Prophylactic medication for bipolar disorder reduces frequency of acute episodes and intervening subsyndromal symptoms (1). Psychological treatment provides additional efficacy (2). One effective approach—relapse prevention—reduces frequency of acute episodes, reduces hospitalization, and improves function (3). Such approaches are cost-effective within collaborative care for at least two to three years (4,5) and are more effective in mania than depression (4,5,6).
Implementation of psychological interventions for bipolar disorder in routine clinical practice is problematic. None of the interventions investigated in randomized controlled trials have used existing service structures, and most require staff with extensive experience of bipolar disorder and psychotherapy training, severely impeding availability in routine clinical care (3). Relapse prevention (6) and psychoeducation interventions (4,5,7,8) require less extensive psychotherapy training but still require clinicians to have experience in managing bipolar disorder.
Multidisciplinary community-based mental health services are common internationally, but there are few evaluations of the effectiveness of psychological interventions in routine care and none confined to bipolar disorder.
This brief report describes a cluster randomized controlled trial examining the feasibility and effectiveness of implementing an enhanced relapse prevention intervention in the United Kingdom's National Health Service (NHS). Detailed rationale and design for the study are given in an article by Lobban and colleagues (9). The intervention is enhanced compared with previous relapse prevention interventions (6) because it has an increased focus on coping strategies for depression; it is short, structured, and designed to be offered by community staff; and it involves a relative or close friend in the intervention where appropriate.
After the study received ethical approval from the U.K. National Research Ethics Service, all community mental health teams in the four participating NHS Trusts in North West England were invited to take part. The study began in 2005 and was completed in 2007. Participants were then followed for one year, and follow-up data will be reported elsewhere. Before randomization, a minimum of four staff from each community mental health team, each working with at least three people with a diagnosis of bipolar disorder, gave written informed consent. Stratified by trust, community mental health teams were then randomly allocated to receive training in enhanced relapse prevention or to continue to offer treatment as usual.
The key aims of training were to enable staff to work with service users to increase knowledge and understanding of bipolar disorder, identify triggers for relapse, increase ability to recognize early warning signs of relapse, develop coping strategies in response to early warning signs in order to increase control over mood changes, maximize the use of social support systems by involving relatives or friends in the management of mood where appropriate, and improve the response of services to early warning signs by developing shared care plans.
The intervention was detailed in a manual provided to staff that was also designed to be used by service users and relatives during the intervention. The manual has six sections, each corresponding to a two-hour session of training for the staff, and one intervention session for the service user and relatives.
Training was delivered by a senior mental health nurse with previous experience in training nursing staff and working with people with bipolar disorder but without formal accreditation or training in psychological therapies. Training was provided in the workplace of the staff to maximize attendance. Training followed the structure of the manual and used slide presentations, video role-plays using professional actors, open discussion, and practice role-plays with feedback on performance. Specialist supervision to support staff in using the intervention with existing service users with bipolar disorder was offered to staff in each team in six weekly group sessions at their workplace at a mutually convenient time.
Staff had to attend at least four of the six sessions of training in order to be considered fully trained in enhanced relapse prevention, balancing the need to receive comprehensive training and the reality of training people in real-world clinical settings. Self-rated perceived skill in working with people with bipolar disorder to prevent relapse was assessed with a Likert scale that was given at baseline, posttraining, and postsupervision. Possible scores ranged from 1, completely unskilled, to 10, completely skilled. Staff confidence was assessed with a questionnaire designed for the study that assessed ten key elements of implementing enhanced relapse prevention. Self-rated confidence total scores were calculated by applying scores of 1, not able, to 4, very able, and summing across the ten elements. Finally, the trainer used the same scoring method to rate each person on how competent he felt he or she would be in implementing the ten key elements of enhanced relapse prevention. Ratings were made posttraining and postsupervision. Feedback on training and supervision was collected with a questionnaire designed for the study. Both were rated on a 4-point scale on quality, the extent to which they met the staff's needs, satisfaction, and usefulness in supporting people with bipolar disorder. All ratings by staff and trainer were collected independently and blindly.
Statistical analyses were carried out with SAS, version 9.1. The SAS procedure surveymeans was used to produce 95% confidence intervals for means adjusted for the design, which was stratified at the trust level with clusters at the level of the community mental health team.
A total of 112 staff members from 23 teams agreed to take part in the study. Eleven teams (56 staff members; median team size, five; range of four to six) were randomly allocated to continue to offer treatment as usual, and 12 teams (56 staff members; median team size, four; range of two to eight) were randomly assigned to receiving training in enhanced relapse prevention. This brief report focuses on the 12 teams who received enhanced relapse prevention training.
Fifty staff members provided information on their professional background. The majority (N=40, 80%) were community psychiatric nurses, and the rest were social workers (N=8, 16%) or occupational therapists (N=2, 4%). Only four of 50 (8%) reported any previous training in bipolar disorder.
Thirty-eight staff members (68%) attended at least four of the six sessions of training, which indicated that they were fully trained in enhanced relapse prevention. The median proportion of fully trained staff in a community mental health team was 68.8% (range of 0 to 100); only one team had no fully trained staff.
Staff who did not attend any training sessions reported less baseline self-rated perceived skill in working with people with bipolar disorder to prevent relapse (N=16, mean=4.55±1.08, 95% confidence interval [CI]=3.59–5.50) than those who attended one or more sessions of training (N=40, mean=5.61±1.71, CI=4.96–6.25). This difference was not statistically significant.
Self-rated skill of working with people with bipolar disorder increased from baseline (N=49, mean=5.37±1.36, CI=4.99–5.75) to posttraining (N=36, mean=6.64±1.96, CI=6.00–7.27) (mean change in skill=1.09±.76, CI=.84–1.33) and from posttraining to postsupervision (N=20, mean=7.70±1.76, CI=6.93–8.47) (N=19, mean change in skill=.79±1.87, CI=-.03 to 1.61).
Training was very well received by the 36 staff who provided feedback, all of whom were fully trained. All stated that the training would be fairly or very useful, and 34 (94%) rated the overall quality of the training as good or excellent.
After training staff felt confident about implementing enhanced relapse prevention. The majority of the sample felt able or very able to implement each of the key elements. The trainer rated more than 70% of the staff as able or very able on each of the key elements after training (Table 1).
Twenty-four (63%) of the 38 trained staff attended one or more sessions of supervision: of these, the median number of sessions attended was five (range of one to six), suggesting that once staff attended supervision, they were well engaged.
Twenty of the 24 staff (83%) who attended supervision sessions gave feedback. This was generally positive: of those who gave feedback, 17 (85%) rated the overall quality of the supervision as good or excellent, and 17 (85%) stated that the supervision was fairly or very useful.
After completing supervision, all staff felt able or very able to carry out all the key elements of enhanced relapse prevention, except for two staff members (10%) who felt only slightly able to devise a timeline that mapped previous episodes and identified key events that occurred in relation to mood changes and three (15%) who felt only slightly able to identify interepisode symptoms (Table 1). The mean change in overall self-rated confidence (postsupervision-posttraining) was 2.63±6.47 (N=19, CI=-.28 to 5.54). Confidence increased for 14 staff after supervision, but it was lower for five.
Staff competence in implementing key elements of enhanced relapse prevention was rated by the trainer. The trainer rated 17 of the 24 staff who attended supervision sessions (71%) as able or very able to carry out all key elements of enhanced relapse prevention, except identification of interepisode symptoms, for which only 14 (58%) were rated as able or very able and four (17%) were rated as not able (Table 1). The mean change in overall trainer ratings of competency (postsupervision-posttraining) was 5.83±8.60 (N=24, CI=2.39–9.28). Confidence increased for 21 staff postsupervision, but it decreased for three.
There was demand for skills-based training in relapse prevention for bipolar disorder among multidisciplinary teams of mental health professionals working in the community, with over 50% of eligible teams willing to take part in the study presented here. Very few of those taking part had received previous training specifically related to bipolar disorder. Staff in these teams did not perceive themselves as skilled at preventing relapse before training, and feedback showed that they valued even basic information about bipolar disorder. Staff who did not attend any training sessions perceived themselves at baseline as less skilled in working with people to prevent relapse in bipolar disorder compared with those who did attend, although this difference was not statistically significant. This suggests that nonattendance may reflect lack of confidence in being able to carry out the training, rather than lack of perceived need.
Attendance at training was acceptable, with over two-thirds of staff completing training. However, only 43% of eligible staff attended both training and supervision even though these occurred in the workplace at mutually convenient times. With more resources and with training and supervision given a higher priority by managers, higher attendance rates may have been achieved. Attendance at both training and supervision sessions may be important in terms of the effectiveness of the intervention, because supervision resulted in a high degree of confidence in all staff who provided feedback, although there were no significant improvements in trainer-rated competence on specific skills in delivering enhanced relapse prevention. Overall, those who attended both training and supervision sessions showed a marked improvement in their overall self-rated skill when working with patients with bipolar disorder.
Feedback and evaluation of both training and supervision from those who attended were very positive, although the small number of staff who attended only a few sessions of training or supervision may have been more critical had they offered feedback. Some staff would have valued longer training and closer supervision, but no one asked for less training or supervision. The most often cited reason for nonattendance at training or supervision sessions was lack of time. Lack of time is often cited by health professionals as a reason for not taking part in training or research, but many health professionals will find the time to carry out other professional activities that they value (10). Staff who did not attend may not have felt that this approach was the most effective intervention to use, or they may have had anxiety about their lack of skill, which led to avoidance of training environments. Further investigation is needed to explore these issues.
The main strength of this study lies in its external validity. The training and supervision in enhanced relapse prevention were offered in routine clinical services, with the necessary degree of flexibility this requires. The cluster randomized controlled trial design ensured that the intervention could be implemented and evaluated at the level of the community mental health team without contamination between the enhanced relapse prevention and nontraining arms of the study.
The study had limitations. Confidence and competence were not rated by care coordinators allocated to offer treatment as usual with no additional training (control group). Changes in confidence and competence cannot therefore definitively be attributed to the specific training intervention. There was no objective assessment of competence. Competence ratings were made by the trainer after training and supervising the staff members; these ratings may be biased to show the trainer in a good light and influenced by staff confidence. Feedback and evaluation were available only after training and supervision, so we cannot prove that training improved the confidence and competencies of staff in the specific tasks of enhanced relapse prevention. However, before training, self-report ratings of overall skill in relapse prevention in bipolar disorder were modest, so training and supervision may have increased both confidence and competencies in relapse prevention in bipolar disorder. The study was not powered to detect changes in staff ratings and may not have detected the true impact of the intervention, especially given the reduced sample size for postsupervision ratings. Alternatively, there may have been ceiling effects on competencies, so that supervision did not increase the competencies of staff but increased their confidence in their ability to implement enhanced relapse prevention.
Mental health professionals working with people with bipolar disorder in the community valued skills-based training and supervision in an enhanced form of relapse prevention. Confidence and competence both improved among those who attended, but barriers to attendance need to be addressed.
This study was supported by grant MRC G0301042 from the Medical Research Council and funding from Mersey Care NHS Trust.
Professor Morriss has received fees from Eli Lilly and Company for conference presentations. The other authors report no competing interests.