Introduction
Mindfulness is a form of meditation that is characterized by learning to pay ‘attention in a particular way: on purpose, in the present moment, and non-judgementally’ (Kabat-Zinn, Reference Kabat-Zinn1994). It has been shown to be broadly acceptable and to be helpful across a range of clinical and non-clinical situations: reduced stress and improved forgiveness among college students (Oman et al. Reference Oman, Shapiro, Thoresen, Plante and Flinders2008); reduced psychological distress in a physical illness population (Reibel et al. Reference Reibel, Greeson, Brainard and Rosenzweig2001); long-term benefits to people with anxiety disorders associated with physical illness (Miller et al. Reference Miller, Fletcher and Kabat-Zinn1995); reduction of depressive symptoms in chronic depression (Kenny & Williams, Reference Kenny and Williams2007); reduction in distress associated with psychosis (Chadwick et al. Reference Chadwick, Taylor and Abba2005, Reference Chadwick, Hughes, Russell, Russell and Dagnan2009; Davis et al. Reference Davis, Strasburger and Brown2007; Abba et al. Reference Abba, Chadwick and Stevenson2008).
Carer distress is beginning to emerge as a potential focus for a Mindfulness-based approach (Epstein-Lubow et al. Reference Epstein-Lubow, Miller and McBee2006; Minor et al. Reference Minor, Carlson, Mackenzie, Zernicke and Jones2006), although a Psychlit search (August 2012) did not identify publications investigating its use among carers of mental health service users. It has been shown to improve treatment outcomes for training therapists engaging in its practice for themselves (Grepmair et al. Reference Grepmair, Mitterlehner, Loew and Nickel2007) and it helps to prevent burnout in mental health staff (Perseius et al. Reference Perseius, Kaver, Ekdahl, Asberg and Samuelsson2007). This feasibility study was additionally intended to promote a normalizing view of symptoms by placing key stakeholders in the delivery of care together in a group.
Methodology
The facilitator attended workshops and a 3-day retreat to learn about Mindfulness meditation, subsequently engaging in Mindfulness meditation practice three times a week for 3 months in preparation. The group consisted of eight weekly sessions of 1 hour duration, following the description of Chadwick (Reference Chadwick2006) in person-based cognitive therapy for distressing psychosis: establishing Mindfulness as a new relationship with distressing experience, aiming for acceptance of experiences rather than avoiding or being ‘lost’ in them. The approach is adapted for the psychosis population, conceptualizing the process as group therapy within which the practice is embedded and shortened. Prolonged silence is avoided; practice outside of the group sessions is not required; there is an emphasis on being kind to oneself and grounding awareness in the body and breathing sensations. The programme sequences were: (1) reacting versus letting go; (2) avoiding experience versus turning towards the difficult and (3) judgement versus acceptance.
Advice provided by Trust Research and Development was that this project was service development and ethical approval was therefore not required. Support was obtained from Trust management and a wide-ranging service-focused stakeholder Reference Group. An information sheet about Mindfulness and the study was circulated to the EIP clinicians and feedback sought via email and in team discussion. The priorities associated with developing this were discussed at the service Reference Group. The group comprised eight weekly sessions of 1 h duration. Evaluation was both qualitative and quantitative. A single case methodology was used to produce qualitative feedback, and a Clinical Outcomes in Research Evaluation (CORE-OM; Evans et al. Reference Evans, Connell, Barkham, Margison, Mellor-Clark, McGrath and Audin2002) measure to anchor the idiosyncratic item measurement. There was no requirement for participants to contribute data.
Participants were surveyed at the start and end of the programme about the acceptability of a group format and the mix of participants. They were also asked at the end to rate the degree to which it had been enjoyable and helpful in daily life (from ‘not at all’ to ‘completely’).
Participants were asked to identify problems or difficulties they hoped might be helped by participation (‘How would you know you are spending more time with your “being” head on?’) and to hold these confidential to themselves, but identify each as an item number on a record sheet. The sheet also asked participants to record the frequency (‘How often they trouble you’) and rate the degree of distress [‘How distressing or severe is this (0–100; 100 is most severe/distressing I can imagine’)] incurred in the preceding week. These could be left in a sealed box at the entrance to the group room. Multiple baseline measures were used. Progress was tracked by labelling data submission packs. There was no record kept of who received which pack, but participants were invited to identify themselves as a member of service user, carer or staff groups. They were invited at the end to submit information concerning the nature of each item.
Results
Attendance
There were a total of 19 participants, comprising 11 service users, three carers and five members of staff. Service users attended a median of five sessions (range 1–7); carers attended a median of seven sessions (range 3–8) and staff members attended a median of five sessions (range 4–7). In total, there were 90 attendances out of a possible total of 152. Fourteen participants attended the first session; 12 attended the last. Session 4 had the lowest attendance (eight participants). Two participants submitted baseline measures that showed no change in the 4 weeks preceding the start of the study.
Acceptability of group format and participant mix
This was assessed on a 7-point scale, anchored at ‘not at all acceptable’ to ‘completely acceptable’. Ten participants contributed to this component of the evaluation, although only two provided both pre- and post-group information, both of whom were in the carer participant group.
At the outset, six scored the acceptability of a group format as 7; one participant scored it as 6 and one as 5. Seven scored the acceptability of the participant mix as 7; one as 5. It may be noteworthy that, although the group and its mix were enthusiastically endorsed among respondents the two scores of 5 were given by a staff member.
Four participants contributed to the end-group evaluation component. Two did not identify their grouping and two were carers. All found the group format and group mix completely acceptable (rating 7). Ratings for the usefulness of the group were 4, 5, 6, 6. Ratings for the enjoyability of the group were 4, 6, 6, 7. Usefulness and enjoyability scores of 4 were rated by the same participant, but this individual did not identify their grouping. Interestingly, in spite of this apparent ambivalence, this individual's (CORE) scores dropped from 45 to 25 over the course of the group and mean idiosyncratic measure intensity (see next section) fell from 514 to 332.
Idiosyncratic measures
A total of ten participants contributed self-monitoring data. These were two service users, three carers and three staff. Two did not identify themselves. This amounted to 36 items and their intensity units (IU) were calculated by reported frequency × severity. These showed a very wide range (20–1500) and were significantly deviant from a normal distribution.
CORE validation of idiosyncratic measures
The group improved on the mean CORE measure total score over the course of the study (pre = 31.0, post = 21.4, s.d. = 16.6, effect size = 0.58; Cohen, Reference Cohen1969). Change on idiosyncratic measures was standardized to compare with the CORE measure: change per unit base measure (A – B/A) per unit time (weeks) was calculated for each item. CORE change per unit time (A – B/t) was also calculated. Neither was significantly deviant from a normal distribution and they were highly significantly correlated (Pearson's correlation = 0.744, p < 0.001). This demonstrates that the rate of change on the idiosyncratic measure matches the rate of change in the CORE and provides a validation of the items being monitored.
Idiosyncratic measure change over treatment
There were 36 data points recorded at baseline/start of therapy and 31 available ‘later’ for analysis. The median idiosyncratic IU measure score at the start was 300 and 90 ‘later’ showing a pre-post improvement in intensity [Mann–Whitney (MW) U test z score = −3.82, p < 0.01].
The service user subgroup included eight start data points (median IU = 272.5) and eight later (median IU = 205, MW z = −1.76, p = 0.079). The carer subgroup included 10 start data points (median IU = 290) and 10 later (median IU = 75, MW z = −2.54, p < 0.01). The staff subgroup included nine start data points (median IU = 500) and seven later (median IU = 490; MW z = −0.6, p = 0.522). Those who did not identify themselves with a group provided nine start data points (median IU = 420) and six later (median IU = 40; MW z = −2.26, p < 0.05).
Of the 31 items available for direct pre-post comparison, 27 improved, two deteriorated and two remained the same. Of the two items deteriorating, one each occurred in two participants. Both were members of the staff group and they had other items that improved over the course of the group; one also had CORE scores recorded and these substantially reduced. The two items that remained unchanged were representative of one participant who was a member of the staff group. The latest record of these were in week 3 and no CORE questionnaires were submitted, therefore no information is available concerning later participation in the group, change of these items or other progress on the CORE measure.
Qualitative evaluation and analysis
Nine participants provided data concerning the personal goals they had hoped for. One noticed improvement in tiredness and concentration. The same participant noticed a reduction in distress associated with ‘racing thoughts’ and a ‘lack of peacefulness’, while recognizing a greater degree of awareness of this experience. Another noticed reduction in ‘stress’ over the course of the group. One became less concerned about a problem, enabling them to share it with family and improve family relationships. This participant also became more engaged in daily life, engaged in more activities and became more able to enjoy their experiences.
Another participant noticed improved ability to have restful sleep and reduction in over-eating. One reported dietary improvement and reduction in alcohol consumption, with associated and desired weight reduction. Another reported improvement in a range of persecutory experiences. One participant reported greater interest in life associated with greater relaxation; another reported the experience as ‘refreshing’.
Very detailed feedback from one participant reported having ‘benefited enormously . . . hav[ing] my being head on more often. . . . It has helped me to accept [my negative feelings] as part of being human [and therefore] not get sucked into . . . spiral[ling] out of control.’ There was a report about feeling less guilty and stressed when busy life demands are difficult to fulfil. The ability to stay asleep, disengage from angry rumination and overcome procrastination improved. Another participant reported improved ability to care for and love themselves.
Facilitator experience. The setting, a seminar room in a (predominantly) physical rehabilitation hospital, by a busy road was not a peaceful one. There were many intrusive noises for the participants to work with during meditation including, when all else was quiet, the sound of a wall clock ticking by the seconds. One early gain reported by several participants was being able to settle irritation with these factors. One participant struggled with abusive voices but reported a developed ability to direct attention away from these, gaining some mastery for the first time.
Staff participants sometimes arrived late resulting in some personal responses for the facilitator to work with! Their contributions to the enquiry included experiencing the meditation as ‘relaxing’; working with the content of the previous clinical encounter and finding shopping lists uppermost in their minds. When responding to patient struggles staff tended to offer advice to ameliorate experiences.
Additional participant experience. There was no invitation for freehand feedback on the experience; however, one carer participant gave extensive feedback on the experience on behalf of themselves and the service user to whom they related:
[I] enjoyed being part of the group . . . welcoming and inclusive . . . would have liked a recap of names at the beginning of each session . . . I have benefited enormously from participating . . . it has helped me have my ‘being’ head on more often . . . ___ was reluctant to attend . . . I passed on [group facilitator's] message that he was welcome at any of the sessions and ___ came to 3 more. . . . ___ said the meditation had been refreshing and liked the way there was ‘no pressure on you’.
Discussion
This feasibility study of a mixed-participant Mindfulness meditation group is encouraging. It showed gain for the group as a whole on their target problems, with most benefit accruing to service users and carers (statistical significance showing even in the small carer subgroup) and only marginal gain to staff as a whole. However, one staff member showed substantial gain on the clinical CORE measure over the course of the group.
The facilitator was a very experienced CBT therapist, although a relatively inexperienced Mindfulness practitioner. The delivery of this group Mindfulness intervention would likely have been substantially improved from further Mindfulness training, more extensive personal practice and a Mindfulness-focused supervision. While this report may represent the picture of a realistic interim as such arrangements, experience and expertise develop in the Health Service in the next few years, Teasdale et al. (Reference Teasdale, Segal and Williams2003) caution against its injudicious use. Although they recognize that Mindfulness practice may ‘affect processes common to different pathologies’ they highlight, among other factors: (i) practitioners’ use of individual psychological formulations to target expected effects in achieving good outcomes and (ii) the fact of deterioration in the condition of some patients when Mindfulness-based interventions have been applied. In this study, although patients were not individually interviewed and formulations derived, the practitioner's familiarity with CBT models of psychosis maintenance (e.g. Morrison, Reference Morrison2001) may have reduced risks.
Acceptability, enjoyability and usefulness data are based on a small subgroup of participants. Those contributing end data found both mix and group format acceptable, helpful and enjoyable overall. The one who did not highly rate the degree of helpfulness and enjoyability still showed substantial symptomatic gain over the life of the group. There could be a natural bias in this component of evaluation as those who persisted are likely be those perceiving it more positively. However the lack of fall-off in attendance suggests that factors other than this account for the low rate of data contribution.
Ten of the 19 participants contributed data on idiosyncratic measures and nine of these provided ‘later’ data from self-monitoring. It is unclear how representative these findings are. It is important to note that each participant group is represented in the findings and this allowed some cross-group comparison. It is interesting that no group was more concordant with data collection than others. Those who did not contribute data predominantly did not contribute baseline data so it is unlikely that they were disenchanted with the group or found it unhelpful. Rather it seems more likely that they came with an intention not to contribute data. This was explicitly facilitated by the group leader and in the information sheet.
Benefits accrued over the course of an 8-week group are likely to be consequential upon a range of factors. Service user participants will have been receiving medication and psychologically informed care coordinator support; circumstances change in individuals’ lives. Some service users and carers may have been receiving family interventions, although none were explicitly receiving concurrent CBT. The multiple baseline measurement is intended to limit the confounding influence of other factors. Where this was achieved (two participants), there was no change between baseline measures and start, providing some validation of the helpfulness of the programme. Furthermore, the ongoing good turn-out suggests that the findings from the subgroup who did contribute data may be representative.
It is encouraging that the idiosyncratic measures of distress improved over the life of the group. However, it is interesting to note precisely what experiences these represent and their breadth. These may vary according to individual constructs through which development is perceived. On way of construing or grouping the reports could be: ‘engagement with daily life’, ‘relationship with self or experiences’ and ‘symptoms’. Five out of nine participants reported in only one of these. If development actually improves in all of these areas (four reported gain in two or more), it may be that goals identified in different domains represent different aspects of the participants’ experience (in the same way that thoughts and feelings are different aspects of the same experience) and that it is the labelling (rather than the experience for which help is being sought) that will determine the domain for each goal. A tentative hypothesis is that the participants’ model of distress (whether illness, activity or self-actualization focused) determines how it is described.
Similarly to Chadwick et al. (Reference Chadwick, Hughes, Russell, Russell and Dagnan2009), this paper reports gain on a symptomatic measure (CORE) but is similarly underpowered to show statistical significance. However, that was not the primary aim of this study and the results for our service users, in keeping with Abba et al. (Reference Abba, Chadwick and Stevenson2008), primarily show a reduction in distress in keeping with a reported altered relationship with symptoms. Sorrell et al. (Reference Sorrell, Hayward and Meddings2010) show that greater distance from voices is associated with greater distress and the components of Chadwick's (Reference Chadwick2006) model of therapy (turning towards the difficult and acceptance) that facilitate closer engagement with voices may be important in this changing of relationship with symptoms associated with distress reduction.
The issue of participant mix is interesting. The group set out to explicitly normalize psychological distress. However, it appears that not all staff were entirely comfortable with this and no participant that identified themselves as staff contributed to the end-group evaluation of this aspect. Whether the initial (score 5/7 on acceptability of a group format and on its mixed participant nature) relative reservation suggests this staff member thought that some would be inhibited in the group or that the group would be unable to adequately help the more vulnerable members or some other is unclear. The findings may go some way to addressing the latter concern.
It is clear that the ability to privilege time boundaries, whether for delivering therapy or attending supervision or meditation groups, is particularly difficult in care coordinator roles and future work of this nature should engage staff members in careful consideration of how to manage the time boundaries of such an experience. In an early intervention service, patients are mental health service culture-naive, but the same is not true of staff. This may contribute to their relative unpreparedness. While many espouse a normalizing understanding of psychotic experiences, most staff will have had years of exposure to a stigmatized view of mental ill-health. This may manifest itself in an emotional-intellectual mismatch of preparedness for this experience. In this way, one possibility is that staff who reported it as ‘relaxing’ in the enquiry may be less able to be in touch with the work of meditation while service users and carers have less baggage about the meaning of psychological struggle and be both better able to experience it and less inhibited about sharing it. Another possibility is that the modification of Mindfulness practice intended for a psychotic group substantially lightened the intensity of the experience for others. It may also be that the focus of struggle for staff was on the experience of working with service users. This would understandably be difficult to share in a mixed-participant group.
In CBT supervision, staff report credibility to their patients as a barrier to trying new things: they find it difficult to show vulnerability to people with whom they have a professional role relationship. The entrenched nature of this role may have some explanatory role in considering the less engaged experience of some staff and may also explain some staff's relative reservation about the composition of the group in formal feedback. Taken together, this suggests that staff need more preparation for this kind of approach or else to have access to their own support. Given professional role requirements in our culture, psychological distress may not yet be amenable to normalization in this way.
Conclusions and recommendations
Despite limitations of data collection, these findings fit with a growing body of evidence supporting the use of Mindfulness meditation in clinical practice and within the ‘psychosis’ population and their carers. Findings support feasibility of the approach, although staff may need more careful selection and preparation for a mixed-participant group. Future programmes could helpfully investigate expectations and experience of staff in the groups, examining either the effect of preparation or eliciting barriers in detailed interview.
This report supports continued development of Mindfulness-based group interventions, particularly for service users and their carers, but with provisos that careful evaluation continues and recommendation of one-to-one preparation with staff participants is included. The latter would substantially increase the resource implications.
The qualitative information would be greatly enhanced by a pre- or post-course interview or both. In particular, it would be possible to explore the tentative view that identified domain goals are determined by a participant model of distress.
Mindfulness group facilitators would be likely to benefit from more extensive personal practice and mindfulness-focused supervision. However, this pilot has shown feasibility and effectiveness, as well as a good degree of acceptability of this intervention with only written material preparation for participants and a relatively low level of Mindfulness expertise (albeit in the context of extensive CBT experience) in the facilitator.
Acknowledgements
The author is grateful to Ms. Cate Wood, cognitive therapist and Mindfulness practitioner, for advice in the preparation of this manuscript.
Declaration of Interest
None.
(1) Update knowledge of effectiveness of Mindfulness interventions for psychosis.
(2) Improved understanding of the nature of benefits for people practising Mindfulness.
(3) Consider the role of Mindfulness meditation in supporting carers.
(4) Reflection on challenges of implementing a normalizing view of psychological distress in a mixed group of service users, their carers and staff.
Comments
No Comments have been published for this article.