Results

Non-completion rates were high with all five modules completed by only 28% (n = 14) of participants, and three modules completed by only 46% (n = 23). The most commonly cited reason for non-completion was lack of time. However, the interview data suggested those who completed the intervention were using mindfulness techniques in practice. They described feeling less stressed on the ward, having an increased focus on patients and a greater appreciation of the importance of self-care. Evaluative feedback therefore showed that the intervention was perceived to be effective at promoting mindfulness skills and was relevant to nursing work.

Conclusion

Minimising attrition and enhancing engagement with the intervention should be the key objectives of a future study. Feedback from participants who completed the intervention indicates that a brief mindfulness intervention delivered online may support the delivery of compassionate nursing care.

What is already known about the topic?

• •There are a number of factors which can hinder compassion in healthcare, and which reflect the challenging nature of the role and workplace pressures.
• •There has been only limited research to develop and evaluate educational interventions which support compassionate nursing care within the realities of clinical practice.

What this paper adds

• •A mindfulness-based intervention tailored to nursing practice is a promising way of supporting compassion and addressing factors which undermine compassion in daily practice.
• •Using online delivery offers flexibility in training and can increase nursing students’ access to learning mindfulness techniques.
• •There is a need for further research to determine how to maximise nursing students’ engagement in and completion of online mindfulness interventions.

1. Introduction

Empirical evidence suggests that compassionate care can improve patient experience and increase the likelihood of better patient outcomes (Post, 2011; van der Cingel, 2011). Compassion is also considered a central tenet of nursing care (Chambers and Ryder, 2009). However, high profile care failures suggest that compassionate care does not always occur (Francis, 2013). Indeed, there have been several campaigns and policy initiatives internationally to promote the importance of compassion within healthcare in the US (Lown, Rosen, and Marttila, 2011), New Zealand (Youngson, 2012) and the UK (Cummings and Bennett, 2012; Francis, 2013). Such campaigns have had a mixed response from staff; for example, in the UK, whilst many of the values and aims within the “Compassion in Practice” strategy (Cummings and Bennett, 2012) have been successfully implemented across services, some nurses appear sceptical of the largely top-down approach of the campaign and of its failure to recognise the considerable organisational barriers to providing compassion in practice (Serrant et al., 2016, O'Driscoll et al., 2018)

Multiple factors appear to hinder the delivery of compassionate care in healthcare settings. Low staffing levels can limit nurses’ ability to dedicate time to creating rapport and trust with patients (Ball et al., 2014). Some healthcare delivery settings have an excessive focus on performance and financial targets, which can lead staff to feel that they themselves are not treated by their organisation in a compassionate way. This focus can increase workers’ stress and can be to the detriment of patient care (Cole-King and Gilbert, 2011; Firth-Cozens and Cornwell, 2009). Caring itself can have a psychological impact on professionals which can also hamper compassionate responding. Witnessing suffering on a continual basis can lead to stress-related syndromes such as burnout or compassion fatigue (Figley, 1995; Maslach and Goldberg, 1998; Peters, 2018). Similarly, unsupported, the emotional labour to regulate nurses’ own negative emotions in practice can lead to exhaustion (Delgado, Upton, Ranse, Furness, and Foster, 2017; Hochschild, 1983).

In terms of interventions for healthcare professionals, Blomberg, Griffiths, Wengstrom, May, and Bridges (2016) reviewed 25 compassionate nursing interventions and concluded through narrative synthesis that most studies showed significant benefits of the interventions in relation to the outcomes measured. A review by Durkin, Gurbutt, and J. (2018) of teaching compassion to nursing students found only two such teaching interventions. A wider scoping review by Sinclair et al. (2016) of the compassion literature included a description of ten interventions to promote compassion to others (and self) in healthcare professionals and students. The reviews identified a range of interventions which were mostly face-to-face, and included evaluations of role-play, communication training, journaling, simulations, new models of care and mindfulness. Blomberg et al. (2016) referred to the low quality of most of these studies which were not replicable, often quasi-experimental and lacked theoretical frameworks. Sinclair et al. (2016) recommended that compassion training should be individualised to students and their existing qualities through a process of reflexive and experiential learning.

Given the emotional and practical resources required to deliver compassionate care (Tierney, Bivins, and Seers, 2018) mindfulness may be a promising intervention for supporting staff working in challenging organisational environments. Marlatt and Kristeller (1999, p68) describe mindfulness as a way of attention training, of “bringing one's complete attention to the present experience on a moment to moment basis.” Although there are a number of mindfulness models which propose different mechanisms of action, most share the concepts of enhancing attention, increasing self-awareness, de-centring (changing how we view our thoughts) and improving emotional regulation (Bishop et al., 2004; Hölzel et al., 2011; Shapiro, Carlson, Astin, and Freedman, 2006). Meta-analyses have suggested that regular practice of mindfulness, largely through meditation, has positive effects in terms of reducing stress, depression and anxiety (Demarzo et al., 2015; Goyal et al., 2014; Kuyken et al., 2016). Reviews of mindfulness intervention studies with nurses and nursing students have also confirmed that it can reduce their stress and anxiety (Botha, Gwin, and C., 2015; Smith, 2014; Van Der Riet, Levett-Jones, and Aquino-Russell, 2018). There is also a limited amount of research which suggests that mindfulness may increase empathy and focus on patient-centred outcomes in healthcare staff and students (Barbosa et al., 2013; Irving, Dobkin, and Park, 2009; Krasner et al., 2009; Shapiro and Schwartz, 1998).

Although the exact mechanism by which mindfulness fosters empathy and compassion is unclear, the literature suggests at least five hypothetical pathways through which mindfulness could foster compassion in nursing practice. Additional pathways might be identified, but those listed below provided the conceptual and mechanistic rationale behind the intervention described in this paper. A mindfulness intervention could support compassionate practice by:

• 1)Increasing awareness in the practice setting which could improve professionals’ awareness of and sensitivity towards what patients need in busy environments (Hölzel et al., 2011, Youngson, 2011);
• 2)Increasing empathy through a greater awareness of ones’ own emotions, which is thought to be a precursor to compassion (Block-Lerner, Adair, Plumb, Rhatigan, and Orsillo, 2007);
• 3)Strengthening motivation to be compassionate to others, through mindfulness and compassion enhancing meditative practices (Gilbert et al., 2009; Leiberg, Klimecki, and Singer, 2011);
• 4)Supporting self-compassion leading to a more optimistic appraisal of one's own resources, and a recognition of the commonality of human suffering (Cole-King and Gilbert, 2011; Neff et al., 2013);
• 5)Improving emotional regulation, and thus lowering stress levels known to have a negative impact on empathy and compassion (Firth-Cozens and Cornwell, 2009).

Previous research with healthcare professionals (Beddoe and Murphy, 2004; Cohen-Katz, Wiley, Capuano, Baker, and Shapiro, 2004; Shapiro, Astin, Bishop, and Cordova, 2005; Van Der Riet, Rossiter, Kirby, Dluzewska, and Harmon, 2015) suggests that attrition from mindfulness courses is an issue. Staff generally report insufficient time to attend sessions. Most standard courses usually last at least 26 h over a period of eight weeks (Kabat-Zinn, 1982). However, research suggests that mindfulness-based programmes can be abbreviated without reducing their effects on measures of physical and psychological health (Carmody and Baer, 2009; Cavanagh et al., 2013; Gilmartin et al., 2017). Online delivery may solve the problem of erratic attendance of nursing students and healthcare professionals at face-to-face mindfulness sessions. Despite this potential, the use of online methods to deliver mindfulness training to nursing students has rarely been explored.

The objectives of the randomised, mixed methods evaluation study reported in this paper were to investigate: 1) the feasibility of delivering the online intervention; 2) its acceptability to nursing students; and 3) the feasibility of the trial design and outcome measures to inform a further effectiveness trial.

2. Methods

2.1. Development and description of the intervention

The intervention was developed using the Medical Research Council Framework (Craig et al., 2008) which provides a systematic approach to developing complex interventions. It was developed by a team comprising a researcher, a mental health nurse, and experts who were qualified to teach mindfulness within the NHS and higher education settings (SC, IN, MC and RLG). As part of the development process, stakeholders’ views on appropriate content were also gathered through a small focus group with post-registration nursing students (n = 5) and further consultation with nursing academics. The usability of the first prototype was tested with a sample of qualified nurses and nursing students (n = 6), and a final version was developed incorporating this feedback. The intervention was developed using LifeGuide, a software initiative funded by the Economic and Social Research Council (UK) to develop interactive online healthcare interventions.

Table 1 and the accompanying description below is guided by the template for intervention description and replication checklist (TIDieR) (Hoffmann et al., 2014). All five of the modules comprised didactic information on compassion and mindfulness, and each was framed by common challenges in nursing. This was to increase salience and provide practical approaches to implementing mindfulness into nursing practice. Table 1 shows the informal mindfulness practices which were suggested. These were brief meditations which could be done as part of the working day. For example, participants were advised to use STOP and take a mindful pause when going to the pharmacy, or before entering a patient's room. Formal practices comprised longer guided meditations lasting between 10 and 20 min. For these, participants were directed, with permission, to a website associated with a self-help mindfulness book, “Finding Peace in a Frantic World” (Williams and Penman, 2011) which provides free, downloadable meditation files.

Table 1.

Intervention content by module.

Title	Nursing challenge	Objective	Formal and informal practices
Module 1: Introduction	Nursing on autopilot	Orientate nurses to the practice of mindfulness	Body scan Mindful Hand washing
Module 2: Focusing attention & awareness	Multi-tasking in nursing and interruptions	Promote greater awareness in practice to noticing suffering and how to respond effectively	Body Scan Mindful walking at work and other work-based examples STOP (Stop, Take a breath, Observe, Proceed) or Pause practice
Module 3: Stress!	Stress of nursing: both organisational and personal challenges	Focus on stress reduction (emotional regulation) to allow more compassionate responding to occur	Body and Breath Meditation Three minute-breathing space and other work-based examples
Module 4: Being with distress	Emotional overload of nursing role	Foster the ability to tolerate difficult or negative emotions to provide compassion without being overwhelmed	Sitting meditation (Sounds and thoughts) Mindful movement with work-based examples
Module 5: A Compassionate approach	Trying to deliver compassionate care	Promote self-compassion and self-care) and generate motivation for compassion to others	Loving Kindness Meditation Self-compassion breaks at work Gratitude practice

Each module took between 30 and 60 min to complete and was released on a weekly basis. Participants accessed the modules one to five sequentially, and were unable to access the next module until they had completed the previous one. They were always able to review previously studied modules. They were able to email mindfulness experts with practice questions or an administrator for technical issues. There was also the opportunity to join a peer support forum to connect to other participants. Each module ended with a resources page of links to academic papers, news articles and relevant videos on the internet to support module learning. Standardised emails were sent on module completion, when a new module became available, and as a reminder to access a module or complete follow up assessments. Participants were also emailed weekly with motivational tips. Each module contained a frequently asked questions page on practice itself (e.g. becoming distracted or getting bored) and a list of common barriers to practising (e.g. finding time and losing motivation) and how to overcome them.

2.2. Evaluation of the intervention: Study design and data collection procedures

The intervention was tested for feasibility using a randomised design, featuring a wait list control, with a 2:1 randomisation bias to ensure that sufficient numbers were randomised to the intervention group to provide feedback. Participants registered online and then completed baseline data. They were then randomised automatically and informed about their allocation through the intervention software, which was followed later by a confirmatory email. Participants in the delayed access group (waiting list control) were given access to the intervention at the end of the study at 20 weeks. Participants were recommended to complete the intervention within 8 weeks but could access it after this time and up until the end of the study. Ethical approval was obtained from the university's ethics committee prior to the development and testing of the intervention (PNM:HR-15/16–273).

2.3. Participants

Participants were eligible if they were either second or third-year pre-registration nursing students or post-registration students. There were no exclusion criteria for second and third year students. First year students were excluded as they were not considered to have enough clinical experience to make use of the techniques in practice. Although the intervention was developed as an introduction to mindfulness, students with previous mindfulness experience were still eligible to participate. Recruitment lasted six months via information posted on student online learning boards and posters within the Faculty. Participants were informed that the study was an independent research project and that choosing to participate or not would have no effect on their studies. Participants in the immediate access and delayed access groups who completed their final follow up evaluations were offered £10 and £5 vouchers respectively. Completion certificates were sent to those who completed module five, and interviewees were offered a £10 voucher for their time. Although not powered, the target sample size for the evaluation study was 90, assuming a high attrition rate of up to 50%. This was informed by previous high dropout rates in studies of online mindfulness interventions (Cavanagh et al., 2013). Even with this level of dropout, there would be sufficient numbers in each group for the purposes of testing the feasibility and acceptability of the intervention (Julious, 2005).

2.4. Evaluating the feasibility of the intervention

2.4.1. Completion rates

Data were collected on the percentage of people who: 1) did not begin the first module; 2) completed five modules by the end of the study period; and 3) completed three or more modules by the end of the study period. Participants who completed at least three of the five modules were considered to have had sufficient exposure to the intervention (or dose) which is akin to the threshold of 50% (or four out of eight sessions) set by previous mindfulness studies (Crane and Williams, 2010; Dimidjian et al., 2014; Kuyken et al., 2010).

2.4.2. Practice rates

Data on the amount of practice (formal and informal) which occurred during the previous week. This was collected through participant self-rating at the beginning of each new module, and at 14 and 20 weeks follow up. At each timepoint, participants were asked how many days per week they carried out both formal and informal practice (not at all, one to two days, three or four days, or more than four days). As research suggests that practising mindfulness for at least three days per week can have an impact on psychological outcomes (Crane et al., 2014; Perich, Manicavasagar, Mitchell, and Ball, 2013) participants were subsequently categorised into those reporting either moderate to high practice (three days per week or more) or low practice (fewer than three days) at each of the time points.

2.5. Evaluating the feasibility of the trial design/measures

The feasibility of the trial design was established through collecting data on the number of: 1) registering participants who wanted to be allocated to the particular groups (immediate or delayed access) at allocation; 2) control participants who engaged in mindfulness practices whilst waiting for the intervention, which could underestimate effect sizes in a future trial; 3) control participants accessing the intervention after 20 weeks to determine potential data loss from this group in an effectiveness trial and 4) The percentage of people providing follow up outcome data post-intervention (either after the final module, or at eight weeks if the intervention was not completed by then) and at 14 weeks and 20 weeks after registration. In addition, several possible psychological outcome measures were selected for potential use in a future effectiveness trial. They were chosen either because they addressed direct outcomes of the intervention (i.e. less fear of providing compassion for others) or the hypothesised pathways by which these direct outcomes might be achieved (i.e. greater mindfulness, reduced stress, increased empathy, increased compassion satisfaction, and increased compassion for self). Thus, they all had potential for use as primary and secondary outcome measures in a later effectiveness study.

2.6. Fear of compassion for others scales (FCS) (Gilbert, McEwan, Matos, and Rivis, 2011)

As part of their work with clinical populations, Gilbert et al. (2011) developed three scales to measure a person's tendency to resist engagement in compassionate behaviour towards others and oneself due to fear of personal consequences (e.g. stress, guilt or burden). The FCS comprises three sub-scales (fear of expressing compassion to others, fear of expressing compassion to self and fear of receiving compassion from others). The fear of expressing compassion to others scale comprises ten items (e.g. “I worry that if I am compassionate, vulnerable people can be drawn to me and drain my emotional resources). All items are rated on a five-point Likert scale from 0 to 4, with lower scores indicating that the respondent has a low fear of expressing compassion for others. A Cronbach's alpha of 0.78 was reported within student and therapist populations (Gilbert et al., 2011). The fear of compassion for self scale comprises 15 items (e.g., ‘‘I feel that I don't deserve to be kind and forgiving to myself’’). It has a minimum score of 0 and a maximum of 60 with lower scores indicating that the respondent has a low fear of expressing compassion towards themselves. Gilbert et al. (2011) reported good internal reliability with a Cronbach's alpha of 0.85 for this sub-scale.

2.7. Interpersonal reactivity index (IRI) (Davis, 1983)

The IRI (Davis, 1983) is a multi-dimensional measure of dispositional empathy comprising four, seven-item sub-scales which can be used independently. Only the seven-item empathic concern sub-scale (EC) which measures the extent to which a person has sympathy and compassion for others was used. The EC sub-scale comprises items such as, “I often have tender, concerned feelings for people less fortunate than me.” Overall scores for this sub-scale range from a minimum of 0 to a maximum of 28 indicating high levels of empathy. Internal consistency of the IRI with Cronbach's alpha ranged from 0.70 to 0.78 (Yu and Kirk, 2009). Test-retest reliabilities at 65–70 days ranged from 0.61 to 0.81 (Davis, 1983).

2.8. Mindful attention awareness scale (MAAS) (Brown and Ryan, 2003)

The Mindfulness Attention Awareness Scale (MAAS) conceptualises mindful awareness in everyday experiences as a state or characteristic that varies between people and can be cultivated with practice. Typical statements are: “I find it difficult to stay focused on what's happening in the present.” It comprises ten questions on a six-point Likert scale. A mean score is calculated to provide a scoring range of one to six, with higher scores indicating that a person tends to be more mindful. On original validation, the MAAS had good internal consistency, with Cronbach's alpha of 0.82, and 4-week test–retest reliability of r = 0.81 (Brown and Ryan, 2003).

2.9. Professional quality of life scale (ProQOL)(Stamm, 2010)

The ProQOL is a widely used measure of compassion fatigue and satisfaction. It is a 30-item self-report measure comprising three discrete subscales: 1) compassion satisfaction (pleasure derived from providing compassion); 2) burnout; and 3) secondary traumatic stress (or compassion fatigue.) Respondents indicate how frequently they have experienced each item in the previous 30 days, on a 5-item Likert scale. Higher scores represent high levels of satisfaction, burnout and fatigue. The compassion satisfaction scale has a Cronbach's alpha of 0.88, the burnout scale of 0.75, and the compassion fatigue scale of 0.81.

2.10. Perceived stress scale (PSS) (Cohen, Kamarck, and Mermelstein, 1983)

The Perceived Stress Scale (PSS) (Cohen et al., 1983) is a well validated and utilised ten item scale (Lee, 2012). The scale measures the frequency of stress within individuals over the last month using a five-point Likert scale. The total score ranges from 10 to a maximum of 40 which indicates high levels of stress. The PSS version 10 has good psychometric reliability with Cronbach's alpha of 0.78 reported (Cohen et al., 1983).

2.11. Acceptability of the intervention

Participants were asked to submit evaluations of each module. The evaluations were ad hoc measures developed by the authors to fit the needs of the feasibility study in the absence of appropriate, validated acceptability measures. The questions were scored using a Likert scale from 1 to 5 and are listed in Table 3. Open response options to provide more information relating to the ratings were available for all items.

Table 3.

Outcome mean scores with SD at baseline, post-intervention and at 14 and 20 weeks follow up both by group and overall.

Measure	Overall Immediate access group					Delayed access group
	Baseline - registered participants N = 86 +	Baseline n = 50	Post-intervention n = 22	Fourteen Weeks n = 16	Twenty weeks n = 19	Baseline n = 27	Post-intervention n = 14	Fourteen Weeks n = 11	Twenty weeks n = 12
Compassion for others (FCS) Min=0 to max=40 indicating high fear	12.50 (SD = 6.99)	12.30 (SD = 7.04)	14.13 (SD = 9.3)	11.88 (SD = 8.41)	10.21 (SD = 7.4)	13.04 (SD = 7.34)	12.78 (SD = 4.50)	14.27 (SD = 8.46)	10.58 (SD = 8.36)
Self-compassion (FCS) Min=0 to max=60 indicating high fear	21.34 (SD = 11.32)	19.42 (SD = 11.09)	13.44 (SD = 9.95)	11.06 (SD = 9.84)	10.21 (SD = 6.56)	24.00 (SD = 10.74)	21.64 (SD = 11.02)	24.18 (SD = 13.60)	23.83 (SD = 13.7)
Empathic concern (Subscale IRI) Min=0 to max=28 indicating high empathy	21.9 (SD = 3.78)	22.0 (SD = 3.83)	20.17 (SD = 4.26)	19.41 (SD = 4.12)	20.89 (SD = 3.91)	21.82 (SD = 3.62)	22.43 (SD = 2.93)	++	22.08 (SD = 3.56)
Mindfulness (MAAS) 1 to 6 High scores = greater mindfulness	3.17 (SD = 0.87)	3.32 (SD = 0.86)	3.61 (SD = 0.91)	3.67 (SD = 0.77)	3.78 (SD = 0.96)	3.01 (SD = 0.84)	3.13 (SD = 0.69)	3.28 (SD = 0.81)	3.54 (SD = 0.69)
Stress Scale (PSS) Scores 10–40 High scores=higher stress	22.31 (SD = 5.51)	21.40 (SD = 5.57)	16.69 (SD = 6.28)	17.29 (SD = 4.59)	16.42 (6SD = 0.12)	24.03 (SD = 5.43)	20.38 (SD = 5.67)	19.45 (SD = 4.94)	19.33 (SD = 6.24)
Satisfaction (PROQOL) 22 (low), 23 to 41 (average) 42 or above (high)	Intervention only	38.41 (SD = 5.21)	38.18 ( (SD = 6.03)	37.31 (SD = 5.90)	38.68 (SD = 6.90)	Not collected from control*	47.90 (SD = 5.81)	39.18 (SD = 6.01)	40.750 (SD = 5.46)
Burnout (PROQOL) Less than 22 (low), 23 to 41 (average) 42 or above (high)	Intervention only	24.70 (SD = 6.04)	27.13 (SD = 4.42)	25.25 (SD = 5.39)	25.26 (SD = 5.81)	Not collected from control*	23.39 (SD = 5.60)	24.46 (SD = 5.20)	25.33 (SD = 5.8)
Fatigue (PROQOL) Less than 22 (low), 23 to 41 (average) 42 or above (high)	Intervention only	22.59 (SD = 5.19)	22.04 (SD = 5.49)	20.19 (SD = 5.47)	20.48 (SD = 5.53)	Not collected from control*	22.08 (SD = 4.11)	23.64 (SD = 4.92)	22.92 (SD = 5.52)

+Registered users including those not randomised.

*ProQOL was collected at end of module 2 not at baseline to reduce questionnaire burden on intervention group. Control group could not complete ProQOL at this second time point.

++ Error missing variable.

All participants were also offered the opportunity to be interviewed, either at the end of the final module if they completed this, or alternatively at the end of the study. All participants who responded to the invitation were interviewed. Telephone interviews were conducted by a member of the research team (SC). The interview schedule was constructed by the research team to answer key questions of acceptability. It began by focusing on the acceptability of the intervention itself, including participants’ motivation and expectations when registering for the intervention. In the second part, questions focused on the acceptability of the mindfulness practices, and any perceived impact of these. There was a clear distinction in the interview questions between the intervention itself and mindfulness practices. It was possible that participants might be enthusiastic about mindfulness or the intervention as a learning tool, without necessarily valuing both. The items were piloted for clarity and minor changes made. The final interview schedule can be found in Appendix 1.

2.12. Data analysis

Descriptive statistics were used to examine the mean acceptability scores and the appropriateness of the psychological measures by comparing baseline scores against normative values for nurses and other populations. The interview data were analysed using a thematic analysis process outlined by Braun and Clarke (2006) due to its flexibility of being untied to a particular theoretical stance. All interviews were transcribed by the researcher (SC) as the first stage of data familiarisation and transcriptions were imported into NVIVO (version 11).

The analysis comprised five main stages: 1) familiarisation with the data which was achieved by transcribing and then re-reading transcripts several times; 2) generation of initial codes to organise the dataset into meaningful groups; 3) identification of themes, through locating different codes within broader themes, and then collation of all the relevant coded data extracts into these identified themes; 4) a review of themes to check that the coded data reflected each theme and these themes reflected the overall meaning of the dataset and finally 5) the naming of themes for further interpretative analysis, and an assessment of how they reflected the narrative of the overall dataset. Initially, two transcripts were read and then coded using procedures 1–3. To ensure objectivity, the preliminary coding scheme was discussed a second researcher (IJN), and after some discussion, the coding framework was revised.

3. Results

3.1. Sample

Eighty-six people registered for the intervention although only 77 completed randomisation. Fifty participants were randomised to an immediate access group and 27 to a delayed access group. Overall, nearly two thirds of the sample (63.6%, n = 49) were between the ages of 18–30. The majority (90.9%, n = 70) described themselves as female. The most frequent nursing speciality was adult (general) nursing, followed by mental health and child health nursing. The largest proportion of students (70.1%, n = 54) were pre-registration, with the remainder describing themselves as qualified nurses studying for a postgraduate qualification. Most participants (55.8%, n = 43) had no previous experience of mindfulness, whilst 42.8% (n = 33) had a little experience (e.g. reading a pamphlet/book, looking at a video, etc.). Table 2. shows the characteristics of the participants who were randomised into each group. However, since an effectiveness analysis was not planned, it was unnecessary to test for statistically significant differences in baseline characteristics across the groups (Arain, Campbell, Cooper, and Lancaster, 2010).

Table 2.

Baseline characteristics of participants in feasibility study.

Characteristics N = 77	Immediate access group n = 50	Delayed access group n = 27
Age 18–30 30–50 50–65	27 (54.0%) 20 (40.0%) 3 (6%)	22 (81.5%) 5 (18.5) 0
Gender female	46 (92.0%)	24 (88.9%)
Speciality adult mental health child health primary care health visiting other	26 (52%) 12 (24%) 5 (10%) 1 (2%) 0 6 (12%)	17 (63.0%) 2 (7.4%) 4 (14.8%) 0 1 (3.7%) 3 (11.1%)
Course BSc (2nd and 3rd yr.) PGDip Postgraduate Post registration Missing	19 (38%) 15 (30%) 11 (22%) 5 (10%) 0	15 (55.6%) 5 (18.5%) 5 (18.5%) 2 (7.4%) 0
Time since qual Not qualified Less than 1 year 1–5 years 6–10 years Over 10 years	23 (46%) 2 (4%) 12 (24%) 6 (12%) 7 (14%)	13 (48.1%) 4 (14.8%) 7 (25.9%) 3 (11.1%) 0
Previous mindfulness None A little A lot	26 (52%) 23 (46%) 1(2%)	17 (63%) 10 (37%) 0

3.2. Feasibility of the intervention

Eighteen percent (n = 9/50) of those randomised to the intervention group never began the intervention. Twenty-eight percent (n = 14/50) of the total sample who were randomised completed all five modules, with 46.0% (n = 23/50) completing at least three modules. amongst the minority of participants (28.0%, n = 14/50) who submitted reasons for non-completion, the most commonly reported were insufficient time, breaking the continuity of practice, and the presence of competing events. See Fig. 1 for participant flow through the study.

Fig. 1.

Chart to show participant flow through the five-module intervention.

3.3. Practice rates

Between weeks one and two, 51.7% (n = 15/29) of participants reported moderate to high formal practice and 68.9% (n = 20/29) practised informal techniques. Between weeks two and three, 43.5% (n = 10/23) of participants practised formally at this level, with 60.9% (n = 14/23) reporting moderate to high use of STOP (informal practice) and 65.2% (n = 15/23) reporting this level of their own selected informal practice. Between weeks three and four, 47.6% (n = 10/21) reported moderate to high practice of formal techniques and only 42.9% (n = 9/21) used informal practices to this degree. In the final week, 64.3% (n = 9/14) reported practising the sounds and thoughts meditation at moderate to high frequency, and 57.1% (n = 8/14) reported using the three-minute breathing space at moderate to high frequency, whilst informal practice was undertaken by 64.3% (n = 9/14). Practice data were submitted at the start of each module before completion. Some participants submitted the practice log without completing the module, which accounts for the occasional differences between these two totals.

3.4. Feasibility of study design/measures

When asked before randomisation about their allocation preferences, 51.2% (n = 44/86) of the participants preferred to be placed in the immediate access group, with only 4.7% (n = 4/86) preferring the delayed group, and with 44.2% (n = 38/86) expressing no preference. 46.8% (n = 36/77) of all participants completed the post-intervention measures (after the last module or at eight weeks if they did not complete the final module before this time.) Whilst 35.1% (n = 27/77) completed 14 weeks (post-registration) follow up measures and 40.3% (n = 31/77) completed them at 20 weeks post-registration. In the delayed access group, at 20 weeks, 22% reported that they had accessed mindfulness resources during the wait period. Only three of the 27 participants in the delayed access control group chose to access the intervention at the end of the study period.

Table 3 provides a description of how the outcome measures varied across follow ups. This information is presented here purely to show the appropriateness of these specific measures with this sample. There was no evidence of floor or ceiling effects on any of the scales. Scores on the stress scale reduced by four to six points over time in both groups, although the reduction was larger in the immediate access group. The fear of self-compassion scale showed relatively large score changes, in a positive direction (between five to seven points post-intervention) in the immediate access group with smaller changes observed in the delayed access group. The fear of compassion for others scale showed a small rise (i.e. indicating a greater fear of providing compassion to others) post-intervention in both groups. This decreased again at 14 weeks after registration and remained lower than baseline at 20 weeks in the immediate access group.

3.5. Acceptability of the intervention

Table 4 shows that most modules received scores of between 3 and 4.5 (possible range 1–5) which suggests positive attitudes towards the intervention. Satisfaction means for modules ranged from 3.7 for module one, to 4.3 for module five. Means were between 4.0 and 4.5 for clarity of the modules, between 3.8 and 4.7 for usefulness of material for nurses, and 3.9 to 4.4 for relevance outside of work. Participants rated their confidence in understanding mindfulness between 3.8 and 4.5. The lowest ratings were for participants’ confidence in their ability to find time to practice, as these ratings ranged from 3.1 to 3.9. Although participants were able to email mindfulness experts with practice questions, none emailed the team to ask for help.

Table 4.

Acceptability scores (study specific measure) by module.

Acceptability questions Scores 1 = least positive to 5= most positive	M1 n = 34 Mean (SD)	M2 n = 27 Mean (SD)	M3 n = 23 Mean (SD)	M4 n = 17 Mean (SD)	M5 n = 14 Mean (SD)
How satisfied were you overall with the module?	3.7 (0.6)	3.8 (0.8)	3.8 (0.8)	4.2 (0.4)	4.3 (0.6)
How clear did you find the concepts and ideas in the module?	4.4 (0.7)	4.3 (0.7)	4.0 (0.90)	4.2 (0.7)	4.5 0.5)
How useful was the content to your work as a nurse?	3.9 (0.8)	4.1 (0.9)	3.8 (0.7)	4.5 (0.5)	4.7 (0.6)
How useful was the content to you personally, outside of work?	4.1 (0.9)	3.9 (0.9)	4.0 (0.8)	4.2 (0.5)	4.4 (0.7)
How confident are you in your understanding of the mindfulness techniques that you have just learnt in the module?	3.8 (0.8)	3.9 (0.8)	3.8 (0.8)	3.8 (0.6)	4.5 (0.6)
How confident do you feel about your ability to find time to practise mindfulness over the coming week?	3.3 (0.9)	3.4 (0.8)	3.1 (0.7)	3.5 (1.1)	3.9 (0.9)

3.6. Interviews: acceptability

All 41 participants who had accessed the first module were invited for interview. From this, twelve (29.2%) consented and were interviewed. All twelve had completed at least four modules, with the majority (75.0%, n = 9) completing all five. Seven were studying for post-registration and five for pre-registration qualifications. Eleven identified as female. All were between 18 and 45. Reflecting the structure of the interviews, three distinct themes could be identified in the data, each with its own set of sub-themes, namely: 1) expectations and factors influencing participants’ engagement with the intervention; 2) experience of completing the intervention and practising mindfulness; and 3) the impact of mindfulness on their nursing practice and other areas.

3.7. Intervention engagement

From the interviews, it appeared that engagement with the intervention was necessary for completion which then led to greater amounts of practice. The extent of engagement and persistence appeared to depend on several factors: 1) competing events, 2) expectations and 3) ease of use. In the following analysis, used names are pseudonyms.

Competing events: Events such as placement or exams had a negative impact on how much time and enthusiasm they could dedicate to the intervention.

“It was a really difficult time for me to do it because I was on placement and you don't have a routine – because you know you're doing shifts all over the place – you're doing nights – earlies….” (Jane, Pre-registration).

Not practising as planned made some lose their motivation:

“…I lost interest because I kinda felt like…I wasn't practising as much as recommended. I wasn't doing it properly.” (Claire, Post-registration).

Expectations: Participants who completed the modules also reported early investment in the techniques. A number had expectations that mindfulness could be beneficial for them. These were often formed by a need to deal with stress or challenges, their previous experience of trying to meditate, or through social influence coming from friends and family.

“I do meditation and yoga and I've done meditation retreats before now, but I don't seem to be able to apply that to a clinical setting.” (Imogen, Post-registration).

“I had a colleague who'd done mindfulness classes but I'd never gone to any – but she kind of told me a little bit about them and what she did in them – and it was quite different to how this course has been” (Jane, Pre-registration).

Ease of use: Ease of use was perceived as important. Participants commented on how they had found learning mindfulness online, and several issues were raised during this discussion. The majority had enjoyed learning online and had particularly appreciated the flexibility it afforded. For example, one participant said:

“I thought that having the online format was far more helpful. If it was a weekly session that you had to go to, I don't think I would have been able to do it as often as I did.” (Diana, Pre-registration).

Most participants liked learning online and independently.

“I think going through the materials made sense. I didn't really need any group input. I didn't feel I needed it and I was able to read it and do it wherever and whenever.” (Helen, Pre-registration nurse).

A handful of participants reported that it was sometimes tricky to put new practices in context, and that it might have been useful to have spoken to other participants to clarify how to practise them. Three participants suggested the idea of a hybrid course, which involved a face-to-face session to introduce or to finalise the intervention.

“It would be nice to have a de-brief afterwards as a group or an initial thing at the start of the course. Just to have that contact” (Imogen, Post-registration).

“Maybe at the end. Because I felt like for me…I really enjoyed it whilst I was doing it – but it would have been nice at the end if it had an end session – an opportunity to just discuss what they found helpful and what they hadn't found helpful and how they incorporated it into their practice” (Claire, Post-registration).

However, intervention factors which made the material more difficult to access (e.g. the lack of portability of information or limited mobile access) made it less likely that participants would engage. Two participants suggested that it would have been better as an app. One explained:

“If it had been an app on my phone I would have used it a lot more…. I think there is something about doing it on my own when there are people in the house. And I think that taking a laptop into a room is a bit too conspicuous!” (Carol, Post-registration).

Conversely, two other participants said that what they considered to be the “formality” of the laptop in comparison to their mobile phone, was more appropriate for their learning. None of the participants had emailed the mindfulness experts with practice questions. When asked about why they had not used this facility, all interviewees explained that they had found the instructions on how to practice sufficient for their needs.

Establishing a practice: Several participants identified approaches such as scheduling in advance, that helped them keep going with the intervention. One said:

“I had to do it at a regular time which is I know very difficult for nurses…when you're on nights to do it at a regular time - but I did it in the morning.” (Carol, Post-registration).

Another confirmed:

“The mindfulness was just trying to start a new habit – which is quite difficult at first ……..I had to give myself what I call protected time to do it – even if it wasn't the whole ten minutes – even if it was just five minutes – it was sort of five minutes for me and I actually quite appreciated that.” (Michelle, Post-registration).

The following four themes emerged from the data as effects of greater practice:

Self-care:

Several participants reported that the compassion meditations and self-care material had a beneficial impact on the way that they treated themselves. For two participants the idea of prioritising self-care was a revelation which had a significant impact on the way that they thought about themselves. One participant explained:

“As a nurse you feel this pressure to look after people, and if you don't do that then you're not a good nurse. That's what I felt the pressure was. But after that module, I felt to be a good nurse I have to look after myself, so it's not selfish, it's important to enable me to be a good nurse.” (Claire, Post-registration).

Another respondent agreed:

“…as a nurse you are always looking after other people and I think I've taken from the course that I need to start taking care of me as well because I matter just as much… and it makes me able to do my job properly.” (Michelle, Post-registration).

Emotional regulation:

An increased ability to regulate negative emotions, particularly stress, at home, university and in nursing practice was also mentioned. Several participants reported that they were developing a more considered approach to ward stressors, which made them feel better and improved the quality of the care they gave.

When I did mindfulness, I was able to see myself and I was a lot calmer, really think and assess what was going on which is a luxury because you don't always get that time on the ward. When I did, I felt that my shift was running much more smoothly, I was in charge, everything was going the way I needed things to go, and I was doing what I needed to do.” (Claire, Post-registration).

STOP and the 3-minute breathing space were considered useful tools to use throughout the day when clinical practice was tough. Another also spoke about being able to use the 3-minute breathing space to check in with herself after being with patients:

“It's making sure that I don't carry that emotional baggage with me – okay so write it down and then - if I need to, I can do a three-minute meditation right there – just to make sure that I'm okay.” (Michelle, Post-registration).

Awareness on the ward:

Four participants reported greater awareness in clinical practice; describing how they could come off autopilot to focus on each task as it was performed. This process resulted in paying mindful attention to their patient without distractions. One participant clarified:

“I feel like I can care for patients better – because if they ask me for something I will actually remember. Because sometimes you just get so busy and you're like “yes I'll do that” – and then you completely forget. Now I do remember.” (Diana, Pre-registration).

Another participant commented on how much more efficient at her role she had become since clearing her head of pervasive thoughts throughout the day and felt that it had improved her engagement with patients:

“If I'm having a one to one with a patient the STOP practice works well there. Then it's much easier to formulate what I'm going to say based on what they've said.” (Imogen, Post-registration).

3.8. Practice after the intervention: The presence of feedback loops

Participants who reported feeling a positive impact of the practices on their lives relatively early on, appeared to find it easier to persevere with the intervention. One participant considered:

“I don't know – you very consciously remind yourself of doing it and then as you practice it – I suppose it becomes more natural in some ways more than others – and eventually some of the ways of doing mindfulness became a natural way to respond to stressful situations and tiredness and overload.” (Helen, Pre-registration).

Not practicing as planned made some participants lose their motivation:

“I think that like I kind of dunno – I lost interest because I kinda felt like Cos I wasn't practicing as much as recommended…I wasn't doing it properly” (Carol, Post-registration).

Similarly, the more practice participants completed, the more likely they were to report establishing a habit, or a fondness for certain practices, which in turn led to more practise.

“And I think, when I finally learned how to incorporate it into what I was doing rather than just thinking okay I have to do this right now - I enjoyed it a lot more.” (Claire, Post-registration).

“The more it went on through the weeks, the more it became a bit of a habit, the more it became easier to slot into my day.” (Michelle, Post-registration).

However, although several respondents reported being very keen to maintain their practice because they had witnessed definite benefits to their lives, almost all those interviewed reported that practice levels were lower after having completed the intervention than during it.

“At the time five weeks seemed perfectly suitable but now the course is over, I've kind of fallen into a bit of a relapse. I have found it quite hard to continue with the practices without the daily organised structure of it.” (Diana, Pre-registration).

4. Discussion

The quantitative results from this present study suggested that although participants found the intervention largely relevant to their nursing work, they struggled to complete all five modules and to fit formal mindfulness practices into their lives. The qualitative data illuminated how a sample of these participants had engaged with learning mindfulness online. These participants reported how they had used the informal techniques to enhance their nursing practice, allowing them to slow down, focus more on patients’ needs, regulate their emotions, and in some cases, to be more compassionate to themselves. The literature suggests that such outcomes can lead to greater compassion to others (Block-Lerner et al., 2007; Shapiro and Schwartz, 1998; Sinclair et al., 2016). Indeed, some authors suggest that motivation for self-compassion may develop prior to compassion for others (Rodríguez-Carvajal, García-Rubio, Paniagua, García-Diex, and de Rivas, 2016).

The attrition rate for the intervention reported in this paper was very high, with 54% failing to complete at least three modules. Lost to follow up rates of 53.2% post-intervention and 59.7% at 20 weeks were also high. High rates of attrition in online mindfulness programmes are not uncommon in non-nursing populations, where online delivery is better established (Fish, Brimson, and Lynch, 2016). A meta-analysis by Spijkerman, Pots, and Bohlmeijer (2016) found that attrition from online programmes in a range of clinical and non-clinical populations varied between 8% and 65%. Thus, although face to face mindfulness studies can also suffer from significant attrition, with estimated mean drop-out rates of 29% (Nam and Toneatto, 2016), attrition from online studies can be higher (Fish et al., 2016). These high attrition rates in online mindfulness studies are somewhat counterintuitive given that remote delivery is designed to fit participants’ needs by reducing travel and time commitments. Reasons cited in the literature for non-completion of online mindfulness interventions include lack of time, lack of perceived benefit, and forgetting to practise (Banerjee, Cavanagh, and Strauss, 2017; Fish et al., 2016; Spijkerman et al., 2016), which were echoed by some non-completers in the present study.

The literature and this study suggest that persuading nursing students to remain engaged with an online mindfulness intervention, particularly one requiring home practice, may be a challenge. We attempted to make the time commitment required by this intervention relatively low. However, there is likely to be less motivation amongst nurses or nursing students with no outstanding clinical problems or psychological symptoms to alleviate, and with no obvious consequences to disengaging from the intervention. We suggest, given the nature of mindfulness it is probably not feasible to conscript participants into mindfulness practice; however, mindfulness techniques could be more formally integrated into the curriculum (McConville, McAleer, and Hahne, 2017). In addition to providing devoted time to practise, this may also increase the perceived relevance and importance of mindfulness for delivering nursing care. For post-registration nurses, the answer may be to attach credits as part of continuing professional development to encourage persistence and denote that such training is valued.

Improving the intervention may also reduce attrition. The qualitative data suggest that the intervention was generally well received by those who used it, and there were few suggestions on how it could be improved. However, as can be seen from the comments on ease of use, a few participants suggested that the intervention may have been easier to use as an app. Although this feeling was not universal, some material such as meditation practices could be made available through a mobile app, which could also allow the scheduling and recording of practice through calendar links. Three interviewees also suggested the potential for meeting up in addition to the online learning. The potential for virtual meet ups or face to face meetings either at the beginning or end of the intervention could be explored as a way of providing participants with additional encouragement to continue.

For significant benefits to accrue, participants need to practise mindfulness techniques in their everyday life. Both completion and practice rates are poorly reported in many studies (Gilmartin et al., 2017). There was evidence of significant self-reported practice in the present study; the number of participants reporting high to moderate formal weekly practice (e.g. more than three times a week) ranged from 43.5% to 57.1% over the course of the study. The proportion of participants reporting high to moderate weekly informal practice ranged from 47.6% to 69% of the sample. However, as both formal and informal practice was measured by self-report, it may have been subject to inaccurate recall or social desirability bias. Overall, informal practice appeared to be easier for nursing students to incorporate into their daily routine. Research on the impact of informal practice on psychological outcomes is in its infancy (Cavanagh et al., 2018), but an investigation into the contribution of informal practice alone to improving clinical care may be worthwhile.

4.1. Strengths and limitations

The strengths of this study were that an evidence-based intervention was developed using a systematic approach within a standard framework, guided by expert tutors and informed by the experience of clinical nurses. This development process was transparent, and the resulting intervention remains replicable and available for future testing with a larger population. Substantial qualitative and quantitative feedback was collected across the study to improve the intervention for future iterations. The feasibility and acceptability of the intervention was also thoroughly tested using a randomised design which also assessed future study procedures. The study showed that a waiting list control design may not be appropriate for future effectiveness studies, as a significant number of participants in the control group accessed other mindfulness resources during their wait time, and very few decided to join the study after the wait period was over.

The mean scores and standard deviations for all the psychological outcomes are provided to determine their appropriateness for use with similar samples of nursing students and to assist with the derivation of sample sizes in future effectiveness studies which use these measures. However, it is not appropriate to analyse between or within group changes in feasibility studies which are not powered to look at effectiveness (Lancaster, Dodd, and Williamson, 2004). With such a small sample, it would be impossible to determine whether changes between or within groups were due to chance findings through random measurement error, an intervention effect, or other variables extraneous to the intervention (e.g. placement or exams). This means that the intervention would now need to be tested in a larger effectiveness study to obtain evidence of an impact on compassion levels before it could be recommended for use in nursing education for this purpose.

A major limitation was that attempts to obtain feedback from non-completers through an online exit survey or interview, were unsuccessful. This could have potentially skewed findings as most of the quantitative and qualitative data collected were from those who had completed at least three modules and who may have been amongst the more enthusiastic. As none of the participants who struggled to complete the intervention consented to be interviewed, this limited the depth of important data that could be collected to explain high attrition and disengagement. Future research would need to determine more effective ways of eliciting the views of participants who find the intervention least helpful. This would help to ascertain whether their lack of engagement related to the intervention itself or to their initial perceptions of the benefits of mindfulness.

Although significant progress has been made in developing new ways of delivering mindfulness material (i.e. online, by video/webinar, or smartphone apps) for the general public and clinical populations, this has rarely been extended to healthcare professionals. Mindfulness interventions are also rarely adapted for use in clinical care. To date, there have only been two small quasi-experimental US studies evaluating online mindfulness interventions with nurses, which broadly support the findings in this paper. One delivered mindfulness training through a university online learning system and reported a decrease in participant stress levels (Spadaro and Hunker, 2016). A second evaluated a hybrid intervention of face to face attendance with back-up recordings of sessions accessible online (Sanko, McKay, and Rogers, 2016). Although the latter reported high dropout, participants’ stress levels declined and they also self-rated improvements in clinical performance, compassion and communication skills. The key aim of the research in this current paper was to develop a mindfulness intervention which supported compassion, and that was tailored specifically for nurses. Both face to face and online mindfulness interventions for particular groups (e.g. parents) or health conditions (e.g. cancer) comprise mindfulness materials adapted for the common or important issues that these groups encounter. It therefore seems reasonable to argue that those used by healthcare professionals should provide the same salience. If interventions are not targeted in this way, they may fail to persuade clinicians to invest time in them or allow them to transfer mindfulness effectively into their clinical roles. This study begins this process of targeting online mindfulness interventions for nurses.

5. Conclusion

The UK study reported in this paper builds upon previous research and further expands the scope of mindfulness-based interventions to address compassion and quality in clinical practice. Recent research suggests that compassion training practices may convey greater benefits in terms of prosocial behaviour and compassion towards others than standard mindfulness practices (Brito-Pons, Campos, and Cebolla, 2018; Hildebrandt, McCall, and Singer, 2017). Therefore, the inclusion of specific compassion practices, such as those in this current online intervention, may be central to interventions supporting compassion in the workplace.

Mindfulness interventions, which offer dual benefits of supporting oneself and enhancing one's care of others, offer a promising approach to engaging professionals in compassion training. The results from this study suggest that whilst many participants did not complete the intervention, those who did reported it to be beneficial. Future research will draw upon participant feedback to develop an optimised intervention and to reduce attrition rates, in preparation for a larger study.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.