6.1. Introduction
The psychological treatment of depression in children and young people is dissimilar from that provided for depressed adults. Although, as with physical treatments for depression, there has been some extrapolation from approaches used for adults (for example cognitive behavioural therapy [CBT] and psychodynamic therapy), in routine practice these formal individual therapies are not the most common psychological approaches. Whereas adults with depression are often treated for the disorder specifically, children with depression are often not thought of as ‘having’ depression but as affected by a set of emotional, behavioural, learning, relationship and family problems which need to be considered together, and may still need to be addressed together, even if depression in the child is a primary concern.
Thus, psychological therapies for depression in children and young people may not be thought about as distinct from working with children, adolescents and families with a wide range of psychosocial difficulties. This is probably especially true with pre-adolescent children, but for adolescents as well depression is very likely to be seen as a sign of a more complicated situation, or result of earlier stresses within a system (the family or school for example).
Nevertheless, as adolescents move towards adult independence (or fail to do so), it is more likely that therapies designed for adults will be thought to be appropriate and extended to these young people. It may be for this reason, as well as because depression in children and young people is most prevalent in adolescence, that the studies reviewed here have generally been conducted with young people of secondary school age, and our conclusions should not be assumed to apply to younger children, without further investigation.
Children rarely initiate mental health assessment and treatment although adolescents may seek help for emotional difficulties in a wide variety of ways (see Chapter 5). While many clinically depressed adults recognise that they are depressed (or at least that they are ill) and seek treatment, both would be rare in children and adolescents. At a mild level, children are highly unlikely to be referred, unless it affects their behaviour in some obvious way (e.g. self-harm, withdrawn or aggressive behaviour at school, or failing academic performance). Depression in children and young people commonly presents as recurrent and unexplained physical symptoms, which may be difficult to recognise as depression, even for the healthcare professionals consulted (e.g. a GP, school nurse, paediatrician). Even when they recognise the underlying problem, parents, child, teachers and others involved may well find it difficult to accept the need for psychological or psychiatric treatment, not uncommonly because of feelings of anxiety, anger and shame, or indeed because of stigma and lack of knowledge about mental health problems generally.
It is important, therefore, to note that the recognition of depression, and the likelihood of children or young people receiving effective treatment and care, is mediated through the differing perceptions and reactions of parents, health and non-health professionals already involved in the child's life, the children/young people themselves and specialist mental healthcare professionals. It is perhaps no surprise that in this context, many children and young people who are depressed have a tendency to think that they are the problem, rather than thinking that they have a problem with which they may be able to get help.
6.2. Psychological therapies
6.2.1. Introduction
Psychological therapies for depression in children and young people include a number of approaches, involving different activities, people and amounts of time, and different theoretical assumptions. Treatment may, for example, involve talking with the child, and perhaps others in the family, to clarify the reasons for the child's unhappiness, withdrawal and other symptoms, with the aim of recognising possible factors (e.g. bereavement, parental mental ill health, bullying) which may be linked to the child's depression. Alternatively it may be focused on identifying roles and communication patterns within the family, on changing the child's depressed behaviour (e.g. staying in bed all day, dropping out of school, self-neglect, social isolation), or on enabling the child to find new ways of expression and communication such as through art therapy.
6.2.2. Current practice
The current system of NHS CAMHS provision is described in Chapter 8. The wider context of non-NHS services, together with self-help resources, is described in Chapter 5.
Within the NHS, depression in a young person may first be noticed by one of a range of primary care and community professionals, including GPs, practice nurses, counsellors, school nurses, community paediatricians. Sometimes these professionals may offer treatment for the depression, particularly if it is not severe and/or it is in the context of long-term physical illness. Some depressed children and young people will be referred to CAMHS. Nevertheless, it is estimated that the majority of children and young people with depression will not be recognised as such and will not, therefore, receive any specific help (Andrews et al., 2002; Coyle et al., 2003).
When a child/young person with depression is referred to CAMHS, whether the first-line treatment offered will be physical or psychological varies considerably. Where the initial approach is social/psychological, this is likely to begin as a generic approach, involving clarification of the problem with the family and child, and locating the child's depression within a wider psychosocial context. Only a small proportion of children in most services will be referred for a specific psychological intervention, such as CBT, individual child psychotherapy or family therapy. Most cases will thus be treated with a psychological approach, involving elaboration and formulation of the problem, which has not been systematically evaluated for treatment efficacy.
A very small proportion of depressed young people may be so severely self-harming or incapacitated that they will be admitted to inpatient adolescent units. Here, it is more likely that a young person would receive a formal psychological therapy, such as those evaluated in outcome studies. In addition, they would be more likely to receive a form of group therapy, in addition to care within a milieu which provides intensive adult supervision and monitoring of physical behaviour and safety.
6.2.3. The evidence base for psychological therapies
There is considerable variation in the evidence base for different specific psychological therapies for depression in children and young people. This variation is a result of a number of factors, including: the cost, ethics and complexity of undertaking randomised controlled trials of psychological therapies, especially those involving even moderately long therapies, or where the goal of change is more than symptomatic improvement; the willingness of therapists to participate in a research study; the level of skill and experience needed for some psychological therapies; and the paucity of funding available for psychological therapy research, especially when compared with the large amount of funding provided by the pharmaceutical industry for drug treatments. Moreover, the funding that is available for psychological therapy research for depressed children and young people has generally been lower than funding for research into externalising (behavioural) problems. This lack of evidence should not, therefore, be taken as evidence of ineffectiveness; there is clearly a need for more research to create an adequate evidence base.
The evidence base for the psychological therapies for depressed children and young people, such as it is, is largest for CBT (especially group), with rather less evidence concerning family therapy, and less still for or against individual child psychotherapy. It is important to recognise that the level of outcome research does not reflect the prevalence of different psychological therapy approaches in current practice within CAMHS in the NHS (for example, group CBT for depression is rarely practised, while short-term family work is extremely common).
6.2.4. Research limitations
In the early years of this research, it was rare for referred children to be studied; such outcome research as there was, was confined to community samples of young people with sub-clinical levels of depressed mood, or clinically depressed young people who were recruited through advertising or screening of large non-referred samples. There is evidence that children who are referred to CAMHS show more complex and entrenched sets of problems, not simply comorbid psychiatric disorders but what may be chronic problems in their social and academic functioning, and psychiatric and social problems in their parents and families (e.g. Hammen, et al., 1999).
There has, however, been a gradual move in recent years to recruit ‘real life’ clinical samples, and to include children and young people with comorbid diagnoses in the studies. These changes introduce new practical problems, such as the need for large sample sizes (e.g. to examine the impact of comorbid diagnoses), as well as difficulties of interpretation. Nevertheless, there is an obvious necessity to increase the external validity (generalisability) of studies' findings.
The current review includes studies of both referred and recruited samples. We also include studies with samples defined by depression symptom checklists as opposed to clinical diagnosis. Both factors need to be borne in mind in interpreting the research findings for application to NHS patients presenting with diagnosable levels of depression.
Finally, a significant limitation within the parameters of this guideline is that some important studies have not selected their sample, or reported on outcomes, in terms of depression, but have looked at the effectiveness of a treatment approach across the range of disorders and comorbid conditions that present to the service (Fonagy et al., 2002). For example, a psychotherapy service that treats many depressed young people and collects outcome data on internalising symptomatology cannot be included because the data are not depression-specific (Baruch et al., 1998; Baruch et al., 1999). Evaluations of services currently need to be carried out in ways that make reporting possible in terms of diagnoses or disorder-specific symptom scales, if their outcomes are to be included in diagnosis-based guidelines. Alternatively – and probably more appropriately, given the way that services are provided to children and young people guidelines could address treatment outcomes across groups of related disorders, in this case across internalising disorders (the range of anxiety and depressive disorders and their combinations) as opposed to depression in isolation. However, that would require a shift in research culture, away from the medical model of seeking the most effective treatment for a DSM illness category, towards the biopsychosocial model in which formulation is expected to be complex, and treatment assumed to be broad-based, to fit the multiple causative and mediating factors which impinge on children's emotional and social development and current functioning.
Future research needs more closely to address the needs of NHS healthcare professionals for guidance on treatment choice. It is positive that the culture of research on psychological therapy for children and young people has moved somewhat closer to clinical reality, in its focus on multimodal/multisystemic therapies and developing therapies that can be successfully applied outside the university clinic. The next step may be to make the research more relevant to CAMHS professionals who do not tend to think of the child in isolation from his or her social context, or use a diagnostic category as the basis for treatment choice.
6.2.5. Databases searched and inclusion criteria
Table 1Databases searched and inclusion criteria for clinical effectiveness of psychological interventions
View in own window
| Electronic databases | MEDLINE, EMBASE, PsycINFO, Cochrane Library |
| Date searched | Databases: inception to January 2004 (key journals searched using the electronic table of contents service February to September 2004) |
| Study design | RCT |
| Patient population | Participants aged 5–18 years old with recognised symptoms of depression |
| Interventions* |
–
Cognitive behavioural therapy (CBT) –CBT + separate parent sessions –Interpersonal psychotherapy (IPT) –Psychoanalytic/psychodynamic child psychotherapy –Self-modelling –Relaxation –Social skills training –Family therapy –Guided self-help –Control enhancement training –Control group (waitlist, non-directive supportive therapy, therapeutic support group, ‘standard care’, clinical management, behavioural problem-solving, life skills training) |
| Outcomes | Remission, symptom levels, functional status, discontinuation from treatment for any reason |
- *
These interventions can be grouped into a much smaller number of major approaches, with considerable overlap between the different ‘brands’ of, for example cognitive-behavioural approach. We regard the major approaches in current practice as: individual CBT; group CBT; structural/behavioural family therapy; systemic family therapy; psychoanalytic/psychodynamic child psychotherapy; other non-directive therapy which is primarily supportive. Within this framework both social skills training and IPT would be regarded as ‘brands’ of the CBT family.
6.2.6. Studies considered7
The review team conducted a new systematic search for RCTs that assessed the efficacy of psychological therapies for children and young people with depression.
Eighteen trials met the eligibility criteria set by the GDG: 14 from the US, three from the UK/Europe, and one from Puerto Rico. In total, data on 1520 participants were used. The trials were published between 1986 and 2004, and were between 4 and 36 weeks long. In addition, two studies, one comparing social skills training with non-directive supportive therapy (REED1994), and one comparing a combined cognitive behavioural family education intervention with waitlist (ASARNOW2002) were excluded from the analysis due to a lack of usable data. A further three studies were excluded because there was no or an inappropriate control group (MUFSON1994, NELSON2003, SANTOR2001). Further information about both included and excluded studies can be found in Appendix R on CD-ROM.
Active intervention versus another active intervention
There were three trials involving a comparison of CBT with non-directive supportive therapy or clinical management (BRENT1997, TADS2004, VOSTANIS1996), one of CBT versus IPT (ROSSELLO1999), one of CBT versus relaxation (WOOD1996), and one of CBT versus family therapy (BRENT1997) ().
Study information for trials of CBT versus another psychological intervention.
In addition, there were two trials of group CBT versus behavioural problem-solving/life skills training (ROHDE2004, STARK1987), two trials involving a comparison of group CBT versus group relaxation (KAHN1990, REYNOLDS1986), one trial involving a comparison of group CBT versus self-modelling (KAHN1990), one trial involving a comparison of group relaxation versus self-modelling (KAHN1990), and one trial of family therapy versus individual psychodynamic psychotherapy (TROWELL) ().
Study information for trials of group CBT/relaxation or family therapy versus another psychological intervention.
There was one trial of IPT versus ‘standard care’ (MUFSON2004), two trials that examined the impact of adding separate parent sessions to group CBT (CLARKE1999, LEWINSOHN1990) and one trial of family therapy versus non-directive supportive therapy (BRENT1997) ().
Study information for trials of group CBT, group CBT plus separate parent sessions, family therapy, and IPT.
6.2.7. Psychological interventions versus waitlist/control group
Evidence summary table for various psychological interventions versus waitlist/control group.
Evidence summary table for various psychological interventions versus waitlist control.
6.2.8. Psychological interventions versus other psychological interventions/control
Evidence summary table for CBT versus other psychological interventions/control intervention.
Evidence summary table for CBT versus other psychological interventions/comparator intervention.
Evidence summary table for various psychological interventions versus other psychological interventions/‘standard care’/comparator intervention.
Evidence summary table for group CBT (+ parent) and guided self-help versus waitlist/group CBT.
6.2.9. Clinical summary
For individual outcomes, the quality of the evidence was generally moderate to low, reflecting the paucity of data and relatively small sample sizes of those studies included in the review.
6.2.9.1. Psychological therapies in general
The evidence regarding the effectiveness of psychological therapies shows that a number of therapies are effective at treatment endpoint (see below for details), but no psychological therapies have been shown to maintain a significant superiority to non-active control treatments at 1-year (or more) follow-up. The overall conclusion seems to be that while a range of therapies produce gain during treatment which is reasonably well maintained at follow-up, where a minimal treatment comparison group is included, this group tends to catch up over the following several months. An accelerated resolution of depression (by say 6 to 12 months compared with a control group) is a very important achievement for the emotional, social and cognitive life of a child or adolescent. Thus, finding that minimally treated children catch up over time does not mean at all that treatment was not effective. Nevertheless, a significant proportion of children and young people do remain depressed at the end of treatment, or are highly at risk of later relapse, even where group results are encouraging. There is some evidence that treatments that have specifically planned booster or follow-up sessions may be effective in maintaining treatment gains, but there clearly needs to be continuing research on the treatment of ‘resistant’ depression. These findings also argue for maintaining a range of treatments to help those who do not respond to first and even second-line treatments. Thus, for example, an unpublished study (TROWELL) found that a very high proportion of moderately to severely depressed young people offered one of two relatively intensive and long-term treatments (family therapy or individual child psychotherapy) improved and stayed well. This study obviously needs to be replicated to establish for which children or young people longer-term treatment may be needed.
6.2.9.2. Individual therapies
Individual CBT
The overall evidence for the effectiveness of individual CBT is inconclusive. Two studies have failed to show that CBT is more effective than waitlist (ROSSELLO1999, 12 weekly sessions of 60 minutes duration) or general clinical management (TADS2004, 15 sessions of 50 to 60 minutes duration for 12 weeks). However, three studies of clinic-referred samples have indicated clinically important improvement compared with comparison therapies, namely relaxation (WOOD1996, 5 to 8 weekly sessions), non-directive supportive therapy (VOSTANIS1996, 9 sessions; BRENT1997, 12 to 16 weekly sessions of 60 minutes duration) and systemic family behavioural therapy (BRENT1997). These studies indicate that CBT is likely to reduce the length of the depressive episode compared with these therapies. These differential effects were not sustained at longer-term follow-up although this was mainly due to ongoing improvements of comparison therapies, rather than relapse in those receiving CBT.
Interpersonal therapy (IPT)
There is limited evidence from three studies (MUFSON1999, MUFSON2004, ROSSELLO1999) indicating the efficacy of IPT (12 sessions of 35 to 60 minutes duration for 12 to 16 weeks) compared with waitlist or ‘standard care’ in increasing the chance of remission and reducing depressive symptoms. There is also limited evidence from one study (MUFSON2004) that IPT (12 sessions of 35 minutes duration for 16 weeks) improves overall functioning when compared with ‘standard care’. In direct comparison with individual CBT the evidence was inconclusive.
Individual psychodynamic psychotherapy
There are no published studies of the effectiveness of psychoanalytic psychotherapy with an untreated or placebo control group. The results of a multi-centre study comparing psychoanalytic psychotherapy (30 sessions of 50 minutes duration for 36 weeks) and family therapy (14 sessions of 90 minutes duration for 36 weeks) have been prepared for publication. The study recruited moderate to severe cases, many of whom had concurrent diagnoses of dysthymia. Preliminary results indicate high rates of remission and excellent maintenance of gains at follow-up, in both treatment arms, with limited evidence that family therapy reduced depressive symptoms more rapidly, but individual therapy may have a more sustained effect (TROWELL).
6.2.9.3. Group therapies
Group CBT
There is considerable evidence from a number of studies to suggest that group CBT (8 to 16 sessions of 40 to 60 minutes duration for 5 to 8 weeks) is an effective treatment of adolescents for increasing the chance of remission and reducing depressive symptoms compared with waitlist conditions/no treatment/‘standard care’ (CLARKE1999, CLARKE2002, KAHN1990, LEWINSOHN1990, REYNOLDS1986, STARK1987, WEISZ1997). However, the majority of evidence for this is from recruited samples from the USA and the effects of therapy were not maintained at longer follow-up, although this was mainly due to ongoing improvements of comparison treatments, rather than relapse in those receiving group CBT. Group CBT has been directly compared with other therapies such as group relaxation, problem-solving and self-modelling, with either inconclusive or limited evidence favouring group CBT. It is not possible to determine how it compares with other more frequently used therapies. The evidence is inconclusive as to whether this would be an effective therapy for clinic-referred young people with clinical depression.
6.2.9.4. Therapies involving the parents
Family therapy
There is limited evidence from one study (DIAMOND2002) about the efficacy of family therapy (12 to 15 sessions of 60 to 90 minutes duration for 6 weeks) compared with a waitlist condition. A second study (BRENT1997) comparing family therapy (12 to 16 sessions of 60 minutes duration for 12 to 16 weeks) with non-directive supportive therapy was inconclusive, and as described above, suggested poorer outcomes relative to individual CBT. A recent multi-centre study of clinical cases (TROWELL), in which family therapy (14 sessions of 90 minutes duration for 36 weeks) was compared with individual psychodynamic psychotherapy (30 sessions of 50 minutes duration for 36 weeks), but not with an untreated or inactive comparison condition, showed high rates of remission and symptom reduction, with good maintenance of gains.
Parent involvement in psychological therapies
There was inconclusive evidence from two studies (CLARKE1999, LEWINSOHN1990) to determine whether the additional involvement of parents in a group CBT therapy (14 to 16 sessions of 120 minutes duration for 7 to 8 weeks) increased effectiveness.
6.2.9.5. Guided self-help
There is evidence from one small trial (ACKERSON1998) suggesting that guided self-help (for 4 weeks) may improve depressive symptoms when compared with waitlist.
6.3. Association between primary outcomes and characteristics of therapist/patient
6.3.1. Introduction
There is little evidence relating to the association between outcome and therapist characteristics.
In contrast, there were many aspects of service user characteristics identified in a variety of studies, which correlated with outcome. Many of these are well known in clinical practice, particularly that comorbidity makes it less likely that therapy will achieve a good outcome. Around 70% of patients treated for depression were found to have comorbid disorders particularly anxiety (Emslie et al., 2003). We found the following statement highly relevant as regards clinical practice:
‘The search for a pure i.e. non comorbid form of very early onset affective illness may be a futile undertaking, as comorbidity may be an intrinsic characteristic of children with affective disorders’ (Emslie et al., 2003, p. 445).
Many authors also stressed the importance of assessing the network around the child because many factors, for example, parental mental ill health (including both affective and non-affective disorders), socio-economic disadvantage and family/parental dysfunction (particularly the impact of divorce and bereavement) (Beardslee, 1993) were negatively correlated with outcome and are also important in therapy selection.
6.3.2. Descriptive review
6.3.2.1. Therapist characteristics
The only study identified was Wiesz (1995), which showed better outcome in treating depression if qualified professional therapists were used rather than non-professional workers.
The importance of a better treatment alliance with the patient was also mentioned (Diamond et al., 2002), but the evidence was from a case record study rather than an RCT.
6.3.2.2. Service user characteristics
Comorbidity was the most important factor that negatively correlates with therapy outcome and affects the chances of relapse (Emslie et al., 2003). There was some evidence that depression which is comorbid with anxiety may be helped by CBT as this has been shown to be effective in reducing anxiety (Brent et al., 1998).
Severity of the depression, especially higher levels of chronicity, suicidality and hopelessness, as well as higher levels of cognitive distortion, were all negatively correlated with outcome (Emslie et al., 2003; Brent et al., 1998). This may contribute to the difference in outcomes between clinical and advertised or recruited samples (Brent et al., 1998).
Poor parenting, negative interactions and higher family dysfunction/stress were also correlated with negative outcome (Emslie et al., 2003; Brent et al., 1998). Children with parents who have an affective disorder experience a rate of major depressive disorder 2.6 times greater than those with parents who have no disorder. The disorders of the children whose parents have also been affected also on average start earlier and last longer. There were also multiple risk factors involved since non-affective disorder was present in the majority of parents and often there was psychiatric disorder in both parents. Divorce or separation also had occurred in a substantial number of families. In fact the main effects of parental affective disorder was significant only when it was in combination with divorce (Beardslee, 1993).
The presence of abuse in all forms as well as trauma was shown in a number of studies to be correlated with higher rates of depression and more difficulty in treating it (Becker et al., 1991; Bergen et al., 2003; Meyerson et al., 2002; Sadowski, 2002; Ramchandani & Jones, 2003).
One would expect the motivation of the patient to change to be correlated with treatment outcome but there was no directly reported evidence of this, other than the frequent report that hopelessness in the patient was negatively correlated with outcome. Similarly the effect of parental depression has been highlighted and this too may be mediated through hopelessness about any treatment proposed for the child. This may be directly communicated to the child or enacted through poor treatment adherence (Brent et al., 1998; Emslie et al., 2003).
Parental involvement, treatment attendance, avoiding premature termination and matching parental/patient expectancies to treatment predicted positive treatment outcome. The presence of social support was also shown to be important especially for girls (Ramchandani & Jones, 2003; Emslie et al., 2003; Schraedley et al., 1999).
6.3.3. Clinical summary
Although little is known about therapist factors that influence outcome, there is some evidence that professionally trained therapists have better results than paraprofessionals with this group. As there is some evidence that a positive treatment alliance predicts better outcome, therapists who are better able to create this alliance with depressed young people are likely to be more successful.
Several characteristics of service users and their carers have been found to relate to psychological therapy outcome. Comorbid conditions and more severe or complex symptomatology are associated with less good outcomes. Parental depression/mental ill health, the impact of divorce, separation and bereavement are especially important family factors; feelings of hopelessness and family dysfunction can impact on the child in many different ways. Clinical populations generally present with comorbid conditions and more complex sets of problems within the individual, the family and the network; multi-modal treatments in sequence or parallel are therefore likely to be required.
6.4. Relapse prevention
6.4.1. Introduction
As described in Chapter 3 of this guideline, around 30% of cases recur within 5 years, many within a year of the earlier episode, and some of these young people develop episodes into adult life. Furthermore, as shown in our systematic review, a proportion of cases in all treatment trials remains diagnosable at the end of treatment, or remain symptomatic at a level below the threshold for diagnosis. A very important question for the care of children and young people with depression is thus whether there are ways to reduce the likelihood of either a relapse of depression following remission, or a long-term state of unhappiness and poor functioning following partial improvement during treatment. Clinically, it is likely that attention needs to be paid to social factors that may maintain a depressed state, or cause further episodes. Such factors would be likely to include relationship difficulties in the family or peer group, including for adolescents difficulty in establishing sexual relationships and achieving greater independence from parents. Difficulties arising from cultural and ethnic differences may be important, as may physical illness or any kind of disability, persistent comorbid disorders, and concern about family members (for example, parental psychiatrist illness).
A systematic search of the literature identified no RCTs concerning the prevention of relapse of depression in children and/or young people that met the eligibility criteria set by the GDG. None of the other reports identified presented compelling evidence.
6.4.2. Databases searched and inclusion criteria
Table 12Databases searched and inclusion criteria for studies of relapse prevention
View in own window
| Electronic databases | MEDLINE, EMBASE, PsycINFO, Cochrane Library |
| Date searched | Databases: inception to February 2004 (key journals searched using the electronic table of contents service February to September 2004) |
| Study design | Controlled trials |
| Patient population | Participants aged 5–18 years with recognised symptoms of depression |
| Interventions included | – Cognitive behavioural therapy (CBT)
– Assessment only |
| Outcomes | Relapse |
6.4.3. Studies considered12
The review team conducted a new systematic search for controlled trials that assessed the efficacy of psychological therapies for children and adolescents with depression for the prevention of relapse.
Table 13Study information table for trials of CBT booster/continuation treatment versus assessment only/no treatment
View in own window
|
Total no. of trials (total no. of participants)
|
1 controlled trial (N = 41)
|
1 trial with historical control (N = 29)
|
| Study ID |
CLARKE1999
| KROLL1996 |
| Diagnosis | MDD or dysthymia
(DSM-III-R) | MDD (DSM-III-R) |
| Length of follow-up | 12 & 24 months | 6 months |
| Age | 14–18 years | 10–17 years |
Two trials met the eligibility criteria: one controlled trial (CLARKE1999) comparing group CBT booster sessions (1-2 meetings) with assessment only using a 24-month follow-up, and one study (KROLL1996) comparing continuation with CBT (after acute phase treatment) with a historical control group using a 6-month follow-up.
6.4.4. Continuation/booster treatment
CLARKE1999 randomly assigned participants who had completed an acute phase treatment of group CBT to one of three 2-year follow-up conditions: (1) booster sessions (one to two meetings) and independent assessments every 4 months; (2) assessment only every 4 months; or (3) assessment only every 12 months. For the purpose of this review, relapse (defined as meeting criteria for unipolar depression) was analysed at 12 and 24 months in those participants who had recovered by the end of the acute phase treatment.
At the end of 12 and 24 months, the evidence was inconclusive regarding the risk of relapse, although there is only a small probability that group CBT booster sessions prevented relapse.
KROLL1996 compared the risk of relapse in participants who continued to receive group CBT for 6 months (after a course of five to eight sessions of CBT during the acute episode) with a historical control group drawn from a previous study of CBT (WOOD1995). All participants had remitted from MDD by the end of the acute phase.
By the end of 6-month follow-up, there was limited evidence that continuation of group CBT may reduce the risk of relapse (RR = 0.35; 95% CI, 0.11 to 1.14).
6.4.5. Clinical summary
We found no evidence to properly assess whether psychological therapies can prevent relapse in children and/or young people with depression. The evidence from non-randomised studies suggests that continuation of group CBT, but not booster sessions, may reduce the risk of relapse. Nevertheless, until further research is conducted using adequately designed relapse prevention studies, no conclusion can be reached.
