The Emotional Resources Group: A Mixed Methods Practice-Based Study of a Transdiagnostic Emotion Regulation Group Intervention

Abstract

Objectives: This study aimed to understand if the emotional resource group (ERG), a transdiagnostic group intervention, significantly improved emotion regulation (ER) in the short and long term. Methods: Participants completed pre- and post-measures of emotional regulation, psychological distress, wellbeing, self-efficacy and confidence in managing mental health problems. Measures were repeated at three and six months post intervention. 14 participants completed a qualitative interview. Results: ER ability and secondary measures improved following the intervention, but the improvement was not maintained over the three and six month follow-up period. From the themes identified through thematic analysis, the ERG was part of a wider process of learning and change for most participants. Participants described changes in awareness, and implementation of new skills to manage their emotional difficulties. Conclusion: Overall the ERG was a positive experience for most participants, providing improvements in ER and other aspects of mental health. However, it is unclear if these improvements are sustained in the longer term. There may be many reasons for this, including fidelity to the ERG model and the short-term nature of the group. The ERG should be considered an option for participants with ER difficulties and further intervention should be offered when the ERG ends.

1. Introduction

Emotion regulation (ER) ability is an important target of interventions across mental health presentations, and most therapeutic interventions include emotional regulation strategies [1,2,3,4,5,6].

Though it is agreed ER is an important treatment target, a key challenge with developing interventions is there is little consensus on what ER is and what can be done to improve it [7]. An understanding of what ER is and what strategies improve it is essential to ensure the development of effective interventions that specifically target ER, rather than addressing other difficulties.

Two influential theories of ER are the appraisal-based model [8] and the Theory of Constructed Emotion (TCE) [9]. The appraisal-based model of emotions states that, irrespective of external stimuli, the emotion we experience largely depends upon how we appraise the situation [10,11]. Experimental and neuroimaging studies support the appraisal-based model and show that differing interpretations of events can lead to different intensities or types of subjective emotions [8,12,13]. An operationalisation of how appraisal theory applies to ER is the extended process model of emotional regulation [7,14]. The process model states that emotion generation involves a typical sequence of processes culminating in the experience and response to any discrete emotion. As such, the model suggests there are different strategies and points at which one might intervene to regulate an emotion [14], and these different choices will lead to different outcomes. The extended process model also introduces the idea of valuation, that we weigh our emotions on the dimension of “good for me” and “bad for me” [7]. Depending on this valuation at various stages, and to what extent we have awareness of these stages in any given context, we may make different choices around which strategies for ER we attempt.

The Theory of Constructed Emotion (TCE) further develops these ideas, and states that our emotions are an active construction based on our previous experiences and conceptual knowledge of emotions, and are predictive rather than reactive to stimuli [9]. In this theory, affect (which is distinct from emotion) is an allostatic, interoceptive prediction implicitly generated by past experiences. It is only when affect if processed through a conceptual lense, that is becomes an emotion. This implies that human emotions are not an automatic response to stimuli; rather, they are an active creation, based on past experience and conceptual knowledge, serving the purpose of creating optimum conditions for our continued survival.

The appraisal-based model and TCE share similarities, with emotions seen as active constructions shaped by the individual and their circumstances [15]. Based on these theories, the focus for ER interventions is broader, encompassing emotion generation processes as well as strategies used once the emotion is in conscious awareness. Both models also emphasise the importance of meaning-making. In the Gross model, this is about cognitive appraisals and evaluations; in the TCE this is about emotional granularity and the usefulness of any particular emotion construct in any particular situation [7,9]. Both models also propose that emotions emerge over time, suggesting that interventions aiming to improve ER should have different strategies for different time courses. Based on these models, a range of skills should be included in ER interventions, including skills focusing on understanding and identifying emotions (e.g., mindfulness, beliefs about ability to change emotions, valuing emotion), and skills that increase awareness of emotions and ability to select strategies (e.g., sense of self-efficacy, distraction, reappraisal).

1.1. Emotion Regulation Group Interventions

Designing ER interventions for groups may have many benefits, some practical (e.g., allowing for efficient use of clinical time) but also in terms of group process. Group interventions can have benefits to services as they can be less time and resource intensive than individual therapy; and although individual approaches tend to be better researched and evidenced, some studies have shown comparable benefits [16,17,18].

Various ER group-based interventions have been developed. Two widely used interventions are Emotion Regulation Group Therapy [19] and Systems Training for Emotional Predictability and Problem Solving [20]. These groups can improve ER skills and other domains [19,21]. Additionally, ER modules from DBT skills training can also improve ER without the addition of other skills training modules [22,23]. However, these interventions come from adapting therapies for specific diagnoses and could potentially not be applicable transdiagnostically as they have not been evaluated for a range of presentations that might have difficulties with ER [21,24]. There are some transdiagnostic ER group interventions, for example, a group adaptation of the Unified Protocol [25] which draws on a range of approaches including CBT, DBT, mindfulness and ACT, again showing a range of benefits [25,26,27,28]. These interventions include elements that are not focused solely on ER, such as CBT being used as an adjunct, so it possible that group outcomes may be related to these elements in addition to more specific ER-focused strategies. Additionally, they utilise specific skills from wider interventions (e.g., distancing techniques from ACT). While the full package of the intervention (e.g., individual ACT therapy) has been shown to be effective, it is unknown if individual components of these interventions are effective independent of the wider intervention [29].

1.2. The Need for Trauma-Informed Interventions

Trauma is a significant risk factor for developing mental health problems, and the majority of people presenting to mental health services will have experienced trauma [30,31,32,33]. Given the high prevalence of trauma within mental health services, it is important that interventions are trauma-informed, but none of the interventions mentioned are designed to be so. This is particularly crucial as ER has been found to mediate the link between childhood trauma and later mental health problems [34]. This would suggest that trauma itself might also be considered as an important transdiagnostic concept, underlying ER difficulties across a range of disorders. It is therefore important to ensure that interventions target both the ER difficulties and the impact of trauma.

1.3. The Emotional Resources Group

The ERG [35] is a 6 session transdiagnostic intervention explicitly focused on ER skills. The group is based on the TCE, and therefore proposes that participants’ subjective feelings in a situation are shaped by prior experience and learning, and may be more or less adaptive to their current situation [29,35]. Two frameworks of ER guide facilitators in assisting participants to learn which ER skills should be implemented in which contexts. These are the Gross process model [8], which highlights the importance of timing for ER, and Koole’s [36] functional model, which considers the function of ER strategies on different targets. Skills have been included in the ERG only if they have been shown to be effective independently of a wider model of intervention, or if they are skills widely used in trauma-focused work, to ensure a trauma-informed approach [35].

Each session focuses on reviewing participants’ emotional reactions, recorded during the week using a “trigger log” to help reflect on the historical origin of sensitized emotional reactions, and goodness-of-fit of various emotion concepts to current situations. It also teaches new skills and provides psychoeducation on emotions. The group also has a strong focus on skill implementation. Further information about the group content can be found in Bacon et al. [29].

An initial evaluation showed positive effects on emotion regulation, as well as well-being and self-efficacy [29]. However, there was no control group or randomisation, so it is unclear if changes are related specifically to the intervention. Additionally, there was no follow-up. Given that one aim was to create lasting change, there is a need to measure beyond pre- and post-scores. The ERG is also shorter than other similar interventions, and a longer-term follow-up help understand whether six weeks is enough to change ER in the long term.

Research to date has focused on quantitative measures. While using validated outcome measures to establish efficacy of psychological interventions is part of routine clinical practice [37], it would also be beneficial to understand participant’s experiences of the ERG. More generally, it has been proposed that using only routine outcome measures can fail to detect potentially harmful interventions or likely treatment responders [38]. This may be especially important given the use of the TCE to underpin the group: the evidence base is still small, and may be counter-intuitive to participants as it contrasts with people’s naïve experience of emotion [9].

1.4. The Current Research

The present study completed a follow-up of the ERG. Qualitative interviews were also used to explore participants’ experiences and to better understand the impact of the group on ER skills.

The following research questions were addressed:

Does the ERG significantly increase emotional regulation ability in the short and long-term?

Does the ERG have a secondary impact on subjective wellbeing and confidence to manage mental health?

What are participants’ experiences of the ERG group?

2. Method

2.1. Design

The study was carried out in NHS settings as a piece of practice-based research. As such, there were limitations on what data was available, the authors did not have access to patients apart from those receiving the group intervention so were not able to randomise allocation to treatment condition. A mixed method design was used to evaluate the implementation of the ERG. The study was non-randomised and did not have a control group. Participants were referred to the group based on their need for the specific intervention and if active treatment is available it is not ethical to withhold treatment for the purposes of randomisation or having a control group.

Referrals for the group were sent in by various practitioners once a date for the group was set, so there was no waitlist that could be used for a control. There was no other similar group being run by the service so no active control could be used.

2.2. Measures

Measures were collected pre/post intervention and after three and six months. The measures used were based on a previous evaluation [29], with the addition of the CORE-10 to understand the impact of the ERG on psychological distress.

Difficulties in Emotion Regulation Scale (DERS)—this is the primary measure. The DERS is a widely used measure with good psychometric properties, with Cronbach’s alpha of 0.93 and 0.92 in recent studies [24,29].

Mental Health Confidence Scale (MHCS)—this measures the sense of self-efficacy that people have and is used to assess if participants’ confidence in living with mental health challenges increases throughout the intervention. It also has good psychometrics, with a Cronbach’s alpha of 0.88 in a similar study [29,39].

Warwick Edinburgh Mental Wellbeing scale (WEMWBS)—this measure gives a single measure of wellbeing and is also well validated, with previous studies finding a Cronbach’s alpha of 0.84 [29,40]. This is used to understand if there is a secondary impact of the ER group on general wellbeing.

Work and Social Adjustment Scale (WSAS) is a measure of functional impairment and can be used across different groups and disorders [41]. The WSAS has good validity and is sensitive to treatment related change [42]. It also measures social functioning, a dimension not captured in other outcome measures [43].

The CORE-10 is a brief assessment measure that is valid across a range of common mental health presentations and widely used within routine clinical practice [44]. It has clear clinical cut-offs and a reliable change index [45].

2.2.1. Treatment Fidelity

All facilitators received routine supervision, but there was no ERG specific supervision. All groups used the ERG group materials described in Bacon et al. [29]. Formal fidelity measures were not used. Information on facilitator training and competence was not available to researchers.

2.2.2. Participants

Participants were included if they had attended the ERG in the NHS board between March 2019 to August 2020, were over 16 years old and had capacity to consent. Participants had a range of diagnoses, including depression, anxiety, OCD, PTSD and personality disorders. Participants were able to access general mental health support during the ERG, though individual psychotherapy appointments were not scheduled while the groups were running.

In the data collection period, seven groups were completed. Figure 1 shows the sample attrition.

2.3. Procedure

Participants were made aware of the research at the start of the ERG. Participants had time to consider the information before giving consent. The pre/post data was collected by group facilitators. The first author contacted consenting participants to arrange follow-up appointments. A qualitative interview was offered to all participants. The transcript and all other data were securely stored in line with legal requirements.

2.4. Qualitative Interview

Data was collected through semi-structured interviews. The interviews were completed either face-to-face or over the telephone.

2.5. Ethics

The study obtained ethical approval from University of Edinburgh Research Ethics Committee. Caldicott approval was given by the health board and further NHS ethical review was not required.

2.6. Analysis

2.6.1. Quantitative Measures

The primary analysis was paired t-tests analysing the pre/post data to examine whether the ERG improves ER and other measures within the intervention period. The secondary analysis used a repeated measures ANOVA to ascertain if any improvements were maintained over time. In measures where this was possible, reliable change and clinically significant change was calculated.

2.6.2. Qualitative Interviews

The interview transcripts were assigned a pseudonym, transcribed and personally identifiable information was removed. Thematic analysis (TA) was used to identify themes. TA is a way of identifying, analysing and reporting patterns in data [46]. This research followed Braun and Clarke’s orientation to TA, which locates TA within a qualitative paradigm, often called “Big Q” [47]. Big Q is non-positivist, assuming there is no universally correct, “objective” version of reality, and emphasises the importance of context, which seemed important as a proportion of the data was collected during the COVID-19 pandemic.

3. Results

3.1. Quantitative Results

Prior to analysis the data was inspected for missing data. There was no missing data at baseline, and missing data at other time points. This was due to participants being lost to follow-up. Casewise deletion was used for the analysis examining all four time points.

Due to the level of attrition between the end of the group and the three month follow-up, a separate analysis—a paired-samples t-test—was completed for the larger data set that was available for the two time points of baseline and intervention end. In all five measures, there was a statistically significant difference between the two time points, indicating improvement between the start and end of the group.

For the DERS, the primary outcome measure, a reliable change and clinically significant change was calculated previously (Bacon et al., 2018). Based on this, 42% (n = 13) of participants showed reliable change between pre and post ERG. A value for clinically significant change was also calculated, by calculating the halfway point between the pre-group mean score (a clinical group) and a “normal” population mean (using data from Gratz & Roemer [24]), giving a value of 104.14. Any participant whose scores were below, or equal to this value was described as showing clinically significant change, as was the case for 32% (n = 10) of participants.

Although there was a statistically significant difference between pre and post scores on the CORE-10, only 55% of participants showed reliable improvement and 13% showed clinically significant change according to criteria by Barkham et al. [45].

Clinically significant change on the WSAS is thought to be a decrease in scores by at least 8: this was seen in 19% (n = 6) of participants [42]. Only two participants improved to an extent that their scores were considered subclinical. This suggests quite a modest improvement in functioning across participants.

The WEMWBS has no clinical cut-offs, but scores can be split into low (14–42), average (43–60), and high mental wellbeing (60–70). By this measure, 61% (n = 19) of participants started and finished the group with low wellbeing, 16% (n = 5) improved from low to average wellbeing during the group, 16% (n = 5) started and finished the group with average wellbeing, and 6.5% (n = 2) declined from average to low wellbeing during their time in the ERG. Considering this, although there is a statistically significant improvement, it is likely the ERG has only a small effect on improving mental wellbeing.

The MHCS has possible scores between 16 and 96, and has no clinical cut-offs or reliable change index. Mean scores indicate that participants’ confidence to manage their mental health difficulties increased following the ERG.

To establish if there were statistically significant differences between group means at the four different time points, a one-way repeated measures analysis of variance (ANOVA) was completed for each of the five measures. Scores on the WEMWBS showed a statistically significant improvement. Other variables showed improvements, but these were not statistically significant. Post hoc analysis also revealed the only significant differences between the time points were between pre- and post-intervention on the WEMWBS.

Table 1. Descriptive statistics and findings.

Measure	Baseline n = 31		End of ERG n = 31		T (30) T (27)	Cohen’s d	3 Month n = 12		6 Month n = 12		Within F F(3,33) F(1.657,18.232) ε		Partial ω2
	Mean	SD	Mean	SD			Mean	SD	Mean	SD
Difficulties in Emotion Regulation Scale (DERS) ‡
	128.94 a	21.97	114.06 b	26.79	4.23 **	0.76	118.33 a	24.74	113.08 a	29.28	2.60	§ p = 0.69 ns	0.11
Subscales: NONACCEPT	n = 28		n = 28
	20.75 a	5.69	19.36 a	6.32	1.75 ns ¶	0.33
GOALS
	20.21 a	4.10	18.71 b	4.45	2.25 ns ¶	0.43
CLARITY
	15.52 a	3.62	14.82 b	3.23	2.43 ns ¶	0.46
AWARENESS
	19.71 a	3.63	18.75 a	3.79	1.48 ns ¶	0.28
IMPULSE
	21.71 a	5.21	18.86 b	5.10	4.18 * ¶	0.84
STRATEGIES
	29.68 a	6.33	26.21 b	6.26	3.91 * ¶	0.74
CORE-10 ‡	24.71 a	6.70	19.87 b	8.18	3.62 **	0.65	22.25 a	8.25	20.42 a	8.93	1.88	p = 0.19 ns	0.05
Work and Social Adjustment Scale (WSAS) ‡	28.02 a	7.94	25.39 b	8.31	2.28 *	0.41	25.58 a	7.79	25.58 a	8.70	1.05	p = 0.70 ns	<0.00
Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS) †	32.53 a	8.68	39.50 b	9.78	4.48 **	0.81	37.75 a	7.46	38.50 a	7.65	2.90 *		0.11
Mental Health Confidence scale (MHCS) †	46.35 a	12.19	53.87 b	15.62	3.93 *	0.61	55.75 a	12.26	56.42 a	10.82	2.38	p = 0.09 ns	0.08

Notes: † higher score indicates improvement, ‡ lower score indicates improvement. Means with different superscripts (e.g., ^a,b) are significantly different for within participant comparisons at the level of p < 0.05. Significance of F and T tests, * denotes p < 0.05, ** denotes p < 0.001. § Greenhouse-Geisser correction applied to degrees of freedom due to violation of sphericity assumption. Effect sizes for Cohen’s d are the following: 0.2 a “small” effect size, 0.5 represents a “medium” effect size and 0.8 a “large” effect size. Effect size for partial ω2 is described as 0.01 as a “small” effect size, 0.06 as a “medium” effect size, and 0.14 as a “large” effect size [48]. ¶ indicates that a Bonferroni correction has been applied to correct for multiple tests. Significant p values for these are 0.05/6 = 0.0083. and are indicated with a *.

outlines the full quantitative analysis findings.

For the primary outcome measure, the DERS, mean scores shows a decrease between ERG start and end, an increase from end of ERG to three months and then a decrease from three months to six months. The trendline in Figure 2 shows there was a general downwards trend over time, although this was not significant. Overall, from these results it can be seen that participants’ emotion regulation skills improved significantly during the intervention, and show an unclear pattern of variability up to six months. The low sample size due to attrition makes it difficult to interpret with confidence whether treatment gains are maintained or lost.

3.2. Qualitative Results

14 participants completed qualitative interviews. Basic participant information is in

Table 2. Participant characteristics.

Participant Pseudonym	Gender	Age Range	Self-Described Difficulty	Previous Use of Mental Health Services	Individually Therapy Arranged after ERG
David	Male	36–45	OCD	No	Yes
Stacey	Female	46–55	Anxiety and depression	Yes	Yes
Sally	Female	16–25	EUPD/BPD	Yes	Yes
Nicola	Female	26–35	Trauma	Yes	No
Yvonne	Female	16–25	Not stated	Yes	Yes
Florence	Female	56–65	Depression	Yes	No
Phoebe	Female	26–35	Trauma	Yes	No
Eilidh	Female	16–25	Not stated	Yes	No
Danielle	Female	46–55	Trauma	Yes	No
Mary	Female	36–45	Anxiety and depression	Yes	Yes
Andrea	Female	46–55	Anxiety and depression	No	Yes
Charlotte	Female	26–35	Depression	Yes	Yes
Gabby	Female	16–25	Not stated	No	Yes
Heather	Female	26–35	EUPD/BPD	Yes	Yes

.

Figure 3 shows the hierarchy of the themes and subthemes.

Although a theme does not have to be present across all participants to be valid, it can be an indication of the strength of the theme and whether it is being dominated by particular voices.

Table 3. Theme presence across participants.

Theme Participant	An Ongoing Process of Learning and Change	Journey through the Group; from Ambivalence, to Awareness to Behaviour Change	Building My Toolbox	Increased Awareness and Some Implementation of New Behaviour	Not a Perfect Fix but It Has Helped	Limited Nature of the Group
David	√	√	√	√		√
Stacey	√		√	√
Sally	√	√	√			√
Nicola	√	√	√	√	√	√
Yvonne	√	√	√	√	√	√
Florence	√			√
Phoebe	√	√	√	√	√	√
Eilidh	√	√	√	√	√	√
Danielle	√	√	√	√	√
Mary	√	√	√	√	√	√
Andrea	√	√	√	√	√	√
Charlotte	√	√	√	√	√	√
Gabby	√	√	√	√	√	√
Heather	√	√	√	√	√	√

indicates which participants contributed to different themes.

3.3. Themes

3.3.1. An Ongoing Process of Learning and Change

This overarching theme is about the group being part of a longer more complex journey to recovery. For many, there was a journey within the group, from being unsure, to growing in awareness and learning new skills. There was considerable diversity within the overarching theme and in the five subthemes. For some, the group was a very significant step and was life-changing; for others, it helped, but the benefits did not last.

Andrea: “I got a lot of information… I’m still trying to put stuff in practise, but it’s harder. When you’re alone”.

Nicola: “I got a lot more out of it than I thought I would…this is making me enjoy my life a bit more… this group, for emotional dysregulation, it’s…almost the perfect thing for it”.

For most participants, the journey continues beyond the group, with most participants engaging in individual therapy after the group, as detailed in Table 2.

3.3.2. Journey through the Group; from Ambivalence, to Awareness, to Behaviour Change

Eleven participants described a journey through the group, often starting with ambivalence or even opposition to the group. This seemed to change as the content of the intervention prompted increased awareness of themselves and their own struggles with emotion, and for many this new awareness resulted in behavioural changes as participants started implementing strategies taught in the group.

Heather “I felt I wasn’t getting anything out of it. But when I took a step back, I was actually realising I got more out of it than I thought I had…it brought out an awareness in me”.

Mary “not only did I get the strategies… I really got to learn about me, which made a huge difference… Because once you know more about yourself…. the better you use the strategies... I got more than what I went for”.

3.3.3. Building My Toolbox

This subtheme describes the process of the group as one of gaining specific skills and resources that will be beneficial in the future. Many participants spoke about “the folder”, a handout of all the slides and exercises, which they used as a physical reminder to prompt them when they were feeling overwhelmed.

Andrea “if I’m in a very bad way…it’s good to have [the folder] there… try and put some stuff into practice… it’s nice to know that I have it there, it’s something that I can fall back on”.

Nicola “I’ve got that wee card that I pull out and it’s like, what are you feeling… how can you respond, what actions can you take… you’re able to carry [the card] with you”.

Participants reflected that the educational environment of the group was helpful, with the emphasis on learning skills rather than focusing on individual problems.

Yvonne “… it wasn’t so much like what’s wrong with you, it was like you’re being taught about like, all these things that are bothering you. You’re going to sit here and look into it all and understand it for yourself. Instead of, like, aw what’s wrong with you…”

David “… it was very informative… explaining how your mind works and the explanation towards surrounding different distraction technique… that I found helpful”.

However, others felt that the group atmosphere was too formal, and found this off-putting.

Phoebe “I thought it was quite closed off… it was like a meeting… I just felt pure pressured and like I didn’t want to do it…”

3.4. Increased Awareness and Some Implementation of New Behaviour

People described multiple changes due to the group. These seemed primarily to be about growing awareness of self and emotions, plus some ability to put new skills into practice. Self-efficacy and confidence also appeared to be something people gained from the group.

Mary “I know now that there probably are things that I can do once I recognise and put a name to [the emotion]…I’d maybe be more able to ask for help now … I can name my emotions more, and explain it better now”.

Nicola “it takes all the emotions that are in your head and breaks it down into smaller chunks… makes emotions seem a bit more logical…see before…I just launched into things… But now…if there’s a problem I’ll sit and think about it…. instead of just firing straight into it, I’m breaking it down for myself…it’s just good being able to do that with emotions now, whereas before I didn’t have the tools”.

There was also something important for some participants about having words to describe their experience and knowing that their experience was valid. For example, several participants realised through the group that they dissociated but had not known there was a word for it.

Phoebe “…so much clearer for me to understand and put a name to it… dissociation… I didn’t know it was a thing, I just thought it was something that I done”.

Particular aspects of the course were mentioned as bringing about changes. Consistent across almost all participants was the transformative power of the “zone of tolerance” exercise, a diagrammatic way of explaining emotional regulation and how to manage hypo- and hyper-arousal. This was a new idea for participants and helped them understand and explain their emotions.

Heather “…the red, amber and stuff, I found that a lot easier. I could just point to where I was sitting at, my mood… that was a new thing for me… felt that was pretty good”.

Other skills that were commented on were breathing exercises, visualisations, the prompt card using the acronym STOPP (Stop and step back, Take a breath, Observe, Pull back: put in some perspective, Practice what works) and trigger logs. Broadly, related to the subtheme “Building my toolbox”, it seemed having a variety of tools and opportunities to practise them led to some changed behaviour, where participants described continuing to use skills from the group.

3.5. Not a Perfect Fix but It Has Helped

Related to the previous subtheme was the idea that the group, while beneficial, was not a perfect fix. For some, this was something they realised through the group; others came already with this expectation. Participants seemed largely to have acceptance around this and a recognition that the group did move things forward for people, sometimes in profound ways.

Nicola “… it really did help… I know it’s not like foolproof, I still struggle with it, at least like it’s something now, to help me, instead of nothing before”.

Danielle “I’m able to deal with things in a different manner than what I would have done in the past…like I said, doesn’t always work, I’ve had the odd blip…But I think that’s like the mist coming over your eyes, and you wipe that away and go like, right, get a grip”.

Most participants said they would recommend the group to others, even if their own experience was suboptimal, perhaps again reflecting the recognition that there is value in it for many people.

3.6. Limited Nature of the Group

Participants struggled with the perceived limited nature of the group and many wanted longer in the group, or more out of the group. Participants felt that the skills they were taught were not always helpful, particularly when in a state of hyper-arousal, and many stated it was not possible to use the skills at this point.

Andrea “… it’s harder. When you’re alone…when you’re being triggered… like I still struggle to get myself out of being triggered… sometimes it doesn’t work”.

Nicola “There were some things that I know are useful in certain situations, but if I’m feeling really really bad they’ll maybe not help”.

There were also worries that the skills from the group would fade and get less useful over time. For some participants, they felt that this had already happened, and they had lost the benefit of the skills they had learned.

Charlotte: “It helped at the time but I’m back to where I was”.

Mary “I felt it really was helping me … if we’d kept going, if it had gone on for even longer, those techniques that we’d learned would have been even further embedded… the more you practise and the better you become at using them”.

Many participants suggested that the group should be longer, or have booster sessions, and felt this would help further implement skills. Generally, aspects of this theme reflect the quantitative results, which showed an improvement after the ERG but that this benefit was not maintained over the long term.

4. Discussion

The qualitative and quantitative results suggest the ERG significantly improves ER ability, as well as secondary measures such as general distress, wellbeing and ability to manage mental health difficulties. This seems only to be a short-term effect based on the quantitative results, which is also somewhat supported by the qualitative results with the theme “Limited nature of the group” where participants felt changes were short-lived, and participants found it hard to implement what was learned in the group after the group ended.

The quantitative results also show quite modest changes in terms of clinical significance, which the qualitative themes of “Limited nature of the group” and “Not a perfect fix but it has helped” corroborate.

However, in contrast to the quantitative results, some participants in the interviews described significant and long-lasting changes that they believe were due to the ERG in the themes “Building my toolbox”, “Increased awareness and some implementation of new behaviour” and “Journey through the group; from ambivalence, to awareness, to behaviour change”. It may be that for these individuals the quantitative measures chosen do not accurately capture a subjective sense of better wellbeing, ER and ability to manage their difficulties.

There may be several explanations for the lack of long-term impact. Behavioural theories of learning (e.g., Hayes, [49]), emphasise the role of context in learning. Context is seen as a key driver of new behavior [50,51], and in this case the group intervention is an essential aspect of the context. During the intervention, the group initially provides the context in which new behaviours (e.g., trying out skills) might be enacted, and the group also provides a source of mutual positive reinforcement. When the group ends, the contextual cues and reinforcers provided by it cease. For example, participants may remember they have the group to attend: a contextual cue to fill out a trigger log. This behaviour might be reinforced both negatively (e.g., “If I don’t fill this out I risk being judged by others”) and positively (praise and verbal reinforcement from facilitators and peers). If this behaviour does not continue after the group, this might be perceived as a failure to establish generalisation cues between salient aspects of the group and the wider context outwith it, but perhaps more pertinently as a lack of intrinsic positive reinforcement from the new behaviours: filling out a trigger log was not meaningful or useful enough to be continued. In summary, the changed behaviour has been contingent on the person being part of the group, and without the group there is a lack of context and appropriate reinforcement.

Although there is an emphasis on skills practice, the time-limited nature of the group means there are limited opportunities to practise and to make links between the examples explored in the ERG and ER choices in daily life. The theoretical underpinnings of the group, especially the TCE, posit that participants’ emotional states are based on years of prior experiences: it may therefore take a lot of dedicated practice to update these predictions.

Many participants felt the group was too short or suggested booster sessions. Dialectical behaviour therapy (DBT) groups are frequently successfully run as open, rolling programmes [52,53]. DBT has a similar patient population to the ERG and shares a skill-based focus [4], indicating a rolling approach might be of value. Rolling and open groups would provide more opportunities for skills practice over a longer time period, and may assist with longevity of improvements. However, with open and rolling groups there is a possibility some benefits of the group process could be lost. Additionally, a lack of resources may preclude services’ ability to provide additional sessions for participants.

When reflecting on long-term impact, it is important to be mindful that the group is not intended to be a panacea for ER difficulties. Explicitly in the manual and in material delivered is the idea of realistic expectations, that the intervention will be the start of change, not its endpoint [35]. This links to the theme “An ongoing process of learning and change”—i.e., for the majority of participants the group is part of a wider journey. Most participants had previous contact with mental health services and the majority went on to individual therapy. This supports the suggestion that skills practice and implementation may be driven by the context of “being in therapy”. Future implementations of the group should consider ensuring that further intervention is accessible after the intervention ends, and perhaps integrating the group as part of wider systems of care. Individual therapy could also be delivered in parallel with the ERG, as with DBT and other ER interventions [4,19].

As outlined in the introduction, one difficulty in ER research is the lack of clarity in conceptualisation of what ER is and how it can be improved. This research builds on evidence from Bacon et al. [29] of the efficacy of an ER group intervention with a clear articulation of theory and theory informing practice, as the ERG is directly built on theoretical understandings of ER. While this theoretical clarity and the importance of this may be clear to clinicians, it is unknown to what extent this influenced participants’ experiences and learning in the group. While there was considerable data on the practical aspects of the course and the skills that participants found helpful, there was very little allusion to the theoretical components of the intervention. As discussed, some of the theoretical components such as the TCE are quite a different stance to common understanding of emotions, so it would be helpful to explore participants reactions to new understandings about emotion and ER.

However, other aspects drawn directly from ER theory, such as the importance of increased awareness of emotions, being able to label emotions and choose an appropriate strategy, do seem to be well understood by participants.

Conversely, there are some tentative indications in the interviews that some of the other theoretical underpinnings may not be fully understood. A key implication of the TCE is that learning needs to be experiential, and related to this, a core component of the ERG is participants’ practising “triggering” an emotion, then using ER skills to regulate their emotions. Many participants did not seem to understand the point of the trigger experiments, and connect them with learning to use ER skills in a hyper-aroused state. Possibly because of this, some participants felt it was impossible to use ER skills when hyper-aroused, meaning when the group ended they were still unable to manage emotions when they were dysregulated.

As will be discussed, one reason for these difficulties in understanding key aspects of the group could be the facilitator’s understanding of the theory of the ERG.

4.1. Limitations and Considerations for Further Research

A considerable amount of the data collection took place during the COVID-19 pandemic. Initial research showed worsening general mental health in the population during the pandemic [54,55] and it may be that the lack of a prolonged benefit of the group could be due to the negative impact of the pandemic. Indeed, several interview participants reflected that their mental health had been worse due to the pandemic.

Another limitation is the lack of a control group, meaning it is unknown whether improvements—or lack of—were due to the intervention or other factors in participants’ lives. In future studies, an active control group would be best, such as a general supportive group, which has been used successfully in other similar studies [26,56]. This is particularly important given findings of a systematic review showing that while ER skills are improved by a range of ER interventions, there is limited evidence that this is superior to any other group intervention not specifically targeting ER [57].

A further limitation is lack of information about facilitators’ training. In Bacon et al. [29], facilitators observed, then co-facilitated a group, and attended quarterly supervision to enhance fidelity. In this study, there was no information on whether facilitators had received training on the ERG or observed the group. This means it is unknown how closely the ERG model was adhered to, which is key given the emotion theory the group is built on, the TCE. If this rationale is not both understood by facilitators and successfully imparted to participants, the skills are unlikely to be used effectively or practised after the group has finished. This in turn would mean skills were unlikely to be utilised effectively in the long term.

4.2. Conclusions

This mixed-methods study shows that the ERG leads to at least short-term improvements to ER and secondary measures of general distress, wellbeing and ability to manage mental health difficulties. Participants describe a range of benefits, but some difficulties in implementing change when the group ends. It is unclear if short-term improvements are sustained long-term. There may be many reasons for this, including difficulties with facilitator training and fidelity meaning the full model is not being implemented, as well as the short-term nature of the group meaning that there is not sufficient time for learning to produce full implementation.

The ERG should be considered an option for participants struggling with ER in accordance with a matched care model, and further intervention should be offered as appropriate when the ERG ends. Adaptations to the ERG model might usefully be considered, such as booster sessions or rolling groups to provide further opportunities for embedding skills.