Family discussion group therapy for major depression: a brief systemic multi-family group intervention for hospitalized patients and their family members

Beneficial factors in family discussion groupsof a psychiatric day clinic: perceptions by thetherapeutic team and the families of thetherapeutic process1

Gilbert M. Lemmens,a Saskia Wauters,b MagdaHeireman,c Ivan Eisler,d Germain Lietaere andBernard Sabbef

This paper reports a pilot investigation of the perception of helpfulevents by the therapeutic team and the families in two family discussiongroups (FDGs) of a psychiatric day clinic. All participants of the FDG,including therapists and observers, filled in questionnaires measuringevents helpful for the individual, for the family and for the group aftereach FDG session. The results showed that the therapeutic team and thefamilies diverged in their overall perception of which factors wereimportant in family discussion group therapy. The therapeutic team sawthe relational climate and the structural aspects of the group (includinggroup involvement and support from the group), and specific thera-peutic interventions as more helpful than the families. The processaspects in the group members (including the experiencing of commu-nality and gaining insight) were, on the other hand, more frequentlymentioned by the families than by the therapeutic team. The clinicalimplications of these findings and suggestions for future research arediscussed.

2003 The Association for Family Therapy and Systemic Practice

The Association for Family Therapy 2003. Published by Blackwell Publishing, 9600Garsington Road, Oxford, OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.Journal of Family Therapy (2003) 25: 41–630163–4445

a Consultant Psychiatrist, Department of Family Therapy, University HospitalLeuven, Kapucijnenvoer 33, B-3000 Leuven, Belgium. E-mail: [email protected]

b Research worker in Family Therapy, Department of Family Therapy,University Hospital Leuven, Belgium.

c Senior Lecturer in Family Therapy, Department of Family Therapy, UniversityHospital Leuven, Belgium.

d Senior Lecturer in Clinical Psychology, Section of Psychotherapy, Institute ofPsychiatry, Kings College, University of London, UK.

e Professor of Psychology, Catholic University of Leuven, Belgium.f Professor of Medical Psychology, University of Antwerpen, Belgium, Associate

Professor of Psychiatry, UMC St-Radboud Nijmegen, The Netherlands.1 Accepted by the previous editor of the journal.

Introduction

Multiple family therapy (MFT) has been used in the treatment of awide variety of clinical conditions including affective disorders(Anderson et al., 1986), schizophrenia (Kuipers et al., 1992;McFarlane et al., 1995; Dyck et al., 2000), eating disorders(Slagerman and Yager, 1989; Dare and Eisler, 2000; Scholz andAsen, 2001), substance abuse (Clerici et al., 1988), child sexualabuse (Asen et al., 1989) and chronic medical illnesses (Gonzales etal., 1989; Steinglass, 1998). While there is empirical evidence forthe efficacy of multiple family interventions, for example, in reduc-ing relapse rates in schizophrenia (Baucom et al., 1998; McFarlane,2002), there has been little research on the process of change inMFT. The described curative factors and mechanisms of change inthe literature consist mainly of impressionistic accounts by thera-pists of what they believed to be the important factors in their ownpractice. Identified factors include learning by analogy (Laqueur,1972), the establishment of a community with shared experiences(Steinglass, 1998), overcoming stigmatization (Asen et al., 2001),generating new and multiple perspectives on illness and family(Steinglass, 1998; Asen, 2002; Bishop et al., 2002), and peerconfrontation (Leichter and Schulman, 1974). Other suggestedtherapeutic factors are the identification with the experience oftheir counterparts in other families (Rolland, 1994), experiencinghope and progress (Laqueur, 1976) and trying out new adaptivepatterns of relating and coping (Strelnick, 1977; O’Shea andPhelps, 1985).

Our own experience of running a family discussion group project(FDG) in a psychiatric day clinic and the positive feedback fromstaff and families encouraged us to explore in a systematic way theexperiences of the families and the therapeutic team during thegroup sessions. We were interested in particular in what is perceivedas helpful by the participants in order to gain a better understand-ing of how the therapy in these groups works, and to make the treat-ment more effective. Research on helpful factors provides a veryparticular perspective on the process of therapy, which has impor-tant strengths but also some clear limitations, including a lack ofspecificity and the risk of bias due to the subjective nature of thedata that rely on memory and self-report. It has been used widely inresearch on experiential-humanistic therapies in keeping with theunderlying philosophical assumption of these approaches, which

42 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

takes the client’s ongoing awareness of his of her own experiencesas the primary datum for therapy (Greenberg et al., 1994) but is alsoclearly relevant to research on family therapy.

This study used an adapted version of the methodology of open-ended questions exploring helpful factors in experiential grouptherapy (Bloch et al., 1979; Dierick and Lietaer, 1990; Crouch et al.,1994; Barrett-Lennard, 1998) to identify beneficial factors at threedifferent levels of perception: events helpful for the individual, forthe family and for the group.

Method

Sample selection

All patients were recruited from the day clinic of the UniversityPsychiatric Clinic Lovenjoel. Family discussion groups, each withfour to six patients and their families, were conducted over a two-year period as part of the therapeutic programme (Naeyaert et al.,1998). The inclusion criteria for this study were: (1) involvement inthe therapeutic programme of the day clinic, and (2) having at leastone significant relationship with a family member. Patients with anacute psychotic episode and patients who had already participatedin a previous group were excluded from the study. No other selec-tion criteria were used. All participants were given written andverbal information about the research project. A group cycle startedafter four or five patients and their relatives gave written, informedconsent to participate in the study.

Participants

The characteristics of the patients including age, DSM-IV diagnosis(American Psychiatric Association, 1994) and group compositionare given in Table 1. Five patients and six relatives participated inthe first group. Of those five families, one family stopped after twosessions and one family joined only after session four and thereforecompleted only two sessions. The second group consisted of fivepatients and six relatives. In this group, one family stopped afterthree sessions. Data on all families, including those that attendedonly some of the group sessions, were included in the analyses.

Two family therapists/psychiatrists (a male and a female) ledboth groups. The observation team behind a one-way screen

Beneficial factors in family discussion groups 43

2003 The Association for Family Therapy and Systemic Practice

44G

ilbert M. L

emm

ens et al.

2003 T

he Association for Fam

ily Therapy and System

ic Practice

TABLE 1 Description of the family discussion groups

FDG group Family Composition of group Mean age DSM-–IVDiagnosisof patients(N=2) (N=10) of patients

1 5 2 Female patients, 3 male 30 Identity problem (301.82)/somatization disorderpatients, 1 sister, 1 girlfriend, (21–37) (300.81)4 spouses Adjustment disorder with mixed anxiety and

depressed mood (309.28)Severe major depressive disorder without psychoticfeatures, single episode (296.23)Severe major depressive disorder without psychoticfeatures, recurrent (296.33)Schizoaffective disorder (295.70)

2 5 4 Female patients, 1 male 38 Psychotic disorder NOS (298.9)patient, 2 mothers, 1 father, (24–52) Moderate major depressive disorder, single episode3 spouses (296.22)

Moderate major depressive disorder, single episode(296.22)/anxiety disorder NOS (300.00)Severe major depressive disorder without psychoticfeatures, single episode (296.23)Schizophreniform disorder (295.40)

consisted of five mental health professionals from the day clinicand/or trainees in family therapy.

Organization of the family discussion groups

The FDG 1 and 2 cycles consisted of six and five sessions respec-tively. The sessions were held once a fortnight and lasted approxi-mately ninety minutes with a fifteen-minute ‘coffee break’ after onehour. Each session was videotaped. The used model was a systemicmultiple family therapy group (Lemmens et al., 1999, 2001). Majorgoals included increasing of the coping and problem-solving strate-gies in the families, enhancing the communication between thefamily members, and challenging family myths and rigid familyscripts. The discussion focused mostly on the impact of the psychi-atric illness on the family unit, the coping strategies of patients andfamily members, family interactions and belief systems (Rolland,1998), resources and family life-cycle issues. Information aboutillness and treatment was given to the group mainly in the contextof exchanges of the group members’ experiences and not in a struc-tured way, such as occurs in psychoeducational interventions.

Measurements and data collection

The session evaluation questionnaires were adapted for the studyfrom questionnaires exploring perceptions of therapeutic factors ingroup therapy and growth groups (Dierick and Lietaer, 1989, 1990;Crouch et al., 1994) (Appendix 1). The open-ended questions gaveinformation regarding specific concrete experiences that the partic-ipants considered helpful for (1) the individual, (2) the family, and(3) the group.

All family members as well as therapists and observers were askedto fill in questionnaires after each session. Data were collected for atotal of ten patients, twelve family members, four therapists and tenobservers during two different group cycles. Patients and familymembers participated on average in 3.5 sessions, therapists andobservers in 4.7 sessions. All present participants responded.

Data analysis

The content analyses were performed by two research workers andthe male family therapist about eighteen months after the ending

Beneficial factors in family discussion groups 45

2003 The Association for Family Therapy and Systemic Practice

of the last group. The inter-rater reliabilities were calculated by anindependent statistician.

All responses to the open questions were broken down by the tworesearch workers into meaningful segments, each containing just oneelement of the group process that was perceived as helpful. For exam-ple, after the first session a patient responded in the following way tothe question about events helpful to herself: ‘/the similar experiences ofthe others/(experiencing of communality) and/the understand-ing/(empathy and feeling understood)’, for the family: ‘/the similarexperiences of the spouses/(experiencing of communality)’, and for thegroup: ‘/the feeling of togetherness/(group cohesion and feeling good)’.

The categorical scoring system was adapted from studies explor-ing perceptions of therapeutic factors in psychotherapy and growthgroups (Dierick and Lietaer, 1989, 1990), which quantify qualitativecomments about the therapeutic process in order to study the rela-tive importance of different therapeutic factors (Bloch andReibstein, 1980; Crouch et al., 1994; Barrett-Lennard, 1998). In thisscoring system, all beneficial treatment experiences were scored inthree different main categories, each containing different subcate-gories. The main categories, which each pointed to a cluster ofsome specific helpful aspects of the therapeutic process, were (A)the relational climate and structural aspects of the group, (B) thespecific interventions by group members and therapist, (C) theprocess aspects in the group member (i.e. the process that theparticipant experiences). Unclassifiable responses, or responseswhich were left blank, no helpful events reported or a positiveresponse containing no details, were combined to form a furthercategory (D).

The three raters piloted the scoring system using responses of aseparate family discussion group (not included in this study).Based on this pilot, thirty-three additional subcategories aboutfamily and illness issues and aspects of the multi-family groupformat were added (for example, the category ‘feeling under-stood by group members with the same position in the family’).Every sixth questionnaire was then selected and its responsesegments were independently scored by the three raters. Thecalculation of the reliabilities between the raters and between thedifferent categories, together with a new qualitative analysis of thedifferences in rating, pointed to the necessity of another adapta-tion of the scoring system, which consisted mainly of mergingsome of the subcategories. Another one-sixth of the questionnaire

46 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

material was selected and the inter-rater reliability (Kappa coeffi-cient: 0.88, 0.90, 0.91) calculated. The rest of the response segmentswere then further independently scored by two persons and theinter-rater reliability (Kappa coefficient: 0.87) for all scoredresponse segments between these two raters remained sufficientlyhigh. Because of the close resemblance between the answers fromthe three levels (individual, family and group), the same categorysystem was used for the analysis of all responses. A short descriptionof the final category system together with some examples ofresponse segments are given in Appendix 2.

In order to investigate the importance of the different beneficialfactors within each level, frequencies and percentages of theresponse segments were calculated.

Results

Helpful factors as perceived by the families

In total there were seventy-seven responses of twenty-two patientsand family members for each question (individual, family andgroup). Table 2 shows that the families indicated a higher level ofbeneficial experiences for the individual (75 per cent) and thegroup (64 per cent) than for the family (55 per cent), for whichlevel 40 per cent of the sessions were reported as containing nohelpful events. A closer look at the higher frequency of ‘no helpfulevents’ responses revealed that the majority of these came from thefamilies in FDG2, especially for the individual and the group level.In this group, two of the families (one of whom dropped out oftreatment) responded after almost every session (at all three levels)that there were no helpful events to report.

The distribution of the helpful response segments (i.e. excludingcategory D: see Table 2) shows clearly that the perception of bene-ficial factors by the families was quite constant for the three differ-ent levels (individual, family, group). About a quarter of the helpfulfactors was situated in the relational climate and structural aspectsof the group, and almost three-quarters in the process aspects in thegroup member. Specific interventions were rarely mentioned bythem. A short description of the most important beneficial factorsfor the three levels will now be given.

The two most frequently mentioned factors by the families werebecoming conscious of/having insight in the family (‘despite the

Beneficial factors in family discussion groups 47

2003 The Association for Family Therapy and Systemic Practice

48G

ilbert M. L

emm

ens et al.

2003 T

he Association for Fam

ily Therapy and System

ic Practice

TABLE 2 The distribution of helpful factors in the perception of the therapeutic team and the familiesT/O P/F

Ind Fam Group Ind Fam GroupMain categories/subcategories (%) (%) (%) (%) (%) (%)

A. Relational climate and structural aspects of the group 35.4 23.7 53.1 23.5 27.8 24.41 dedication, commitment and involvement 12.2 5.3 17.0 3.7 1.8 2.6

• within the group 11 4.4 17.0 2.5 1.8 2.6• within the family 1.2 0.9 0 1.2 0 0

2 space and freedom 1.2 3.5 8.2 1.2 5.5 3.83 empathy and feeling understood 0 0 0.7 6.2 5.5 1.34 confirmation, appreciation, and support 15.8 5.3 1.4 7.4 5.5 2.6

• from the group 11 2.6 1.4 4.9 1.8 2.6• from the therapists 1.2 1.7 0 0 0 0• from the family 3.7 0.9 0 2.5 3.7 0

5 authenticity and transparency 1.2 0.9 0 2.5 3.7 2.6• within the group 1.2 0.9 0 1.2 1.8 2.6• within the family 0 0 0 1.2 1.8 0

6 group cohesion and feeling good 0 0.9 12.2 1.2 5.5 97 informal, beneficial contact outside the sessions 0 0 1.4 0 0 08 structural aspects and facilitating group composition 4.9 7.9 12.2 1.2 0 2.6

B. Specific interventions by group members or therapist 21.9 15.8 12.2 3.7 1.8 1.39 offering therapeutic techniques 2.4 3.5 4.1 1.2 1.8 0

10 stimulating, clarifying and interpretative interventions 9.8 7.0 4.8 2.5 0 1.311 reaction and feedback 8.5 4.4 0 0 0 012 interrupting a disturbing process 1.2 0.9 3.4 0 0 0

C. Process aspects in the group member 42.7 60.5 34.7 72.8 70.4 74.413 self-revelation 8.5 4.4 4.8 6.2 3.7 1.314 self-exploration and interpersonal exploration 8.5 7.9 0.7 4.9 0 0

• self-exploration 0 0.9 0 0 0 0• discussing mutual relationship with family members 6.1 4.4 0 2.5 0 0• expressing feelings towards family members 2.4 1.7 0 2.5 0 0• interpersonal exploration with other group members 0 0.9 0.7 0 0 0

15 discussing different topics 6.1 14.9 7.5 9.9 9.3 1.3• existential and family-related 3.7 6.1 2.0 3.7 5.5 0

Beneficial factors in fam

ily discussion groups49

2003 T

he Association for Fam

ily Therapy and System

ic Practice

TABLE 2 ContinuedT/O P/F

Ind Fam Group Ind Fam GroupMain categories/subcategories (%) (%) (%) (%) (%) (%)

• illness-related 2.4 4.4 2.0 2.5 1.8 1.3• other 0 4.4 3.4 3.7 1.8 0

16 experiencing of communality 9.8 14.9 12.9 17.3 27.8 28.2• with other group members 7.3 9.6 10.9 14.8 25.9 24.4• with other families 2.4 5.3 2.0 2.5 1.8 3.8

17 learning by observation 1.2 5.3 2.0 3.7 1.8 6.418 becoming conscious of/having insight into: 4.9 11.4 1.4 21 14.8 16.7

• one-self 0 0 0 3.7 1.8 5.1• family 1.2 3.5 0 11.1 13 2.6• illness 2.4 6.1 0.7 4.9 0 2.6• other 1.2 1.7 0.7 1.2 0 6.4

19 experiencing the helpful potential of the group/ 0 0.9 3.4 7.4 9.3 14.1experimenting with new behaviour/capacity to helpothers

20 experiencing hope and progress 2.4 0.9 0 2.57 1.8 1.321 experiencing relief, relaxation and humour 0 0 2.0 0 1.8 3.8

TOTAL ABC 100 100 100 100 100 100

Frequency of response segments of ABC 82 114 147 81 54 78D. Remaining categories (per cent of total responses) 5 8 6 26 50 37

22 blank responses 0 0 0 0 5 823 no helpful events reported 5 6 6 25 40 2924 yes, but unclassifiable or unclear answers 0 0 0 1 5 025 yes with no explanation 0 2 0 0 0 0

Frequency of total responses 66 66 66 77 77 77% of total yes responses 95 94 94 75 55 64

Notes: T/O = therapists and observers, P/F = patients and family members, Ind = individual level, Fam = family level, group = group level, %= percentage of response segments of ABC. (If a response contained different meaningful response segments of the same main category, thisresponse is scored only once in that main category; if a response contained different meaningful response segments of different main cate-gories, this response is scored in every one of those main categories; the frequency of the total response-segments of main categories ABC ishigher than the frequency of all yes responses with meaningful comments, because each response can contain different response-segments.)

differences between me and my mother, I like her’, ‘the distance in our rela-tionship is caused by our lack of commitment’, ‘you don’t have to sacrificeyourself all the time for your partner’), and experiencing of communal-ity, especially with other group members (‘the similar feelings andquestions of the others’, ‘other spouses also get negative reactions about theirrelationship with a “depressed person” ’, ‘others also suffer from anxiety’,‘you’re not alone’). Another important factor for all three levels wasthe helpful potential of the group/experimenting with new behavi-our/capacity to help others (‘it was an interesting conversation’, ‘thewhole session’). Further, the families benefited apparently fromdiscussing different topics (‘the topic of having another child’, ‘the topicof the impact of depression on the family’), mostly for themselves and thefamily level, and less for the group level and self-revelation (“dare totalk about oneself”) mostly for the individual level.

Several aspects of the relational climate and structural aspects ofthe group were mentioned as helpful by the families for the threelevels, although not very frequently. Most important group aspectswere group cohesion and feeling good (‘the feeling of togetherness’, ‘thefeeling that everybody belongs to the group’), empathy and feeling under-stood (‘the understanding for the patient’s viewpoint’, ‘the understandingfor everybody’s situation’, ‘the understanding of my husband by the otherfamily members’), and confirmation, appreciation and support (‘it’s asupport’, ‘the support from my husband’). Specific interventions wererarely mentioned.

Helpful factors as perceived by the therapists and observers

In total there were sixty-six responses of fourteen therapists andobservers for each question (individual, family and group). Table 2shows that, overall, the therapeutic team almost always identifiedbeneficial experiences happening for the three investigated levels.More than 90 per cent of their responses pointed to beneficial expe-riences for each level.

Unlike the families, the distribution of the beneficial factors,however, varied quite considerably between the different levels ofthe therapists and observers’ viewpoint. The relational climate andstructural aspects of the group were seen to be particularly impor-tant for the group, while the process aspects in the group memberswere considered to be more beneficial for the family. At the indi-vidual level these two categories were seen by the therapists andobservers as almost equally helpful (relational climate and group

50 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

structure accounting for 35 per cent and group member process forjust over 40 per cent). In contrast with the families, the profession-als also regularly mentioned specific interventions as helpful (over20 per cent at the individual level). A more detailed description ofthe mentioned factors for each level follows.

Experiencing of communality, especially with other groupmembers (‘the similar feelings and questions of the others’, ‘the similarproblems’), and discussing different topics (‘the topic of having anotherchild’) were also, like the families, perceived as beneficial by thera-pists and observers for all three levels. Further, they found self-revelation (‘for a partner to talk about his fear of his wife not improving’,‘a partner told that she couldn’t cope any more’ ) an important beneficialfactor mostly for the individual level. Therapists and observersmostly mentioned insights into the illness (‘staying in bed can also bea way of protesting’, ‘an illness offers opportunities for learning’) and forthe family level as helpful, where the families benefited apparentlymore from insights into family functioning.

Most important group aspects were: dedication, commitmentand involvement within the group (‘the active participation of all groupmembers’, ‘the commitment of a mother’), structural aspects and facili-tating group composition (‘the same age of the group members’), confir-mation, appreciation and support from the group (‘the support thata wife got from another woman when her husband didn’t agree with her’),and finally group cohesion and feeling good (‘the good atmosphere inthe group’) and space and freedom (‘some difficult topics could be openlydiscussed’). The latter two factors were mostly helpful for the grouplevel. Surprisingly, empathy and feeling understood were rarelymentioned as beneficial group aspects by the therapeutic team.

Specific interventions most frequently mentioned by the thera-pists and observers were stimulating, clarifying and interpretativeinterventions (‘the therapist asking a woman if she wanted to add some-thing after her husband had spoken’), and reaction and feedback (‘thatother group members pointed out to a patient that she wasn’t left alone by herhusband’). These interventions were especially important for theindividual and the family level.

Discussion

In this pilot study we investigated helpful factors in two familydiscussion groups at a psychiatric day clinic and looked at how theywere perceived by the therapeutic team and by the families. Before

Beneficial factors in family discussion groups 51

2003 The Association for Family Therapy and Systemic Practice

discussing our findings the potential limitations of the study needto be addressed.

The methodology we used, that of the open-ended question-naire, can obviously access only those therapeutic processes thatoperate at a conscious level and that are sufficiently salient for thesubjects to mention them. While this is a limitation, the method hasthe advantage of being able to reveal a range of helping factorswhich are not easily accessible by other techniques of processmeasurements such as structured questionnaires (McKenzie, 1987;Alexander et al., 1994). By categorizing qualitative data, some of therichness of the information is inevitably lost, but on the other handit allows comparisons to be made which can reveal informationabout the relative importance of the different perceived beneficialfactors (Greenberg et al., 1994).

The comparison of the self-perception of the families with theoutside observation of the therapists and observers, while clearly ofconsiderable interest, is not straightforward and also raises poten-tial problems. First, the multiple observations of the therapeuticteam (i.e. they had the opportunity to observe simultaneously morethan one patient or family member during a session) compared tothe single observation of the families makes it difficult to makedirect comparisons, particularly when considering the frequency ofno helpful events reported (there is clearly a difference betweennot experiencing any helpful events oneself and reporting as anobserver that nothing helpful occurred for any of the families in thegroup). Further, the responses of the families and the therapeuticteam are obviously not directly comparable because, for example,the questions were not the ‘same’ when put to them. We mightexpect both groups to have different criteria, different meaningsascribed to ‘helpful’, or different demand characteristics. In a way,differing views may have arisen from dissimilar reference pointsrather than from opposing opinions about therapy. One could alsoargue that the professionals and the family members enter into theprocess with different expectations and a different understandingof what might be happening, and that therefore one is not compar-ing like with like. While this is undoubtedly the case and makes anydifferences found more difficult to interpret, it does not detractfrom the value of making such comparisons. Indeed we wouldargue that it is in the confrontation of the different viewpoints andour attempts to understand the differences that can lead to newinsights into the therapeutic process (Bloch and Reibstein, 1980).

52 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

As Bloch and Reibstein (1980) point out, the finding that theparticipants report helpful events, which can reflect particularpossible therapeutic factors, does not in itself mean that thesefactors automatically result in a positive outcome, for which nomeasurements were performed in this study. The experimentaldesign with the adaptation of questions, hypotheses and methodo-logy from experiential group psychotherapy research for this studyalso risked the overemphasis of the group aspects of the familydiscussion group. The examination of helpful events for threedifferent levels as well as the inclusion of different family categoriesin the scoring system were designed to overcome this. The size ofthe sample was clearly a further limitation, requiring that anyconclusions drawn are treated with a degree of caution, andprecluded the possibility of investigating specific group effects onthe clinical perception of helpful events.

The findings of this study showed, first, that most participants,the therapeutic team as well as the families, experienced helpfulevents during the group sessions. The professionals respondedmore often that beneficial experiences happened in sessions, andthey also mentioned more beneficial factors for each beneficialexperience than did the families. As mentioned earlier this is, atleast in part, likely to be due to their making simultaneous observa-tion of more than one family, but is also likely to have been influ-enced by their enthusiasm for the multiple family therapy format.Perhaps for similar reasons, they rarely indicated that ‘no helpfulevents’ happened in contrast to the families of whom 25 to 40 percent chose the ‘no helpful events’ category at one or more levels.Most of the ‘no helpful events reported’ came from FDG2, whichcould have been due to a number of factors. Plans to move the dayclinic caused a sense of insecurity in patients and day clinic staffduring this group, and may have influenced the normal groupprocesses and the therapeutic context in this group. Another factorwas the less clear identification between the different families inFDG2 compared to FDG1, which consisted mainly of patients andtheir spouses within a more closely related family life-cycle stage. Itis also possible that some family problems, although perhaps veryrelevant at the time, could not be discussed in the multiple familygroup format. Perhaps those families would have benefited morefrom single family therapy.

The families made relatively little differentiation between helpfulevents at the three predetermined levels (individual, family, group).

Beneficial factors in family discussion groups 53

2003 The Association for Family Therapy and Systemic Practice

Events, experienced as helpful for the individual, seemed to be seenautomatically as beneficial for the family as well as the group, as if ahelpful event produced a general and overall beneficial effect andas if therapy was rather experienced as a whole. This could alsoexplain that if no helpful events were reported by them, it was oftenmentioned for all three levels. In contrast, therapists and observersseemed to believe that particular beneficial factors played a moreimportant role for certain levels in the group. Their view of thetherapy process seemed more complex (or more fragmented) as asum of different factors, which added to the treatment experience.

Some of the differences between families and professionals werestriking. For instance, specific interventions barely figured among thelist of helpful factors from the families’ point of view in contrast tothe professionals who saw them as a significant contributor to thetherapeutic process. This is consistent with other studies of group(Dierick and Lietaer, 1989, 1990; Yalom, 1995) and individualpsychotherapy (Glass and Arnkoff, 2000). The findings might beinterpreted as showing that the group process plays itself outprimarily between the group members, and that the therapist’s roleis primarily to facilitate this process. Alternately it may mean thatspecific interventions such as reaction and feedback, stimulatingand clarifying by the therapists were necessary to create a thera-peutic context but were in themselves less salient for the familiesthan the resulting group processes and were therefore ‘less visible’to them. The above difference may of course be explained in anumber of ways and at one level may be accepted as simply repre-senting two different observer perspectives, both of which are validin their own right. Nevertheless, they point to the fact that as thera-pists we need to be more aware of how the therapeutic process isexperienced by families and the meaning that they attach to theseexperiences.

Consistent with previous observations (Laqueur, 1976; Strelnick,1977; O’Shea and Phelps, 1985; Steinglass, 1998), this study foundthe experiencing of communality to be the most important helpfulfactor in the perception of the families. The following exampleillustrates this for the different levels.After the first session of FDG1, a spouse of a depressed patient mentionedas helpful for him: ‘the other partners have similar experiences and feel-ings’. He then indicated that ‘the similar experiences and feelings of theother patients’ were helpful for his family. His spouse noted in her answeras helpful for herself: ‘the similar experiences of the other group

54 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

members, and their understanding’, and as helpful for her family: ‘thesimilar experiences of the spouses of the patients’. After the second sessionhe indicated that ‘the predominantly similar feelings of the differentgroup members’ were helpful for the group.

In the initial phase of a family discussion group therapists oftenfocus on finding ‘common denominators’ as a therapeutic strategyto increase group cohesion (Laqueur, 1976). The high rate ofresponses from the families in this category (and the fact that it washigher than the therapists and observers) highlights the impor-tance of this factor and suggests that it is perhaps even more bene-ficial than perceived by the therapists. In the above-mentionedexample, experiencing of communality not only makes the patientfeel that she is not alone with her depression but simultaneouslymakes the spouse recognize himself in another husband of adepressed patient. These simultaneous processes help both of themto realize that their reactions, feelings and struggles are normal andmake them feel less isolated. Family members often got in touchwith other families with similar problems for the first time in ourgroups.

The second important helpful factor in the eyes of the familieswas becoming conscious of/having insight, particularly in their familyprocess. Despite the limited number of group sessions and the pres-ence of an important psychiatric disease in the family, they quicklygained insight into their own family as well as the illness. Althoughinformation about family functioning, psychiatric illnesses andtreatment was not structured in any way, the multi-family groupformat itself probably stimulated the generation of these ‘insights’.The large number of differences and similarities between the groupmembers on various levels (individual, family, group) help thegroup members to broaden their views and to generate differentperspectives (Leichter and Schulman, 1974). It emphasizes the factthat family discussion groups offer a unique forum to gain differentinsights by combining the experiences of patients, families andtherapists. But it also raises the question of which therapeutic strat-egy would be the most productive in generating these insights:psychoeducation by the therapist, the shared experiences of theparticipants, or a combination of the two. Somewhat surprisingly,the therapists and observers only rarely mentioned insights as help-ful (with the exception of insights into the illness at the familylevel). This could have been due to the fact that the therapists and

Beneficial factors in family discussion groups 55

2003 The Association for Family Therapy and Systemic Practice

observers tended not to focus on ‘dysfunctional’ dynamics, interac-tions and cognitive processes in the individual families and may nothave therefore expected to find insight into individual or familyfunctioning to be an important therapeutic factor. Equally, as weargued above, it could also be the effect of the use of a methodo-logy, which tends to downplay factors that do not have a very highsalience for the observer.

The possibility of experiencing the helpful potential of the group/toexperiment with new behaviours in relation to different persons (a familymember, another father, a child) in the group (Igodt, 1983) alsoseemed to be helpful mainly in the perception of the families.However, most response segments in this category gave rather amore general description of the helpful potential of the group thanpointing at factors such as experimenting with new behaviours or acapacity to help others.

The exchange on different topics in the family discussion groupalso seemed to be helpful in the eyes of all participants and foralmost all levels. The group creates a forum for discussing familyand illness-related topics, which may be difficult to attain within asingle family unit. Moreover, discussing these issues in the multi-family group context has the effect of normalizing communicationpatterns and contents between the family members, and also allowsfor the beneficial effect of the ‘outsider witness group’, which otherfamilies often played in these groups (Asen, 2002). The discussionprobably stimulates other cognitive processes, such as self-reflectionand insight. Self-revelation was apparently more helpful in theeyes of the therapists and observers than of the families. When self-revelation occurs in a group, it is usually very noticeable and, to theobservers, a very visible event. It is not at all clear why it is seldomidentified by the families. Perhaps they attach more importance tothe processes such as starting a discussion or gaining insight whichcan follow from it rather than the actual self-revelation itself. It canalso be diluted or disappear in the discussion of the group.

About a quarter of all therapeutic factors was situated in the rela-tional climate and structural aspects of the group in the perception of thefamilies for the three levels. They probably have more of an indirecteffect, as they are an inseparable part of the therapeutic frameworkof a group (like water for a fish). For the latter reason, familiesprobably take them for granted. Almost all subcategories were moreor less mentioned by them. The most frequently mentioned benefi-cial factors were group cohesion, empathy and feeling understood,

56 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

and confirmation, appreciation and support. These are in theireyes necessary group aspects for having a successful family discus-sion group. Therapists and observers seemed to have paid moreattention to these therapeutic factors, especially for the individualand the group levels. They viewed a group as more helpful ifdifferent family members were present, if the participants wereactively involved in and dedicated to the group, where they expe-rienced confirmation and support, group cohesion and an openatmosphere.

Conclusion

This pilot study has revealed important information about what thetherapeutic team and families experience as helpful in a familydiscussion group of a psychiatric day clinic. Important differencesin beneficial therapeutic factors between both viewpoints were iden-tified, and provide some insight into how families themselvesperceive the process of change during treatment and how familytherapists can facilitate family learning in this form of psychother-apy. The study of course raises further questions that need to beaddressed by future research. We are currently undertaking an RCTof family discussion group and single family therapy which shouldhelp clarify some of the questions raised by this study.

ReferencesAlexander, J.F., Holtzworth-Munroe, A. and Jameson, P. (1994) The process and

outcome of marital and family therapy: research review and evaluation. In E.A.Bergin and S.L. Garfield (eds) Handbook of Psychotherapy and Behavior Change.New York: John Wiley and Sons.

American Psychiatric Association (1994) Diagnostic and Statistical Manual of MentalDisorders, 4th Edition. Washington, DC: American Psychiatric Association.

Anderson, C.M., Griffin, S., Rossi, A., Pagonis, I., Holder, D.P. and Treiber, R.(1986) A comparative study of the impact of education vs. process groups forfamilies of patients with affective disorders. Family Process, 25: 185–205.

Asen, E. (2002) Multiple family therapy: an overview. Journal of Family Therapy, 24:3–16.

Asen, E., Dawson, N. and McHugh, B. (2001) Multiple Family Therapy: TheMarlborough Model and its Wider Applications. London: Karnac.

Asen, K.E., George, E., Piper, R. and Stevens, A. (1989) A systems approach tochild abuse: management and treatment issues. Child Abuse and Neglect, 13:45–57.

Barrett-Lennard, G.T. (1998) Carl Rogers’ Helping System: Journey and Substance.London: Sage.

Beneficial factors in family discussion groups 57

2003 The Association for Family Therapy and Systemic Practice

Baucom, D.H., Shoham, V., Mueser, K.T., Daiuto, A.D. and Stickle T.R. (1998)Empirically supported couple and family therapy for marital distress and adultmental health problems. Journal of Consulting and Clinical Psychology, 66: 53–88.

Bishop, P., Clilverd, A., Cooklin, A. and Hunt, U. (2002) Mental health matters: amulti-family framework for mental health intervention. Journal of Family Therapy,24: 31–45.

Bloch, S. and Reibstein, J. (1980) Perception by patients and therapists of thera-peutic factors in group psychotherapy. British Journal of Psychiatry, 137: 274–278.

Bloch, S., Reibstein, J., Crouch, E., Holroyd, P. and Themen, J. (1979) A methodfor study of therapeutic factors in group psychotherapy. British Journal ofPsychiatry, 134: 257–263.

Clerici, M., Garini, R., Capitanio, C., Zardi, L., Carta, I. and Gori, E. (1988)Involvement of families in group therapy of heroin addicts. Drug and AlcoholDependence, 21: 213–216.

Corsini, R. and Rosenberg, B. (1955) Mechanisms of group psychotherapy:processes and dynamics. Journal of Abnormal and Social Psychology, 51: 406–411.

Crouch, E.C., Bloch, S. and Wanlass, J. (1994) Therapeutic factors: interpersonaland intrapersonal mechanisms. In A. Fuhriman and G.M. Burlingame (eds)Handbook of Group Psychotherapy; An Empirical and Clinical Synthesis. New York:John Wiley & Sons.

Dare, C. and Eisler, I. (2000) A multi-family group day treatment programme foradolescent eating disorder. European Eating Disorders Review, 8: 4–18.

Dierick, P. and Lietaer, G. (1989) Therapeutische factoren in groepstherapie engroeigroepen: een exploratorisch onderzoek bij groepsleden en therapeuten.Psychologica Belgica, 29: 1–23.

Dierick, P. and Lietaer, G. (1990) Member and therapist perceptions of therapeu-tic factors in therapy and growth groups: comments on a category system. In G.Lietaer, J. Rombouts and R. Van Balen (eds) Client-centered and ExperientialPsychotherapy in the Nineties. Leuven: Universitaire Pers Leuven.

Dyck, D.G., Short, R.A., Hendryx, M.S., Norell, D., Myers, M., Patterson, T.,McDonell, M.G., Voss, W.D. and McFarlane, W.R. (2000) Management of nega-tive symptoms among patients with schizophrenia attending multiple-familygroups. Psychiatric Services, 51: 513–519.

Glass, C.R and Arnkoff, D.B. (2000) Consumers’ perspectives on helpful andhindering factors in mental health treatment. Journal of Clinical Psychology, 56:1467–1480.

Gonzales, S., Steinglass, P. and Reiss, D. (1989) Putting the illness in its place:discussion groups for families with chronic medical illness. Family Process, 28:69–87.

Greenberg, L., Elliott, R. and Lietaer, G. (1994) Research on experientialpsychotherapies. In E.A. Bergin and S.L. Garfield (eds) Handbook ofPsychotherapy and Behavior Change. New York: John Wiley & Sons.

Igodt, P. (1983) La thérapie d’un groupe de familles et sa mise en pratique d’aprèsPeter Laqueur. Thérapie Familiale, 4: 81–97.

Kuipers, L., Leff, J. and Lam, D. (1992) Family Work for Schizophrenia: A PracticalGuide. London: Gaskell.

Laqueur, H.P. (1972) Mechanisms of change in multiple family therapy. In C.J.Sager and H.S. Kaplan (eds) Progress in Group and Family Therapy. New York:Brunner/Mazel.

58 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

Laqueur, H.P. (1976) Multiple family therapy. In P.J. Guerin (ed.) Family Therapy:Theory and Practice. New York: Gardner Press.

Leichter, E. and Schulman, G.L. (1974) Multi-family group therapy: a multidi-mensional approach. Family Process, 13: 95–110.

Lemmens, G., Heireman, M., Naeyaert, E. and Sabbe, B. (1999)Familiediscussiegroepen: een netwerk van verbindingen. Tijdschrift voorFamilietherapie, 6: 21–34.

Lemmens, G., Heireman, M. and Sabbe, B. (2001) Familiediscussiegroepen:methodiek en processen. Systeemtherapie, 13: 72–85.

McFarlane, W.R. (2002) Multifamily Groups in the Treatment of Severe PsychiatricDisorders. New York: Guilford Press.

McFarlane, W.R., Link, B., Dushay, R., Marchal, J. and Crilly, J. (1995)Psychoeducational multiple family groups: four-years relapse outcome in schiz-ophrenia. Family Process, 34: 127–144.

MacKenzie, K.R. (1987) Therapeutic factors in group therapy: a contemporaryview. Group, 11: 26–34.

Naeyaert, E., Sabbe, B., Heireman, M. and Lemmens, G. (1998) Multiple familytherapy en familiediscussiegroep: ontwikkeling en een voorbeeld van toepass-ing in dagbehandeling. Tijdschrift voor Psychotherapie, 24: 231–252.

O’Shea, M. and Phelps, R. (1985) Multiple family therapy: current status and crit-ical appraisal. Family Process, 24: 555–582.

Rolland, J.S. (1994) Families, Illness, and Disability; An Integrative Treatment Model.New York: Basic Books.

Rolland, J.S. (1998) Beliefs and collaboration in illness: evolution over time.Families, Systems & Health, 16: 7–25.

Scholz, M. and Asen, E. (2001) Multiple family therapy with eating disorderedadolescents: concepts and preliminary findings. European Eating Disorder Review,9: 33-42.

Slagerman, M. and Yager, J. (1989) Multiple family group treatment for eatingdisorders: a short term program. Psychiatric Medicine, 7: 269–283.

Steinglass, P. (1998) Multiple family discussion groups for patients with chronicmedical illness. Families, Systems and Health, 16: 55–70.

Strelnick, A.H. (1977) Multiple family group therapy: a review of the literature.Family Process, 16: 307–325.

Yalom, I.D. (1995) The Theory and Practice of Group Psychotherapy, 4th edition. NewYork: Basic Books.

Appendix 1

The session evaluation questionnaires

A The session evaluation questionnaire (family version)1 Have you experienced something in this session that you have

perceived as PARTICULARLY HELPFUL OR IMPORTANTFOR YOURSELF? This can relate to a short moment or alonger part of the session.

Beneficial factors in family discussion groups 59

2003 The Association for Family Therapy and Systemic Practice

O NO, ACTUALLY NOTO YESIf yes, describe it below

2 Have you experienced something in this session that you haveperceived as PARTICULARLY HELPFUL OR IMPORTANTFOR YOUR FAMILY? This can relate to a short moment or alonger part of the session.

O NO, ACTUALLY NOTO YESIf yes, describe it below

3 Have you experienced something in this session that you haveperceived as PARTICULARLY HELPFUL OR IMPORTANTFOR THE GROUP? This can relate to a short moment or alonger part of the session.

O NO, ACTUALLY NOTO YESIf yes, describe it below

B. The session evaluation questionnaire (therapeutic team version)1 Have you experienced something in this session that you have

perceived as PARTICULARLY HELPFUL OR IMPORTANTFOR A GROUP MEMBER? This can relate to a short momentor a longer part of the session.

O NO, ACTUALLY NOTO YESIf yes, describe it below

2 Have you experienced something in this session that you haveperceived as PARTICULARLY HELPFUL OR IMPORTANTFOR THE FAMILIES? This can relate to a short moment or alonger part of the session.

O NO, ACTUALLY NOTO YESIf yes, describe it below

3 Have you experienced something in this session that you haveperceived as PARTICULARLY HELPFUL OR IMPORTANTFOR THE GROUP? This can relate to a short moment or alonger part of the session.

O NO, ACTUALLY NOTO YESIf yes, describe it below

60 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

Appendix 2

A brief description of the category system

A Relational climate and structural aspects of the group1 Dedication, commitment and involvement: motivation of

group members to work on their problems, interest andattention for the group members, the active participation inthe group, the involvement with other group members.(a) Within the group: ‘the active participation of all group

members’(b) Within the family: ‘the commitment of the family members for

the patients’.2 Space and freedom: an atmosphere of openness and toler-

ance, there’s room for all types of personal contributions andreactions, but also respect for one’s own limits and intimacy:‘there was an open discussion’.

3 Empathy and feeling understood: one listens empathically,the group member feels understood on a deeper level inwhat he/she tries to contribute: the understanding of my spousefor me ‘being ill’.

4 Confirmation, appreciation and support: getting apprecia-tion as a person, feeling supported when having a difficulttime, regard for certain interventions made in the group.(a) From the group: ‘the support for my husband from a family

member of a patient’(b) From the therapists: ‘the support of the therapist for a

patient’’(c) From the family: ‘the support within the families’ .

5 Authenticity and transparency: being more transparent thanin everyday life, talking in a personal way, acting the way theyreally are.(a) Within the group: ‘you could be yourself, and yet share things’(b) Within the family: ‘that I can be myself in the relationship

with my parents’.6 Group cohesion and feeling good: a feeling of belonging

together and of being accepted in the group, an atmosphereof confidence and safety: ‘I feel good in the group’.

7 Informal, beneficial contact outside the sessions: ‘the contactduring the coffee break with the other family members’.

8 Structural aspects and facilitating group composition: events

Beneficial factors in family discussion groups 61

2003 The Association for Family Therapy and Systemic Practice

related to the group composition such as presence of all ages,and the structure of the session, including respecting timelimits: ‘I found it important that my wife and children were present’.

B Specific interventions by group members or therapist9 Offering therapeutic techniques: techniques that lead the

members to reach their own feelings, such as homeworkassignment: ‘the pleasant homework task at the end of the session’.

10 Stimulating, clarifying and interpretive interventions: inter-vention to tell more about oneself, to go deeper into a certainissue, saying something from a different frame of reference,making connections, offering a broader perspective or analternate viewpoint: the mother of a patient was frequently asked bythe therapists to explain about the relationship with her daughter.

11 Reaction and feedback: answering to what a group member saysor asks, giving an impression about a group member, or reveal-ing thoughts or feelings which a member evokes: ‘some groupmembers indicated to the parents of a patient to give her more freedom’.

12 Interrupting a disturbing process: ‘that the therapist started totalk about another subject when the group got stuck’.

C Process aspects in the group member13 Self-revelation: bringing personal matters to the group, being

able to or daring to talk about something: ‘to talk about howdifficult it was the last days’.

14 Self-exploration and interpersonal exploration.(a) Self-exploration: approaching a personal problem and

exploring it, expressing feelings: ‘my crying is not alwaysrelated to my pain’

(b) Discussing mutual relationships with family members:relations are discussed and explored, tensions are talkedover, and misunderstandings straightened out: ‘thepatient made it clear to her parents that she didn’t want tocommunicate with her father’

(c) Expressing feelings towards family members: expressingpositive and negative feelings towards family members:‘expressing her feelings of anger towards her sister and parents’

(d) Interpersonal exploration with other group members:discussing mutual relations with other group members,expressing feelings towards other group members: ‘show-ing emotions as crying and laughing in the group’.

15 Discussing different topics.(a) Existential and family-related: ‘talking about love’

62 Gilbert M. Lemmens et al.

2003 The Association for Family Therapy and Systemic Practice

(b) Illness-related: ‘to talk about depression’(c) Other: ‘the subject of communication’.

16 Experiencing of communality: recognizing oneself in thingsthat others mention, not being alone with certain thoughtsand questions.(a) With other group members: ‘we all have similar problems’(b) With other families: ‘the problems, which we have with the

family members, are similar to the other patients and familymembers’.

17 Learning by observation: seeing or hearing how others tackleproblems: ‘you learn from the other persons how they have reacted tosomething’.

18 Becoming conscious of/having insight into: something in thegroup leads to reflection, becoming conscious of somethingkept hidden, putting into perspective one’s own opinion,discovering new or half-conscious things about oneself.(a) Oneself: ‘you have to decide which persons are important in

your life, I spend too much time with “unimportant” people’(b) Family: ‘my brother is always there for me and I am always there

for him’(c) Illness: ‘the social activities of a spouse of a pain patient are

also limited by the pain problem’(d) Other: ‘not talking is also communication’.

19 Experiencing the helpful potential of the group/experi-menting with new behaviour/capacity to help others: thegroup members are capable of helping each other, having anenriching experience, trying out new behaviours which thegroup member finds difficult: ‘the session was important andhelpful’.

20 Experiencing hope and progress: finding comfort from theexperience of having already advanced quite a bit, looking atthe future with more confidence: ‘to hear and see from othersthat things get better after a while and despite the current difficulties’.

21 Experiencing relief, relaxation and humour: ‘we had a goodlaugh’.

D Remaining categories22 Blank responses.23 No helpful events reported.24 Yes, but unclassifiable or unclear answers.25 Yes, with no explanation.

2003 The Association for Family Therapy and Systemic Practice

Beneficial factors in family discussion groups 63