Home > Clinical Guides > Depression: The Treatment and Management... > HIGH-INTENSITY PSYCHOLOGICAL INTERVENTIONS
  • We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Mental Health (UK). Depression: The Treatment and Management of Depression in Adults (Updated Edition). Leicester (UK): British Psychological Society; 2010. (NICE Clinical Guidelines, No. 90.)

8HIGH-INTENSITY PSYCHOLOGICAL INTERVENTIONS

This section covers the high-intensity interventions that were identified in the searches for the guideline update and groups them according to the definitions developed for the previous guideline (NICE, 2004a). Although to some degree cognitive behavioural therapies, behavioural activation, problem solving therapy and couples therapy54 share a common theoretical base, they are reviewed separately.

8.1. COGNITIVE BEHAVIOURAL THERAPIES

8.1.1. Introduction

Cognitive behavioural therapy (CBT) for depression was developed by Aaron T. Beck during the 1950s and was formalised into a treatment in the late 1970s (Beck et al., 1979). Its original focus was on the styles of conscious thinking and reasoning of depressed people, which Beck posited was the result of the operation of underlying cognitive schemas or beliefs. The cognitive model describes how, when depressed, people focus on negative views of themselves, the world and the future. The therapy takes an educative approach where, through collaboration, the person with depression learns to recognise his or her negative thinking patterns and to re-evaluate his or her thinking. This approach also requires people to practise re-evaluating their thoughts and new behaviours (called homework). The approach does not focus on unconscious conflicts, transference or offer interpretation as in psychodynamic psychotherapy. As with any psychological treatment, cognitive behavioural therapy is not static and has been evolving and changing. There have been important elaborations on the techniques of therapy (Beck, 1997) to address underlying beliefs more directly, which have been applied to particular presentations such as persistent residual depressive symptoms that leave people vulnerable to relapse (Paykel et al., 1999; Scott et al., 2000; Moore & Garland, 2003; Watkins et al., 2007). The guideline refers to ‘cognitive behavioural therapies’ to indicate the evolution of CBT for depression over several decades.

Definition

For the purpose of this review, cognitive behavioural therapies were defined as discrete, time-limited, structured psychological interventions, derived from the cognitive behavioural model of affective disorders and where the patient:

  • works collaboratively with the therapist to identify the types and effects of thoughts, beliefs and interpretations on current symptoms, feelings states and/or problem areas
  • develops skills to identify, monitor and then counteract problematic thoughts, beliefs and interpretations related to the target symptoms/problems
  • learns a repertoire of coping skills appropriate to the target thoughts, beliefs and/or problem areas.

In most individual trials of CBT, the manual used was Beck’s Cognitive Therapy of Depression (1979) which advocates 16 to 20 sessions for treatment and relapse prevention work.

Group cognitive behavioural therapy

Trials looking at group CBT, which predominantly uses the ‘Coping With Depression’ approach (Kuehner, 2005; Lewinsohn et al., 1989), were also included. This approach has a strong psychoeducational component focused on teaching people techniques and strategies to cope with the problems that are assumed to be related to their depression. These strategies include improving social skills, addressing negative thinking, increasing pleasant activities, and relaxation training. It consists of 12 sessions of 2 hours’ duration over 8 weeks with groups held twice weekly for the first 4 weeks. The groups are highly structured (Lewinsohn et al., 1984; Lewinsohn et al., 1986) and typically consist of six to ten adults, with two group leaders. One- and 6-month follow-up sessions are also held and booster sessions can be used to help prevent relapse.

Mindfulness-based cognitive therapy

Mindfulness-based cognitive therapy (MBCT) was developed with a specific focus on preventing relapse/recurrence of depression (Segal et al., 2002). MBCT is an 8-week group programme with each session lasting 2 hours, and four follow-up sessions in the year after the end of therapy. With 8 to 15 patients per group, MBCT has the potential to help a large number of people.

MBCT is a manualised, group-based skills training programme designed to enable patients to learn skills that prevent the recurrence of depression (Segal et al., 2002). It is derived from mindfulness-based stress reduction, a programme with proven efficacy in ameliorating distress in people with chronic disease (Baer, 2003; Kabat-Zinn, 1990), and CBT for acute depression (Beck et al., 1979), which has demonstrated efficacy in preventing depressive relapse/recurrence (Hollon et al., 2005). MBCT is intended to enable people to learn to become more aware of the bodily sensations, thoughts and feelings associated with depressive relapse, and to relate constructively to these experiences. It is based on theoretical and empirical work demonstrating that depressive relapse is associated with the reinstatement of automatic modes of thinking, feeling and behaving that are counter-productive in contributing to and maintaining depressive relapse and recurrence (for example, self-critical thinking and avoidance) (Lau et al., 2004). Participants learn to recognise these ‘automatic pilot’ modes, step out of them and respond in healthier ways by intentionally moving into a mode in which they ‘de-centre’ from negative thoughts and feelings (for example, by learning that ‘thoughts are not facts’), accept difficulties using a stance of self-compassion and use bodily awareness to ground and transform experience. In the latter stages of the course, patients develop an ‘action plan’ that sets out strategies for responding when they become aware of early warning signs of relapse/recurrence (Williams, J.M., et al., 2008).

8.1.2. Studies considered55

In total, 68 studies were identified, of which 46 RCTs were included; 24 studies were found in the update search and 22 were also reported in the previous guideline. Furthermore, 22 trials were excluded in this update search. The main reasons for exclusion were: trials included populations that were not diagnosed with depression; authors replaced dropouts; or more than 50% of participants dropped out of the study.

Summary study characteristics of the included studies are presented in Table 36, Table 37 and Table 38, with full details in Appendix 17b, which also includes details of excluded studies.

Table 36. Summary study characteristics of cognitive behavioural therapies.

Table 36

Summary study characteristics of cognitive behavioural therapies.

Table 37. Summary study characteristics of cognitive behavioural therapy.

Table 37

Summary study characteristics of cognitive behavioural therapy.

Table 38. Summary study characteristics of cognitive behavioural therapies.

Table 38

Summary study characteristics of cognitive behavioural therapies.

8.1.3. Clinical evidence for cognitive behavioural therapies

Evidence from the important outcomes and overall quality of evidence are presented in Table 39, Table 40 and Table 41. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

Table 39. Summary evidence profile for cognitive behavioural therapies.

Table 39

Summary evidence profile for cognitive behavioural therapies.

Table 40. Summary evidence profile for cognitive behavioural therapies.

Table 40

Summary evidence profile for cognitive behavioural therapies.

Table 41. Summary evidence profile for cognitive behavioural therapies.

Table 41

Summary evidence profile for cognitive behavioural therapies.

8.1.4. Clinical evidence summary for cognitive behavioural therapies

Cognitive behavioural therapies versus antidepressants

There were sixteen trials (n = 1793) that reported the effectiveness of CBT compared with antidepressants. Six of those studies were found in the search of the guideline update and ten were reported in the previous guideline. The results for depression scores at post-treatment (BDI: SMD −0.06; 95% CI −0.24 to 0.12; HRSD: SMD 0.05; CI −0.06 to 0.15) and at 1-month follow-up (BDI: SMD −0.02; 95% CI −0.68 to 0.65; HRSD: 0.08; 95% CI −0.59 to 0.74) were not significantly different and this, along with the relatively narrow CIs, suggests broad equivalence between CBT and antidepressants. However, by 12 months’ follow-up the evidence from three trials (Hautzinger [in-pats], Hautzinger1996 and Blackburn1997; n = 137) indicates that CBT has a significant medium effect (BDI: −0.41, 95% CI −0.76, −0.07; HRSD: SMD −0.50; 95% CI −0.84, −0.15) over antidepressants. In terms of leaving the study early, there was a significant higher risk of discontinuation (RR 0.75; 95% CI 0.63, 0.91) in the antidepressant group. A 1-year follow-up of the DIMIDJIAN2006 trial indicates that people who had cognitive therapy were less likely to relapse following treatment than those previously treated with antidepressants (RR 0.82; 95% CI 0.60, 1.11).

Cognitive behavioural therapies versus comparator (waitlist control)

Four low quality studies (two reported in the previous guideline: Beach1992 and Selmi1990; and two found in the update search: DERUBEIS2005 and DIMID-JIAN2006) compared the efficacy of cognitive behavioural therapies versus waitlist control. The effectiveness of CBT for the treatment of depression was large (SMD −0.89; 95% CI −1.45; −0.33) in self-reports and showed an effect in clinician-reported depression scores (RR 0.45; 95% CI 0.23, 0.91).

Combination (cognitive behavioural therapy + antidepressants) versus antidepressants

Nine studies included a comparison between combined treatment of CBT plus anti-depressants and antidepressants alone. Only one of those studies (FAVA199856) was found in the search for this guideline update. The combination treatment of CBT and antidepressants had a lower risk of discontinuation compared with antidepressants (RR 0.81; 95% CI 0.65, 1.01). There is evidence that the combined treatment has a significant medium effect in the reduction of self-rated (SMD −0.38; 95% CI −0.62, −0.14) and clinician-rated (SMD −0.46; 95% CI −0.61, −0.31) depression scores. At 6 and 12 months’ follow-up; however, there was limited data (BDI at 6 months: SMD 0.35; 95% CI −0.69, 1.40; HRSD at 6 months: SMD 0.50; 95% CI −0.53, 1.53; BDI at 12 months: SMD −0.29; 95% CI −0.70, 0.12; HRSD at 12 months: SMD −0.29; 95% CI −0.64, 0.07), which introduced some uncertainty about the relative long-term effectiveness of the combination of these two treatments.

Combination (cognitive behavioural therapy + antidepressants) versus cognitive behavioural therapy

Six studies reported in the previous guideline included a comparison of combination treatment and CBT alone. In contrast with the dataset on the combination of CBT and antidepressants versus antidepressants, it was not possible to identify a benefit for adding antidepressants to CBT (BDI at post-treatment: SMD −0.17, 95% CI −0.44, 0.10; BDI at 1-month follow-up: SMD −0.29, 95% CI −0.94, 0.36; HRSD at 1-month follow-up: SMD −0.08, 95% CI −0.72, 0.57). This might suggest that although the CBT and antidepressants dataset supports combined treatment, clinical benefit could still be derived from CBT alone.

Cognitive behavioural therapies versus comparator (placebo plus clinical management)

There was little evidence of the increased effectiveness of CBT when compared with placebo plus clinical management from two studies (also reported in the previous guideline: Elkin1989 and Jarrett1999; n = 193). There was some indication of higher dropout rates in the placebo groups but the effect (RR 0.44, 95% CI 0.12, 1.61) was not significant and therefore inconclusive. There was a small effect on reducing depression scores at endpoint in favour of CBT (self-rated: SMD −0.15, 95% CI −0.51, 0.21 and clinician-rated: SMD −0.32, 95% CI −0.68, 0.04) when compared with placebo plus clinical management. However, the results were not significant and the CIs were fairly wide so the evidence remains inconclusive.

Cognitive behavioural therapies versus other therapies designed for depression (behavioural activation and interpersonal therapy)

There were three studies that compared cognitive behavioural therapies with behavioural activation in the treatment of depression (DIMIDJIAN2006, Gallagher1982, JACOBSON1996). However, the comparison in the Gallagher1982 study included cognitive therapy following the approach of Beck and colleagues (1979) and Emery (1981) and was compared with behavioural therapy that followed Lewinsohn’s (1975) approach. In addition, the Gallagher1982 study only reported leaving study early data. There were no clinically important differences identified between CBT and behavioural activation (BDI at endpoint: 0.34; 95% CI −0.26, 0.95; HRSD at endpoint: −0.03; 95% CI −0.62, 0.57). From this evidence it is not possible to draw any clear conclusions about the relative efficacy of the treatments. A 1-year follow-up of the DIMIDJIAN2006 trial indicates that people who had cognitive therapy were less likely to relapse following treatment than those previously treated with antidepressants (RR 0.82; 95% CI, 0.60, 1.11).

Four studies included a comparison of CBT versus IPT (Elkin1989, Freeman2002, LUTY2007, MARSHALL2008). Again, there were no clinically important differences between CBT and IPT (BDI at endpoint: 0.21; 95% CI −0.01, 0.41; HRSD at endpoint: 0.13; 95% CI −0.06, 0.32). This evidence although limited suggests that IPT might be as effective as CBT in the treatment of depression.

Cognitive behavioural therapies versus other psychotherapies not specifically designed for depression

There were three studies that looked at the effectiveness of CBT compared with other therapies not specifically designed for depression. Two studies (Gallagher-Th1994 and Shapiro1994) compared CBT with short-term psychodynamic psychotherapy. One study (Rosner1999) compared CBT with gestalt psychotherapy57. The evidence indicates no clinically important differences for the comparison of CBT with short-term psychodynamic psychotherapy in decreasing depression (BDI at endpoint: SMD −0.35; 95% CI −1.30, 0.61) or with Gestalt psychotherapy (BDI at endpoint: SMD 0.17; 95% CI −0.56, 0.91). From this evidence it is not possible to draw any clear conclusions about the relative efficacy of the treatments.

Cognitive behavioural therapies (primary care) versus GP care

Three trials reported in the previous guideline included a comparison between CBT in primary care versus usual GP care. The studies varied in duration: Freeman2002 consisted of 16 sessions over a 5-month period, Scott1992 was 16 weeks’ duration and Scott1997 was 6 weeks. In terms of leaving the study early due to any reason, the evidence suggests that there is a higher risk for discontinuation in those in the CBT (primary care) group (RR 1.54; 95% CI 0.97, 2.46). The evidence here is difficult to interpret as many patients in GP care might have been in receipt of antidepressants and the duration of treatment was shorter than that typical of CBT. At the end of treatment self-report depression scores (SMD 0.01; 95% CI −0.83, 0.85) were not significantly different, and neither were clinician-rated depression scores (SMD −0.33; 95% CI −0.74, 0.08).

Group cognitive behavioural therapies

Three studies reported in the previous guideline looked at the effectiveness of group CBT when compared with other psychotherapies (Bright1999, Covi1987, Klein1984) and no new studies were found that looked at this comparison for the guideline update58. The results show no significant difference in risk for discontinuation (RR 0.94, 95% CI 0.57, 1.53) or depression scores at post-treatment (BDI: SMD −0.17, 95% CI −0.61, 0.26; HRSD: SMD −0.12, 95% CI −0.55, 0.31). However, when self-rated depression scores were analysed by a cut-off of BDI >9, there was a significant difference favouring group CBT (RR 0.60, 95% CI 0.46, 0.79).

A further analysis was carried out looking at group CBT compared with waitlist control or treatment as usual. Four studies evaluated the Coping with Depression programme (see above) (ALLARTVANDAM2003, BROWN1984, DALGARD2006, HARINGSMA2006). The evidence indicates no clinically important difference in risk for discontinuation (RR 1.34, 95% CI 0.44, 4.11). There was a significant medium effect of group CBT in lowering depression scores at endpoint (SMD −0.60, 95% CI −0.84, −0.35) and at 6 months’ follow-up (SMD −0.40, 95% CI −0.83, 0.02). Therefore group CBT (in particular Coping with Depression) appears an effective treatment for people with mild depression.

Cognitive behavioural therapies for the elderly

Three studies looked at the effectiveness of CBT in the treatment of depression in elderly populations. LAIDLAW2008 and Thompson2001 compared CBT with antidepressants. Thompson2001 also included a comparison of the combination of CBT with antidepressants with antidepressants alone. WILKINSON2009 looked at the effectiveness of group CBT in relapse prevention compared with waitlist control.

The evidence was inconclusive regarding leaving the study early. In clinician-rated depression scores, there was a significant medium effect favouring CBT (SMD −0.41; 95% CI −0.79, −0.03). However, the results were not significant for follow-up data (at 3 months: SMD −0.35; 95% CI −0.78, 0.07 and at 6 months: −0.15; 95% CI −0.74, 0.44). The results suggest the effectiveness of CBT seen in adults of working age may be replicated in older adults but some caution is required in interpreting the results.

In the combined treatment of CBT plus antidepressants versus antidepressants alone, there was little to no difference in risk for discontinuation between the two groups (RR 0.92). There were medium effects favouring combined treatments for both self-rated (SMD −0.36; 95% CI −0.84 to 0.12) and clinician-rated depression scores (SMD −0.45; 95% CI −0.93 to 0.03). It should be noted that the CIs for both effects just cross the line of no effect, so these results should be interpreted with caution. The evidence from one trial (WILKINSON2009; n = 43) comparing group CBT plus antidepressants with antidepressants alone in the treatment of depression for the elderly is not significant (BDI ≥12 at 6 months: RR 1.69, 95% CI, 0.68, 4.21; Montgomery–Åsberg Depression Rating Scale [MADRS] ≥ 10 at 6 months: RR 0.26; 95% CI, 0.03, 2.14) and this prevents any clear conclusion being drawn.

Cognitive behavioural therapies – relapse prevention

This section brings together the impact of relapse prevention studies in different areas (group CBT, individual CBT, combination of CBT and antidepressants, and MBCT). A number of studies have addressed the issue of relapse prevention and have developed a number of different approaches both to the patient population identified and the specific CBT approach taken. The approaches include extending the duration of individual CBT, specific group-based approaches including a programme for those with residual symptoms (BOCKTING2005) and MBCT. In total, seven studies (n = 957) found in the search for the guideline update examined relapse prevention in people who had been administered CBT. Three of these studies compare CBT with antidepressants.

Group cognitive behavioural therapy versus treatment as usual

The evidence from one study (BOCKTING2005) indicates a higher risk for discontinuation in those administered group CBT than treatment as usual (RR 2.47; 95% CI 1.01, 6.05). There is insufficient evidence (one study and wide CIs) to determine the comparative effectiveness between the two groups in terms of relapse or remission rates at 68 weeks. Similarly, the evidence indicates a non-significant difference in self-reports of depression in patients with five or more previous episodes of depression (SMD −0.08; 95% CI −0.54, 0.39). It is important to mention that the study that reports this comparison is based on a series of post-hoc analyses and results should be interpreted with caution.

Cognitive behavioural therapy versus clinical management (not shown in tables)

Two studies (FAVA1998 and PAYKEL2005) compared the effectiveness of individual CBT with clinical management (with antidepressants) as part of a relapse prevention programme. They report a significant difference favouring individual CBT in relapse rates when compared with clinical management (RR 0.54; 95% CI 0.37, 0.79). Furthermore, one of the two studies, PAYKEL2005, reports remission at 68 weeks (RR 1.30; 95% CI 0.94, 1.80). However, data at 68 weeks should be interpreted with caution given that there is only one study and the CIs are wide. The two studies mentioned previously are not shown in the tables of study characteristics or in the summary of evidence profiles in this chapter in the interest of brevity and given that these studies report different outcomes from those listed in the tables. These studies, however, appear in the study characteristics tables in Appendix 17b and the forest plots in Appendix 19b.

Combination cognitive behavioural therapy + antidepressants versus antidepressants

When the combination treatment of CBT plus antidepressants was compared with antidepressants alone there were no significant differences in terms of risk for discontinuation (RR 0.96; 95% CI 0.61, 1.52) or relapse (RR 0.80; 95% CI 0.22, 2.85).

Mindfulness-based cognitive therapy

Four studies (CRANE2008, KUYKEN2008, MA2004, Teasdale2000) evaluated the effectiveness of MBCT group treatment in relapse prevention. Two studies (MA2004, Teasdale2000) compared the combined treatment of group MBCT with GP care versus GP care alone. The evidence indicates a higher risk for discontinuation in the combined treatment (RR 19.11, 95% CI 2.58, 141.35) but a significantly lower risk for relapse (RR 0.74, 95% CI 0.57, 0.96). Regarding the reduction of relapse rates, group MBCT when compared with antidepressants showed a small to medium effect of group MBCT lowering depression scores at 1-month (BDI: SMD −0.37, 95% CI −0.72, −0.01; HRSD: SMD −0.31, 95% CI −0.66, 0.05) and at 15 months’ follow-up (BDI: SMD −0.34, 95% CI −0.69, 0.02; HRSD: SMD −0.23, 95% CI −0.59, 0.12).

8.1.5. Health economic evidence and considerations

Two studies were identified in the systematic literature review that evaluated the cost effectiveness of cognitive behavioural therapies for people with depression (Kuyken et al., 2008; Scott et al., 2003). Details on the methods used for the systematic search of the health economics literature are described in Chapter 3, Section 3.6.1. Evidence tables for all health economics studies are presented in Appendix 15.

Kuyken and colleagues (2008) evaluated the cost effectiveness of MBCT compared with maintenance antidepressant medication in 123 patients with a history of depression participating in a primary care-based RCT. The time horizon of the analysis was 15 months and both a health service and societal perspective were taken in separate analyses. Costs included all hospital care, community health and social services and any productivity losses resulting from time off work. The outcome measures used in the cost-effectiveness analysis were the mean total number of relapses/recurrences avoided and the mean total number of depression-free days. Over 15 months’ follow-up, there was no significant difference in total mean costs between MBCT and antidepressant treatment (US $3,370 versus $2,915; p = 0.865). From an NHS and Personal Social Services (PSS) perspective, the ICER was $429 per relapse/recurrence prevented and $23 per depression-free day. From a societal perspective, the ICER was $962 per relapse/recurrence prevented and $50 per depression-free day. The authors suggested that the additional cost of MBCT may be justified in terms of improvements in the proportion of patients who relapsed.

Scott and colleagues (2003) evaluated the cost effectiveness of cognitive therapy added to antidepressants and clinical management compared with antidepressants and clinical management alone in a UK RCT of 154 patients with partially remitted major depression. The setting was either in local clinics or in participants’ homes. The time horizon of the analysis was 68 weeks (including 20 weeks of treatment). The study estimated NHS costs including treatments, clinical management, hospital care, primary care, group and marital therapy and medication for this period. The primary outcomes used in the analysis were relapse rates for the two treatment groups. Overall, the cognitive therapy group was significantly more costly than standard clinical treatment, with a mean difference of £779 per person (p < 0.01). The ICER of cognitive therapy versus standard care was £4,328 per relapse averted or £12.50 per additional relapse-free day. The authors concluded that in individuals with depressive symptoms that are resistant to standard treatment, adjunctive cognitive therapy was more costly but more effective than intensive clinical treatment alone.

8.2. BEHAVIOURAL ACTIVATION

8.2.1. Introduction

Behavioural activation for depression evolved from learning theory that posits two types of learning: operant or instrumental learning and classical conditioning. Although classical conditioning theories for depression have been put forward (for example, Wolpe, 1971) with treatment recommendations (Wolpe, 1979) there have been no treatment trials of this approach. Operant or instrumental learning posits that depressive behaviours are learned through the contingencies around those behaviours. In behavioural therapies, depression is seen as the result of a low rate of positive reinforcement and is maintained through negative reinforcement. Most commonly, patients use avoidance to minimise negative emotions and situations they worry will be unpleasant. Behavioural therapies focus on behavioural activation aimed at encouraging the patient to develop more rewarding and task-focused behaviours as well as stepping out of patterns of negative reinforcement. The approach was developed by Lewinsohn (1975). In recent years there has been renewed interest in behavioural activation (for example, Jacobson et al., 2001; Hopko et al., 2003), as it is now known, as a therapy in its own right, although it has always been part of cognitive behavioural treatments of depression (Beck et al., 1997).

Definition

Behavioural activation is defined as a discrete, time-limited, structured psychological intervention, derived from the behavioural model of affective disorders and where the therapist and patient:

  • work collaboratively to identify the effects of behaviours on current symptoms, feelings states and/or problem areas
  • seek to reduce symptoms and problematic behaviours through behavioural tasks related to: reducing avoidance, graded exposure, activity scheduling, and initiating positively reinforced behaviours.

8.2.2. Studies considered59

There were six studies involving a comparison of behavioural activation. Of these, four were found in the searches for the guideline update and two were from the previous guideline. Two further studies were identified, which were excluded: CULLEN2006 because of a lack of extractable data60 and Thompson1987 because it was unclear which patient numbers were used in their table reporting outcome measures and the dropout data was not fully reported. Comparisons between behavioural activation and cognitive behavioural therapies can be found in the previous section (see Section 8.1). One study, McLean1979, entailed a comparison with psychotherapy. HOPKO2003 compared behavioural activation with an attentional control (the control had the same duration of contact in a group but no therapy was given) in an inpatient setting. A further study, DIMIDJIAN2006, entailed a comparison between behavioural activation and antidepressants, as well as a comparison between behavioural activation and placebo.

Summary study characteristics of the included studies are presented in Table 42, with full details in Appendix 17b, which also includes details of excluded studies. Studies comparing CBT and behavioural activation are reported in Section 8.1 and therefore have not been included in Table 42.

Table 42. Summary study characteristics of behavioural activation.

Table 42

Summary study characteristics of behavioural activation.

8.2.3. Clinical evidence

Because there are a relatively small number of studies for behavioural activation a summary of evidence profile table has not been included here. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

8.2.4. Clinical evidence summary

Behavioural activation versus cognitive behavioural therapy

Studies comparing behavioural activation and CBT are reported in the CBT summary evidence profile tables – see the section ‘Cognitive behavioural therapies versus other therapies designed for depression (behavioural activation and interpersonal therapy)’. In summary, there were three studies included (DIMID-JIAN2006, Gallagher198261 and JACOBSON1996). Gallagher1982 only reported leaving the study early data. There were no clinically important differences identified between CBT and behavioural activation (BDI at endpoint: 0.34; 95% CI −0.26, 0.95; HRSD at endpoint: −0.03; 95% CI −0.62, 0.57). From this evidence it is not possible to draw any clear conclusions about the relative efficacy of the treatments.

Behavioural activation versus placebo

Only one study (DIMIDJIAN2006) included a comparison of behavioural activation versus placebo. The evidence suggests there is no significant difference between treatments in risk for discontinuation (RR 1.23; 95% CI 0.33, 4.64). Similarly, there were no significant differences between treatments in the reduction of depression scores (self-reported, BDI: SMD 0.07; 95% CI, −0.61, 0.75 and clinician reported, HRSD: SMD 0.06; 95% CI, −0.62, 0.73). These results are based on one medium-sized study and given its wide CIs it is difficult to make any firm conclusions from this evidence.

Behavioural activation versus other interventions

One study, McLean1979, compared behavioural activation with an attentional control. The study used a short-term psychotherapy (10 weeks of 1-hour sessions) following Marmor (1973, 1975) and Wolberg (1967), the aim of which is the development of insight of the psychodynamic forces that initiated the patient’s current depression. From this study, only leaving the study early data could be extracted, and their results indicate an increased risk for discontinuation in the control group (RR 0.17; 95% CI, 0.04, 0.71). It should be noted that this evidence is based on one study and the CIs are wide.

The second study, HOPKO2003, compared behavioural activation with a supportive treatment (three times weekly, 20 minutes for 14 days), which was a non-directive discussion with the clinician in which the patient was encouraged to share their experiences. The results at post-treatment favoured behavioural activation (BDI: SMD −0.69; 95% CI, −1.52, 0.14). However, this result is not significant and should be interpreted with caution.

Behavioural activation versus antidepressants

There is limited evidence from one study (DIMIDJIAM2006) of the effect of behavioural activation in the treatment of depression when compared with antidepressants. This limited evidence seems to indicate a low risk of discontinuation in the people administered antidepressants when compared with those in the behavioural activation group (RR 0.31; 95% CI 0.12, 0.83). In terms of depression scores, the results were not significant but tended to favour the antidepressant group in those diagnosed with moderate severity (self-reported scores: SMD 0.15; 95% CI −0.47, 0.78 and clinician-reported scores: SMD 0.14; 95% CI −0.49, 0.77) and in those with high severity (self-reported scores: SMD 0.24; 95% CI −0.29, 0.76 and clinician-reported scores: SMD −0.04; 95% CI −0.56, 0.49). There seems to be little to no difference between behavioural activation and antidepressants in terms of relapse rates at 1 year (RR 1.04; 95% CI 0.49, 2.21).

8.2.5. Health economic evidence and considerations

No evidence on the cost effectiveness of behavioural activation for people with depression was identified by the systematic search of the economic literature. Details on the methods used for the systematic search of the economic literature are described in Chapter 3, Section 3.6.1.

8.3. PROBLEM SOLVING

8.3.1. Introduction

It has long been recognised that depression is associated with social problem-solving difficulties (Nezu, 1987). The reasons for this may be various, relating to the effects of depressed state, lack of knowledge, and rumination. As a consequence, helping patients solve problems and develop problem-solving skills has been a focus for therapeutic intervention and development of therapy (Nezu et al., 1989). There has been recent interest in developing problem-solving therapies for depression for use in primary care (Barrett et al., 1999; Dowrick et al., 2000).

Definition

Problem-solving therapy is a discrete, time-limited, structured psychological intervention, which focuses on learning to cope with specific problems areas and where therapist and patient work collaboratively to identify and prioritise key problem areas, to break problems down into specific, manageable tasks, problem solve, and develop appropriate coping behaviours for problems.

8.3.2. Studies considered62

No new studies found in the search for the guideline update were included. Two studies were found and excluded on the basis of one study not reporting the outcome data (AREAN2008) and one study having a sample size of less than ten (NEZU1986). Three studies were reported in the previous guideline but only two are included in the update (Mynors-Wallis1995; Mynors-Wallis2000). Dowrick and colleagues (2000) which was included in the previous guideline, was excluded from this update because not all patients met criteria for depression (<80%).

Summary study characteristics of the included studies are presented in Table 43, with full details in Appendix 17b, which also includes details of excluded studies.

Table 43. Summary study characteristics of problem solving.

Table 43

Summary study characteristics of problem solving.

8.3.3. Clinical evidence

Evidence from the important outcomes and overall quality of evidence are presented in Table 44. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

Table 44. Summary evidence profile for problem solving.

Table 44

Summary evidence profile for problem solving.

8.3.4. Clinical evidence summary

Only two studies were found that met the inclusion criteria for problem solving and only one study (Mynors-Wallis1995) indicated that this intervention had a significant effect in reducing depression scores (clinician-rated: SMD −0.66; 95% CI −1.21, −0.12; self-rated: SMD −0.69; 95% CI −1.24, −0.14) when compared with placebo. This effect was also seen for dichotomous scores: clinician-rated (RR 0.55; 95% CI 0.33, 0.89) and self-rated (RR0.62; 95% CI 0.39, 0.99). A further study (Dowrick2000) indicated a significant decrease in the number of people diagnosed with depressive and subthreshold depressive symptoms after 6 months of treatment (RR 0.83; 95% CI 0.68, 1.02) when compared with placebo. However this trial did not meet the inclusion criteria for the guideline due to 80% or more of the population in the trial not meeting diagnosis for depression and therefore does not appear in the tables above.

There were no significant differences when problem solving was compared with antidepressants or when the combination treatment of problem solving and antidepressants was compared with antidepressants alone, but the uncertainty surrounding these results makes it difficult to draw any conclusions.

8.3.5. Health economic evidence and considerations

One study was identified in the systematic literature review that evaluated the cost effectiveness of problem solving for people with common mental health problems (including depression and anxiety disorders) (Kendrick et al., 2006a). Details on the methods used for the systematic search of the health economics literature are described in Chapter 3, Section 3.6.1. References to included studies and evidence tables for all health economics studies are presented in the form of evidence tables in Appendix 15.

Kendrick and colleagues (2006a) evaluated the cost effectiveness of problem solving delivered by community mental health nurses (CMHNs) compared with usual GP care and generic CMHN care. The setting was primary care and the study population included adult patients with a new episode of anxiety, depression or reaction to life difficulties (33% with a primary diagnosis of depression). The time horizon of the analysis was 26 weeks and two separate analyses were undertaken from a health service and societal perspective. Costs estimated in each treatment group included nurse training and supervision, primary care, social worker and psychiatrist, hospital care plus out-of-pocket patient costs and productivity losses due to time off work. The outcome measures used in the analysis were QALYs, estimated using utility scores derived from the EQ-5D questionnaire. Total direct health service costs and productivity losses were higher over 26 weeks in the problem solving group compared with GP or CMHN care. Overall, the mean cost difference between the problem solving and GP groups was £315 per patient (p < 0.001). No significant differences in utility scores or QALYs were detected between the three treatment groups at 26 weeks’ follow-up. The results of the incremental analysis showed that both problem solving and generic CMHN care were dominated by GP care. The mixed population in this study limits its relevance to this guideline.

8.4. COUPLES THERAPY

8.4.1. Introduction

Therapists have noted that a partner’s critical behaviour may trigger an episode of depression, and/or maintain or exacerbate relapse in the long term (for example, Hooley & Teasdale, 1989), although other researchers have questioned this (for example, Hayhurst et al., 1997). There has also been some research looking at differences in the vulnerabilities between men and women within an intimate relationship, with physical aggression by a partner predicting depression in women. Difficulties in developing intimacy, and coping with conflict, also predict depression in both men and women (Christian et al., 1994). Like other therapies, couples therapy has evolved in recent years. Systemic couples therapy aims to give the couple new perspectives on the presenting problem (for example, depressing behaviours), and explore new ways of relating (Jones & Asen, 1999). Other developments such as those by Jacobson and colleagues (1993) took a more behavioural approach. In the analysis of couples therapy in this guideline, the focus of the search was not on a specific approach but on couples therapy more generally.

Definition

Couples therapy is defined as a time-limited, psychological intervention derived from a model of the interactional processes in relationships where:

  • the intervention aims to help participants understand the effects of their interactions on each other as factors in the development and/or maintenance of symptoms and problems
  • the aim is to change the nature of the interactions so that the participants may develop more supportive and less conflictual relationships.

8.4.2. Studies considered63

Six RCTs were included in the review of couples therapy. Two studies were found in the search for the guideline update (BODENMANN2008 and JACOBSON1993) and four were also included in the previous guideline. One study (Leff et al., 2000), which was included in the previous guideline, was excluded from the update because more than 50% of the participants dropped out from one arm of the study; this study used a systemic approach based on the Jones and Asen (1999) manual.

Summary study characteristics of the included studies are presented in Table 45, with full details in Appendix 17b, which also includes details of excluded studies.

Table 45. Summary study characteristics of couples therapy.

Table 45

Summary study characteristics of couples therapy.

8.4.3. Clinical evidence

Evidence from the important outcomes and overall quality of evidence is presented in Table 46. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

Table 46. Summary evidence profile for couples therapy.

Table 46

Summary evidence profile for couples therapy.

8.4.4. Clinical evidence summary

In five of the studies included in this review the model used was a behavioural model; two other studies used a model based on IPT. Two studies (Beach1992 and O’Leary1990) indicated a significant large effect in reducing depression self-report scores at post-treatment (SMD −1.35; 95% CI −1.95, −0.75) when compared with waitlist control. In a larger dataset where couples therapy is compared with individual CBT, there were no significant differences in risk for discontinuation (RR 1.22; 95% CI 0.55, 2.71) or depression scores at post-treatment (BDI: SMD −0.10; 95% CI −0.58, 0.38; HRSD: −0.07, 95% CI −0.69, 0.55) or at 6 months’ follow-up (BDI: SMD −0.05, 95% CI −0.67, 0.57) suggesting couples therapy has broadly similar effects to CBT. There is some indication of an effect in reducing self-reported depression scores at 1 year’s follow-up (SMD −0.41; 95% CI −0.90 to 0.09) but this does not persist to 1 and a half years (SMD −0.08, 95% CI −0.70, 0.54). Two studies (BODENMANN2008 and Foley1989) compared couples therapy with IPT. The results from these two small-sized studies had wide CIs and therefore it is difficult to interpret the comparison of the two treatments with any confidence.

8.4.5. Health economic evidence and considerations

No evidence on the cost effectiveness of couples therapy for people with depression was identified by the systematic search of the economic literature. Details on the methods used for the systematic search of the economic literature are described in Chapter 3, Section 3.6.1.

8.5. INTERPERSONAL THERAPY

8.5.1. Introduction

Interpersonal therapy (IPT) was developed by Klerman and Weissman (Klerman et al., 1984) initially for depression although it has now been extended to other disorders (Weissman et al., 2000). IPT focuses on current relationships, not past ones, and on interpersonal processes rather than intra-psychic ones (such as negative core beliefs or automatic thoughts as in CBT, or unconscious conflicts as in psychodynamic psychotherapy). It is time limited and focused on difficulties arising in the daily experience of maintaining relationships and resolving difficulties during an episode of major depression.

The main clinical tasks are to help patients to learn to link their mood with their interpersonal contacts and to recognise that, by appropriately addressing interpersonal situations, they may simultaneously improve both their relationships and their depressive state. Early in the treatment, patient and therapist agree to work on a particular focal area that would include: interpersonal role transitions, interpersonal roles/conflicts, grief and/or interpersonal deficits. IPT is appropriate when a person has a key area of difficulty that is specified by the treatment (for example, grief or interpersonal conflicts). It can be delivered as an individually focused therapy but has also been developed as a group therapy (Wilfley et al., 2000).

The character of the therapy sessions is, largely, facilitating understanding of recent events in interpersonal terms and exploring alternative ways of handling interpersonal situations. Although there is not an explicit emphasis on ‘homework’, there is an emphasis on effecting changes in interpersonal relationships and tasks towards this end may be undertaken between sessions.

Definition

IPT was defined as a discrete, time-limited, structured psychological intervention, derived from the interpersonal model of affective disorders that focuses on interpersonal issues and where the therapist and patient:

  • work collaboratively to identify the effects of key problematic areas related to interpersonal conflicts, role transitions, grief and loss, and social skills, and their effects on current symptoms, feelings states and/or problems
  • seek to reduce symptoms by learning to cope with or resolve these interpersonal problem areas.

8.5.2. Studies considered64

Twenty-two trials were identified; 14 were included and eight were excluded. The most common reasons for exclusion were: that the trials did not report the outcome data, that they included populations without a diagnosis of depression and that they used an unclear control intervention. Of the 14 studies that were included, six were found in the new search for the guideline update and eight were also included in the previous guideline. Three studies included a comparison of IPT with CBT, and these results are reported in Section 8.1. From the 14 included studies there were three examining IPT as a continuation treatment; two studies looked at IPT as a 3-year maintenance treatment; and four studies looked at IPT in an older population. It is important to mention that one study, Reynolds1999, is a four-arm trial of an elderly population, including IPT as an acute treatment, then as a continuation treatment, and finally, for those who recovered, they were randomised to IPT as a maintenance treatment. The terms ‘continuation’ and ‘maintenance’ have been used interchangeably in many trials. In this guideline continuation treatment is defined as a treatment that occurs after the acute symptoms have subsided, when the patient could be considered to be substantially improved and the aim is to achieve remission or significant improvements in symptoms and restore normal function. Maintenance treatment occurs when the episode is considered to have remitted or significantly improved, the patient is stable, but treatment is continued to avoid recurrence.

Summary study characteristics of the included studies are presented in Table 47, Table 48, Table 49 and Table 50 with full details in Appendix 17b, which also includes details of excluded studies.

Table 47. Summary study characteristics of IPT.

Table 47

Summary study characteristics of IPT.

Table 48. Summary study characteristics of IPT as continuation treatment (up to 6 months).

Table 48

Summary study characteristics of IPT as continuation treatment (up to 6 months).

Table 49. Summary study characteristics of IPT as maintenance treatment (3 years).

Table 49

Summary study characteristics of IPT as maintenance treatment (3 years).

Table 50. Summary study characteristics of IPT for the elderly.

Table 50

Summary study characteristics of IPT for the elderly.

8.5.3. Clinical evidence

Evidence from the important outcomes and overall quality of evidence are presented in Table 51, Table 52, Table 53 and Table 54. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

Table 51. Summary evidence profile for IPT.

Table 51

Summary evidence profile for IPT.

Table 52. Summary evidence profile for IPT as continuation treatment (up to 6 months).

Table 52

Summary evidence profile for IPT as continuation treatment (up to 6 months).

Table 53. Summary evidence profile for IPT as maintenance treatment (3 years).

Table 53

Summary evidence profile for IPT as maintenance treatment (3 years).

Table 54. Summary evidence profile for IPT for the elderly.

Table 54

Summary evidence profile for IPT for the elderly.

8.5.4. Clinical evidence summary

Three studies included a comparison of IPT with CBT, and these results are reported in Section 8.1. Only one study, Elkin1989 (n = 123) looked at IPT when compared with placebo. There was a higher risk for discontinuation in the placebo group when compared with IPT (RR 0.57; 95% CI 0.33, 0.99). Furthermore, there was a significant small to medium effect (SMD −0.43; 95% CI −0.79, 0.07 and RR 0.73; 95% CI 0.56, 0.93) for IPT in reducing clinician-rated depression scores at post-treatment when compared with placebo.

Four studies looked at IPT compared with usual GP care (including medication). The data for the Freeman2002 study is unpublished data the review team for the previous guideline obtained from the authors in anticipation of it being published. However, the study had still not been published when the guideline update was being prepared, and it is important to take this into consideration when interpreting the results. The evidence indicated a significant effect in self-reported depression scores at post-treatment (SMD −0.69; 95% CI −1.22, −0.16). In addition, there was a large effect for IPT in reducing self-report depression scores at 3 months’ (SMD −0.88; 95% CI −1.48, −0.28) and 9 months’ (SMD −0.73; 95% CI −1.32, −0.13) follow-up. Similarly, in clinician-rated depression reports there was a large effect at 3 months’ (SMD −0.81; 95% CI −1.41, −0.21) and 9 months’ (SMD −0.98; 95% CI, −1.60, −0.37) follow-up.

Based on the evidence of one study (Reynolds1999) the combination treatment of IPT plus antidepressants when compared with IPT alone had a significant difference in decreasing clinician-rated depression scores (RR 2.26; 95% CI 1.03, 4.97). Furthermore, one study, SCHRAMM2007, showed that when combination treatment was compared with antidepressants alone there was a significant medium effect (SMD −0.40; 95% CI −0.75, −0.05) in the reduction of clinician-rated depression scores post-treatment.

Two studies, Elkin1989 and Schulberg1996, examined the effectiveness of IPT versus antidepressants alone. The evidence showed no significant differences among the two groups (for depression scores: BDI post-treatment SMD 0.04; 95% CI −0.32, 0.40; HRSD post-treatment SMD 0.08; 95% CI, −0.15, 0.30).

Interpersonal therapy as a continuation treatment

The evidence of one study (Schulberg1996) showed a small to medium significant effect (SMD −0.44; 95% CI −0.73, −0.15) for IPT in reducing depression scores after 4 months’ continuation treatment when compared with treatment as usual.

Based on the evidence of two studies with a continuation time of 3 years (Frank1990, Reynolds1990) the evidence indicates that combining interpersonal therapy and antidepressants has a lower risk of relapse when compared with IPT plus placebo (RR 0.17; 95% CI 0.01, 3.51). This significant effect was also seen when combination treatment was compared with antidepressants (RR 0.42; 95% CI 0.02, 9.34) and also when compared with medication clinics (RR 5.50; 95% CI 0.26, 115.22).

Interpersonal therapy as maintenance treatment

Only two studies included a comparison of IPT as a maintenance treatment (Frank1990 and Reynolds1999B). When IPT was studied as a maintenance treatment, combination treatment had a significant effect in lowering the risk of relapse (RR 0.42; 95% CI 0.27, 0.65) when compared with IPT plus placebo and (RR 0.22; 95% CI 0.10, 0.49) when compared with medication clinics.

Interpersonal therapy for the elderly

The evidence for IPT in an elderly population is based on four studies (n = 284). One study (Reynolds1999; n = 33) indicated a significant effect (RR 2.26; 95% CI 1.03, 4.97) for reducing clinician-rated depression scores in an elderly population, favouring combination treatment of IPT plus antidepressants when compared with IPT alone.

Based on the same study (Reynolds1999; n = 42), antidepressants had a significant effect in reducing clinician-rated depression measures (RR 1.60; 95% CI 0.94, 2.75) when compared with IPT.

8.5.5. Health economic evidence and considerations

No evidence on the cost effectiveness of IPT for people with depression was identified by the systematic search of the economic literature. Details on the methods used for the systematic search of the economic literature are described in Chapter 3, Section 3.6.1.

8.6. COUNSELLING

8.6.1. Introduction

Counselling was developed by Carl Rogers (1957) who believed that people had the means for self-healing, problem resolution and growth if the right conditions could be created. These conditions include the provision of positive regard, genuineness and empathy. Rogers’s original model was developed into structured counselling approaches by Truax and Carkhuff (1967) and, independently, by Egan (1990) who developed the three stage model: exploration, personalising and action. Voluntary sector counselling training (for example, Relate) tends to draw on these models. However, although many other therapies now use the basic ingredients of client-centred counselling (Roth & Fonagy, 1996), there are differences in how they are used (Kahn, 1985; Rogers, 1986) and counselling has become a generic term used to describe a broad range of interventions delivered by counsellors usually working in primary care. The content of these various approaches may include psychodynamic, systemic or cognitive behavioural elements (Bower et al., 2003).

Definition

The British Association for Counselling and Psychotherapy (BACP) defines counselling as ‘a systematic process which gives individuals an opportunity to explore, discover and clarify ways of living more resourcefully, with a greater sense of well-being’.

8.6.2. Studies considered65

Three new studies (GOLDMAN2006, GREENBERG1998, WATSON2003) meeting the inclusion criteria were found in the update search. Three studies (Bedi2000, Simpson2003, Ward2000) were reported in the previous guideline, two of which are included in the guideline update. Ward2000 was excluded because it did not meet inclusion criteria: only 62% met diagnosis for depression and this study was not completely randomised. However, as this study was included in the previous guideline a separate sub-analysis has been conducted to determine whether this would have affected the GDG’s conclusions. The results of this sub-analysis do not appear in the tables, but are described in the text below. A further trial (Stiles et al., 2006, a non-RCT) was examined, but it was ultimately excluded because not all patients met criteria for depression and there were concerns about the selection of the study population.

Summary study characteristics of the included studies are in Table 55, with full details in Appendix 17b, which also includes details of excluded studies. Two studies, GOLDMAN2006 and GREENBERG1998, are not listed in Table 55 because these compare two different types of counselling.

Table 55. Summary study characteristics of counselling.

Table 55

Summary study characteristics of counselling.

8.6.3. Clinical evidence

Evidence from the important outcomes and overall quality of evidence are presented in Table 56. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

Table 56. Summary evidence profile for counselling.

Table 56

Summary evidence profile for counselling.

8.6.4. Clinical evidence summary

One study (Bedi2000) compared the effectiveness of counselling versus antidepressants, although some differences in the baseline scores of the patient preference group suggest caution in interpreting the data. There were no significant differences and the evidence remains inconclusive (self-reported depression scores at endpoint: SMD 0.04; 95% CI −0.38, 0.47 and at 12-month follow-up: SMD: 0.17; 95% CI −0.32, 0.66; clinician-rated depression scores at endpoint: RR 1.20; 95% CI 0.80, 1.81) and does not support a conclusion that counselling and antidepressants are equivalent. This caution is support by the 12-month follow-up data; clinician-reported depression scores were significantly reduced in the antidepressant group when compared with counselling (RR 1.41; 95% CI 1.08, 1.83). The results of this study should be treated with some caution as the introduction of a patient preference element to the trial led to considerable differences in baseline severity measures between the two arms.

One study (Simpson2003) compared the combination of counselling plus GP care with usual GP care. There was no evidence of any important clinical benefit of counselling plus GP care (BDI at 6 months: SMD 0.06; 95% CI −0.29, 0.40 and at 12 months: SMD 0.03; 95% CI −0.33, 0.40). A sub-analysis was conducted on Ward2000, which did not meet the inclusion criteria but was raised during the consultation process. This study included a comparison of counselling versus GP care. The results indicated a significant medium effect in self-report depression scores at post-treatment (SMD −0.49; 95% CI −0.83, −0.15) but no significant differences between the two treatment groups on discontinuation and self-report depression scores at follow-up.

The comparison of counselling versus CBT was included in one study (WATSON2003). There is insufficient evidence (only one small-sized study with wide CIs) to reach any definite conclusion about the relative effectiveness of these two treatments (for BDI scores post-treatment: SMD 0.04; 95% CI −0.38, 0.47). This was still the case when a sub-analysis including the Ward2000 study was conducted (SMD −0.07; 95% CI −0.33, 0.20). (Individual outcomes for Ward2000 are: at endpoint BDI: SMD −0.14; 95% CI −0.48, 0.21 and at 12-month follow-up BDI: SMD 0.04; 95% CI −0.31, 0.38.)

Two studies (GOLDMAN2006 and GREENBERG1998), compared two different types of counselling (and therefore are not included in the tables above). GOLDMAN2006 compared client-centered counselling with emotion-focused counselling. The results favoured emotioned-focused therapy (BDI scores: SMD 0.64; 95% CI −0.02, 1.29). GREENBERG1998 examined the effectiveness of client-centered counselling versus process-experiential counselling. The evidence indicates that there was no significant difference between treatments in reduction of self-reported depression scores (SMD 0.13; 95% CI, −0.57, 0.82). These two studies are small in size and therefore results should be interpreted with caution.

The participants in the trials included in this review were predominantly drawn from groups in the mild-to-moderate range of depression (mean baseline BDI scores between 18 and 26) and two trials included people with minor depression (BDI scores starting from 14) (Bedi2000 and Ward2000).

Overall the evidence for counselling is very limited. Some practice-based evidence was also reviewed (Stiles et al., 2006) but the number of patients with depression in the study fell below the cut-off for inclusion. Furthermore, other diagnoses were included in this study. A smaller practice-based study (Marriott & Kellett, 2009), which included only 34% with a diagnosis of depression, compared counselling, cognitive analytic therapy and CBT but it was small and underpowered and it was not possible to reach any conclusion on the differential effectiveness of the treatments. In addition to the limited data available for counselling, interpretation of the results is complicated by the different therapeutic models adopted in the studies. For example, Bedi2000 and Ward2000 follow a Rogerian client-centred model of counselling, Simpson2003 a psychodynamic model, whereas the studies by WATSON2003, GREENBERG1998 and GOLDMAN2006 adopt a process-experiential/emotion-focused model, which is compared in the latter two trials with the Rogerian client-centred model.

8.6.5. Health economic evidence and considerations

Three studies were identified in the systematic literature review that evaluated the cost effectiveness of counselling for people with depression and other common mental health problems (Friedli et al., 2000; Miller et al., 2003; Simpson et al., 2003). Details on the methods used for the systematic search of the health economics literature are described in Chapter 3, Section 3.6.1. Evidence tables for all health economics studies are presented in Appendix 15.

Friedli and colleagues (2000)66 compared non-directive counselling with usual GP care in a UK RCT of 136 people with referral symptoms being caused by depression and anxiety disorders (50% were given a GP diagnosis of depression). The time horizon of the analysis was 9 months and direct NHS costs (hospital inpatient stay, outpatient consultations and medications) and non-health service costs (lost productivity, travel and childcare) were estimated over this period. The primary outcome of the clinical analysis was change in BDI scores. However, as no differences in clinical outcomes were detected between the two groups, the study was effectively a cost-minimisation analysis. Over 9 months, total direct NHS costs were £309 and £474 per person while total non-health service costs were £809 and £469 per person, for counselling and GP care, respectively. The authors concluded that counselling in primary care was not cost effective in the short-term if indirect costs were taken into account but that, overall, referral to counselling was no more clinically effective or costly than GP care.

Miller and colleagues (2003)67 compared counselling with antidepressants in patients with major depression who were recruited from general practice. Sixty five patients were randomised to either treatment modality while a further 183 patients who chose their treatment modality were also analysed. The time horizon of the analysis was 12 months and direct NHS costs (inpatient, outpatient, counselling, GP consultations and medications) were estimated. The primary outcome measure used in the analysis was change in BDI scores. However, no significant differences were detected between the two treatment groups at 12 months. Overall, no significant differences in total mean costs per person were detected between the two randomised groups while the non-randomised counselling group was significantly more costly than the non-randomised antidepressant treatment group over 12 months. The authors suggested that counselling might be a more cost-effective intervention in patients with mild to moderate depression but, for the larger patient group, antidepressant treatment was likely to be the more cost-effective intervention.

Simpson and colleagues (2003) evaluated the cost effectiveness of short-term psychodynamic counselling compared with routine GP care in a UK RCT of 181 patients with a history of depression. The time horizon of the analysis was 12 months and direct healthcare costs (specialist mental health, hospital, primary care and community health and social care services) were estimated for this period. The primary outcome measure used in the clinical analysis was change in BDI scores. However, since there were no significant clinical differences detected between the two treatment groups, the study was effectively a cost-minimisation analysis. Overall, there was no statistically significant difference between the two treatment groups in total costs per person over 12 months (£1046 versus £1074). The authors suggested that there was no cost-effectiveness advantage of counselling over routine care for general practice attendees with chronic depression.

8.7. SHORT-TERM PSYCHODYNAMIC PSYCHOTHERAPY

8.7.1. Introduction

As with other schools of psychological therapy there are a number of variations on the original model of psychodynamic psychotherapy with some approaches focusing on the dynamic of drives (for example, aggression) while others focus on relationships (Greenberg & Mitchell, 1983). Other forms of this therapy have been influenced by attachment theory (Holmes, 2001). Clinical trials of psychodynamic psychotherapy have focused on short-term psychological therapy (typically 10 to 30 weeks) usually in comparison with antidepressants or CBT. It is this brief version of psychodynamic psychotherapy, often referred to as short-term psychodynamic psychotherapy, which is the focus of this review.

Definition

Psychodynamic psychotherapy is defined as a psychological intervention derived from a psychodynamic/psychoanalytic model, and where:

  • Therapist and patient explore and gain insight into conflicts and how these are represented in current situations and relationships including the therapeutic relationship (for example, transference and counter-transference). This leads to patients being given an opportunity to explore feelings and conscious and unconscious conflicts, originating in the past, with a technical focus on interpreting and working through conflicts.
  • Therapy is non-directive and recipients are not taught specific skills (for example, thought monitoring, re-evaluating, or problem-solving).

8.7.2. Studies considered68

In total, 17 studies were found in the search for trials of short-term psychodynamic psychotherapy. Ten studies were included (six were found in the update search and four were also reported in the previous guideline) and seven were excluded. Reasons for exclusion included: trials not being RCTs, papers not reporting outcome data, trials including participants without a diagnosis of depression and authors replacing dropouts. Two studies (Gallagher-Th1994 and Shapiro1994) included a comparison of short-term psychodynamic psychotherapy with CBT, the results of which are reported in Section 8.1. One study (McLean1979), compared short-term psychodynamic psychotherapy with behavioural activation and this is reported in Section 8.2. One study (Guthrie1999) was not included because in the sample population (which was selected on the basis of high attendance at outpatient clinics) only 73.6% met diagnosis for depression and, therefore, did not meet the inclusion criteria (which is >80% of the total population). However, while it is not included in the main analyses and tables, a sub-analysis including this paper was conducted and is reported below.

It should be noted that all the included studies were of short-term psychodynamic psychotherapy and therefore the analysis and subsequent recommendations are limited to short-term psychodynamic psychotherapy, typically 16 to 20 sessions but with a range of 10 to 30 sessions across the included studies.

Summary study characteristics of the included studies are in Table 57, with full details in Appendix 17b, which also includes details of excluded studies.

Table 57. Summary study characteristics of short-term psychodynamic psychotherapy.

Table 57

Summary study characteristics of short-term psychodynamic psychotherapy.

8.7.3. Clinical evidence

Evidence from the important outcomes and overall quality of evidence are presented in Table 58. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

Table 58. Summary evidence profile for short-term psychodynamic psychotherapy.

Table 58

Summary evidence profile for short-term psychodynamic psychotherapy.

8.7.4. Clinical evidence summary

Problems with unextractable data and multiple different comparators limited the analyses it was possible to undertake for this review. The evidence from one study (DEKKER2008) showed a significant medium effect (SMD 0.43; 95% CI 0.03, 0.82) favouring antidepressants when compared with short-term psychodynamic psychotherapy in the reduction of clinician-rated scores at endpoint. However, the results of a further small-sized study (SALMINEN2008) showed no significant differences between short-term psychodynamic psychotherapy and antidepressants when looking at the mean change from baseline to endpoint (SMD 0.03; 95% CI −0.52, 0.58), but given the wide CIs and size of the study it is difficult to establish a clear picture of this comparison. One study (McLean1979) indicated a significantly higher risk of discontinuation in those treated with short-term psychodynamic psychotherapy when compared with behaviour therapy (RR 3.02; 95% CI 1.07, 8.50).

When compared with a waitlist control, one study (MAINA2005) showed a significant and large effect (SMD −1.09; 95% CI −2.04, −0.13) in clinician-rated depression scores at post-treatment, favouring short-term psychodynamic psychotherapy. This study also indicated a large effect (SMD −0.97; 95% CI −1.91, −0.03) for short-term psycho-dynamic psychotherapy in clinician-rated depression scores at post-treatment when compared with supportive therapy. A follow-up study (MAINA2008) showed that adding short-term psychodynamic psychotherapy to antidepressant treatment had a significant medium to large effect at 24 months (SMD 0.52; 95% CI 0.10, 0.95) and at 48 months (SMD 0.59; 95% CI 0.16, 1.01) in reducing clinician-rated depression scores when compared with antidepressants alone. MAINA2005 and MAINA2008 were conducted in a population diagnosed with minor depression or dysthymia. KOOL2003 compared short-term psychodynamic psychotherapy in two different populations with depression: one with comorbid personality disorder and the second without. The results suggest that short-term psychodynamic psychotherapy is more effective in people diagnosed with depression and personality disorder than those without (SMD −1.15; 95% CI −1.62, −0.69 and SMD 1.50; 95% CI 0.81, 2.18 respectively) but the small sample size in the population without personality disorder suggests caution when interpreting this result.

When a separate analysis was conducted with the Guthrie1999 study, the evidence was inconclusive given the small size of the study and the wide CIs (for SCL-90-R at endpoint: SMD −0.16; 95% CI −0.53, 0.22 and at 6-month follow-up SMD −0.24; 95% CI −0.62, 0.13).

In summary, this is a weak dataset characterised by a number of the findings being contradictory and/or difficult to interpret. Some of the difficulty derives from there being a number of different comparators in a small dataset. There is limited evidence for a benefit of short-term psychodynamic psychotherapy (typically 16 to 20 sessions over 4 to 6 months) in a population with dysthymia and subthreshold depressive symptoms over waitlist or usual care and inconsistent findings when compared with antidepressants. Comparisons against other active psychological interventions are also very limited.

8.7.5. Health economic evidence and considerations

One study (Guthrie et al. 1999) was identified in the systematic literature review that evaluated the cost effectiveness of short-term psychodynamic psychotherapy for people who are high utilisers of psychiatric services (with 73.6% having a diagnosis of depression). Details on the methods used for the systematic search of the health economics literature are described in Chapter 3, Section 3.6.1. Evidence tables for all health economics studies are presented in Appendix 15.

The study by Guthrie and colleagues (1999) compared brief psychodynamic interpersonal therapy (equivalent to short-term psychodynamic psychotherapy) with treatment as usual in a UK RCT of 144 patients with non-psychotic disorders (75.5% diagnosed with depression). The time horizon of the analysis was 6 months post-treatment and direct NHS costs (inpatient, outpatient, day cases, A&E visits and medications) and non-health service costs (travel and lost productivity) were estimated during this period. The primary outcome measures used in the economic analysis were quality-adjusted life months (QALMs), which were estimated from utility weights derived from the EQ-5D questionnaire. Overall, total societal costs per person were lower in the brief psychodynamic interpersonal therapy group at 6 months (US $1959 versus $2,465; p = 0.21). The brief psychodynamic interpersonal therapy group also gained more QALMs during this period (4.87 versus 3.48; p = 0.13). While brief psychodynamic interpersonal therapy appeared to dominate treatment as usual, resulting in lower costs but better outcomes, neither the cost nor QALM differences between the two treatment groups were statistically significant.

8.8. RATIONAL EMOTIVE BEHAVIOURAL THERAPY

8.8.1. Introduction

Rational emotive behavioural therapy is a form of CBT developed by Albert Ellis in the 1950s and 1960s (Ellis, 1962). Compared with CBT it has been subject to fewer research trials, and only one study met the criteria of the GDG (DAVID2008). This study compared rational emotive behavioural therapy with antidepressant medication.

Definition

Rational emotive behavioural therapy is a present-focused, relatively short-term therapy usually delivered one-to-one that uncovers and addresses the relationships between thoughts, feelings and behaviours. There is an emphasis on addressing thinking that underpins emotional and behavioural problems. Patients learn how to examine and challenge their unhelpful thinking.

8.8.2. Studies considered69

Only one RCT (DAVID2008) was found and was included in the review. This section reports on the comparison of rational emotive behavioural therapy with antidepressants; comparison with CBT can be found in Section 8.1.

Summary study characteristics of the included studies are in Table 59, with full details in Appendix 17b, which also includes details of excluded studies.

Table 59. Summary study characteristics of rational emotive behaviour therapy.

Table 59

Summary study characteristics of rational emotive behaviour therapy.

8.8.3. Clinical evidence

Because of the small dataset, a summary of the evidence profile is not included here. The full evidence profiles and associated forest plots can be found in Appendix 16b and Appendix 19b, respectively.

8.8.4. Clinical evidence summary

The evidence of one study (DAVID2008) showed no clinically important different effects of rational emotive behaviour therapy in depressed patients when compared with antidepressants (BDI: SMD −0.07; CI −0.44 to 0.29; HRSD: SMD 0.00; CI −0.37 to 0.37). However, the findings were promising in terms of end-of-treatment depressive symptoms and in terms of acceptability (RR 0.63; 95% CI 0.22 to 1.80) and preventing relapse at 6 months’ follow-up (RR 0.20; 95% CI 0.02 to 1.61).

8.8.5. Health economic evidence and considerations

No evidence on the cost effectiveness of rational emotive behavioural therapy for people with depression was identified by the systematic search of the economic literature. Details on the methods used for the systematic search of the economic literature are described in Chapter 3, Section 3.6.1.

8.9. ECONOMIC MODELLING

8.9.1. Background

The aim of this economic analysis is to update the model constructed in the previous guideline (NICE, 2004a), which evaluated the cost effectiveness of antidepressant treatment versus a combination of antidepressant treatment and CBT for the routine treatment of moderate/severe depression. It was anticipated that other high-intensity psychological interventions such as IPT or behavioural activation would be evaluated in an economic model. However, evidence of the clinical effectiveness of IPT or behavioural activation compared with antidepressant treatment for moderate or severe depression was limited. Based on GDG expert opinion, CBT was again chosen as the form of psychological therapy for this analysis as the clinical evidence was superior and CBT remains more widely available in the UK compared with other high-intensity interventions.

Clinical outcome data within the model including rates of discontinuation, remission and relapse remained the same as reported in the previous guideline model. It should be noted that these data were taken from meta-analyses that were undertaken in the previous guideline. Therefore, in this economic analysis, levels of depression severity in relation to the HRSD and BDI were based on those proposed by the APA (2000a) rather than those proposed in this guideline. However, it was necessary to update the economic model in order to better reflect current medical practice within the UK. This included the additional costs of maintenance therapy in both treatment groups while other input parameters, including patient utility scores and unit costs, were also updated.

8.9.2. Methods

A pragmatic decision analytic model was constructed using Microsoft Excel XP. Within the model patients either continue or discontinue their initial treatment, after which they enter remission or no remission health states. Patients in remission can then either relapse or remain in remission health states. A detailed structure of the decision tree is presented in Figure 6. A time horizon of 15 months was chosen to reflect the available comparative clinical evidence. This included 3 months of the initial therapy, followed by 6 months’ maintenance therapy and 6 months’ follow-up. The following strategies were considered:

Figure 6. Structure of the model.

Figure 6

Structure of the model.

  • Strategy A: Antidepressant treatment given for 12 weeks with 6 months’ maintenance therapy and 6 months’ follow-up (AD).
  • Strategy B: Combination of 12 weeks’ antidepressant treatment and 16 sessions of CBT with 6 months’ maintenance therapy and 6 months’ follow-up (COMB).

Originally, three specific strategies for the first-line management of depression were considered. However, similar to the previous guideline, the updated clinical evidence review showed no overall superiority for CBT alone on treatment outcomes over antidepressant treatment. The efficacy evidence combined with the significantly higher treatment cost of CBT compared with the cost of antidepressant treatment resulted in the exclusion of CBT alone from the final analysis.

8.9.3. Model assumptions

Population

Two separate models were constructed for a hypothetical cohort of 100 patients in each treatment group with either moderate or severe depression.

Resource use and unit costs

An NHS and PSS perspective was taken for the analysis based on current NICE guidance (NICE, 2008b). Therefore, only direct health and social care costs were considered in the analysis. In order to cost the two therapy pathways, resource utilisation data were collected as part of the literature review or from GDG expert opinion. Unit costs were obtained from a variety of sources including the British National Formulary (BNF 56, 2008) and the Personal Social Services Research Unit (PSSRU) (Curtis, 2009). Resource utilisation data were then combined with the relevant cost associated with each therapy. All costs were based on 2007/08 prices and were inflated where necessary using Hospital and Community Health Service indices (Curtis, 2009). As in the case of outcomes, no discounting was applied because the time horizon was 15 months.

Antidepressant treatment

The antidepressant treatment protocol consisted of 12 weeks plus 6 months’ maintenance period of 40 mg of generic citalopram per day for both moderate and severely depressed patients (GDG expert opinion). Citalopram was used to represent standard pharmacotherapy for patients with moderate or severe depression because it was the most commonly prescribed antidepressant in 2007 in England (Department of Health, 2008a). As part of patient monitoring, it was assumed that all patients with moderate depression and 50% of patients with severe depression would receive standard GP care while the remaining 50% of patients with severe depression would receive specialist mental health outpatient care (GDG expert opinion). It was also assumed that patient monitoring in both primary and secondary care consisted of two fortnightly visits in the first month followed by one visit per month, while the maintenance therapy period consisted of one GP/specialist visit every 2 months (GDG expert opinion).

Combination therapy

For both moderate and severely depressed patients, it was assumed that combination therapy would consist of 16 sessions of CBT over 12 weeks, in addition to the anti-depressant treatment protocols described above (GDG expert opinion). One CBT session lasts for 55 minutes and is provided by a specialty doctor, clinical psychologist or mental health nurse (Curtis, 2009). During the 6-month maintenance therapy period, it was assumed that both moderate and severely depressed patients would receive an additional two CBT sessions, in addition to the antidepressant (AD) maintenance therapy protocols described above (GDG expert opinion).

Subsequent healthcare

Patients who discontinued initial treatment did not incur the full costs of treatment. To revise costs downwards, it was assumed that patients who discontinued initial treatment would drop out after 4 weeks of treatment, irrespective of treatment group (Rush et al., 2006; GDG expert opinion). For patients in remission who did not relapse during follow-up, it was assumed that no further additional treatment or mental healthcare resources beyond the 6-month maintenance period were required. However, for patients with unsuccessful treatment outcomes, it was assumed that they would continue to consume additional mental healthcare resources over the 15-month time horizon. Cost data for subsequent mental healthcare were taken from a study published by the King’s Fund which estimated annual mental healthcare costs for respondents with mild, moderate and severe depressive disorder based on the UK psychiatric morbidity survey (McCrone et al., 2008). As such, these annual mental healthcare costs may be an under estimate of the actual costs incurred by patients with moderate and severe depression, as one would expect respondents with mild depression to use less mental healthcare on average. These mental healthcare costs included hospital and outpatient care, social services, residential care, GP visits and medication costs. These annual costs were divided into monthly cost estimates and then projected for the periods during which unsuccessfully treated patients would consume subsequent mental healthcare estimated in the model. According to the survey, only 65% of people with depression were in contact or receipt of mental health services. Therefore, these subsequent mental healthcare costs were weighted downwards based on the assumption that 35% of patients would not incur any further healthcare costs. Patients who did not achieve remission following therapy incurred full 3-month treatment costs followed by subsequent mental healthcare thereafter. For patients who relapsed while in remission, it was assumed that the average time to relapse was based on the midpoint of the clinical relapse data elicited in the guideline meta-analysis, which was estimated over a 12-month period (GDG expert opinion). Full details of all resource use and unit cost parameters are presented in Table 60.

Table 60. Resource use and cost estimates applied in the economic model.

Table 60

Resource use and cost estimates applied in the economic model.

Clinical outcomes and event probabilities

The outcome measure used for the economic evaluation was the quality-adjusted life years (QALYs) gained from either treatment. No discounting of outcomes was necessary since the time horizon of the model was 15 months. The key clinical parameter estimates – discontinuation rates, remission rates and relapse rates – were collected as part of the updated clinical systematic review undertaken for the guideline. The dichotomous outcome measure of no remission was defined by scores greater than six on the 17-item HRSD or more than eight on the 24-item HRSD.

For the base case analysis, the baseline absolute rates of remission, discontinuation and relapse for antidepressant treatment as well as the respective relative risks of combination therapy versus antidepressant treatment were taken from the relevant guideline meta-analyses. The guideline meta-analysis of non-remission rates was based on intention-to-treat analysis, with non-completers being considered as an ‘unfavourable’ outcome (that is, as non-remitters). This means that non-remission rates included people who completed treatment but did not remit plus people who did not complete treatment. For the economic analysis, the proportion of non-remitters in the completer group was estimated from the available data, and was subsequently incorporated in the respective branch of the decision tree.

For patients who did not complete their initial therapy, it was assumed that rather than remaining moderately or severely depressed, a small proportion (20%: 95% CI 10, 30) would spontaneously enter remission (GDG expert opinion). For patients in remission, the rate of relapse was estimated as 67% based on a study of patients who were not receiving maintenance antidepressant treatment (Murphy et al., 1984). Therefore, this is likely to be an over estimate of the relapse rate for patients in this analysis who are receiving maintenance therapy. These two probabilities were applied to patients in both treatment arms. For the sensitivity analyses, 95% CIs around the relevant relative risks of combination (COMB) therapy versus antidepressant (AD) treatment were used. Full details of event probabilities are presented in Table 62.

Table 62. Clinical effectiveness parameters applied in the economic model.

Table 62

Clinical effectiveness parameters applied in the economic model.

Utility data and estimation of quality-adjusted life years

In order to express outcomes in the form of QALYs, the health states of the economic model needed to be linked to appropriate utility scores. Utility scores represent the HRQoL associated with specific health states on a scale from 0 (death) to 1 (perfect health); they are estimated using preference-based measures that capture people’s preferences and perceptions on HRQoL characterising the health states under consideration.

8.9.4. Systematic review of published utility scores for adults with depression

Among the studies already assessed for eligibility, eight publications were identified that reported utility scores relating to specific health states and events associated with depression (Revicki & Wood, 1998; Bennett et al., 2000; King et al., 2000; Lenert et al., 2000; Schaffer et al., 2002; Pyne et al., 2003; Sapin et al., 2004; Peveler et al., 2005).

Three studies used the EQ-5D Index instrument, currently recommended by NICE as a measure of patient HRQoL for use in cost-utility analyses (King et al., 2000; Sapin et al., 2004; Peveler et al., 2005). In all three studies, preference values elicited from the UK population sample were used (Dolan & Williams, 1995). King and colleagues (2000) collected patient EQ-5D utility data over 12 months’ follow-up in an RCT comparing usual GP care with two types of brief psychological therapy (non-directive counselling and CBT) among patients with depressive or mixed anxiety/depressive symptoms (BDI >14). Patient utility, reported as median scores, improved from baseline in all three treatment groups at 4 and 12 months. However, no differences in median scores were detected between the three patient groups. The study by Peveler and colleagues (2005) was another HTA based on an RCT comparing the cost-utility of TCAs, SSRIs and lofepramine among UK patients with a new episode of depressive illness (based on GP diagnosis). Patients completed the EQ-5D questionnaire on a monthly basis over 12 months. Again, utility scores improved from baseline at 12 months in all three treatment groups but no differences were detected between groups.

The study by Sapin and colleagues (2004) was based on a multicentre, prospective cohort of patients with a new episode of major depressive disorder recruited in the French primary care setting assessed at 8 weeks’ follow-up. EQ-5D utility scores were stratified according to depression severity (defined by CGI scores), and by clinical response (defined by MADRS scores) at follow-up. At 8 weeks, patients with MADRS scores lower or equal to 12 were considered as ‘remitters’ and others considered as ‘non-remitters’. Patients with a decrease of at least 50% in relation to baseline score were considered as ‘responders’ and others as ‘non-responders’. These two patient groupings also led to the creation of three mutually exclusive groups: ‘responder remitters’, ‘responder non-remitters’ and ‘non-responders’.

The other five studies used a variety of instruments to measure patient utility (Revicki & Wood, 1998; Bennett et al., 2000; Lenert et al., 2000; Schaffer et al., 2002; Pyne et al., 2003). Bennett and colleagues (2000) used a disease-specific measure, the McSad instrument, to estimate utility scores for a cross-sectional sample of patients who had experienced at least one episode of major unipolar depression in the previous 2 years. McSad is a direct utility measure in which rating scale (RS) and standard gamble (SG) techniques were used to obtain utilities for specific health states. The health state classification system contains six dimensions (emotion/self-appraisal/cognition/physiology/behaviour/role-function), each with four levels of dysfunction (none/mild/moderate/severe). Utility scores were generated for three temporary clinical marker states of 6 months’ duration (mild/moderate/severe depression) and chronic states of lifetime duration (self-reported and severe depression).

Lenert and colleagues (2000) estimated utility scores among depressed US primary care patients based on six health states according to level of depression severity (mild/severe) and physical impairment (mild/moderate/severe). Cluster analysis was applied to the SF-12 HRQoL instrument to generate the six health states. Utilities applied to the six health states were elicited through the use of visual analogue scale (VAS) and SG methods. The resulting six-state health index model was then applied to HRQoL data taken from a longitudinal cohort study of patients with current major depression or dysthymia.

Pyne and colleagues (2003) used the self-administered Quality of Well-Being scale (QWB-SA) in a prospective cohort of US patients treated with antidepressants to measure change in patient HRQoL scores over 4 months’ follow-up. The scoring function of the QWB-SA was based on rating scale measurements taken from a random sample of the US population. QWB-SA scores improved during follow-up for treatment responders (defined by a 50% reduction in HRSD-17 scores) but did not improve for non-responders.

Revicki and Wood (1998) used SG techniques in US and Canadian patients with major depressive disorder to generate utility scores for 11 hypothetical depression-related and current health states according to depression severity and antidepressant treatment. The depression-related health states varied depression severity (mild/-moderate/severe) and medication (nefazodone/fluoxetine/imipramine), were framed in terms of 1 month’s duration and described symptom severity, functioning and well-being, as well as medication therapy including side effects.

Similarly, the study by Schaffer and colleagues (2002) used SG techniques to elicit utility scores for ten individual symptom profiles of major depression plus three ‘clinical marker’ depression profiles (mild/moderate/severe) among patients with current or past depression. The individual symptom profiles each consisted of five statements describing a particular aspect of a symptom of depression, incorporating the content of several depression scales and interviews (HRSD, BDI, MADRS, DSM–IV and Structured Clinical Interview for DSM-IV [SCID-IV]).

8.9.5. Summary

Table 61 summarises the methods used to derive health states and estimate utility scores associated with various levels of depression severity and treatments for depression as well as utility scores from each study. Overall, the studies reviewed here reported significant impact of depression on the HRQoL of patients with depression. A number of studies indicated that patients valued the state of severe depression as being close to zero (death) (Revicki & Wood, 1998; Bennett et al., 2000). There was some limited evidence to suggest that generic utility measures such as the EQ-5D may be less sensitive than disease-specific measures such as the McSad health state classification system.

Table 61. Summary of studies reporting utility scores relating to specific health states and events associated with depression.

Table 61

Summary of studies reporting utility scores relating to specific health states and events associated with depression.

NICE currently recommends the EQ-5D as the preferred measure of HRQoL in adults for use in cost-utility analyses. NICE also suggests that the measurement of changes in HRQoL should be reported directly from people with the condition examined, and the valuation of health states be based on public preferences elicited using a choice-based method such as time trade-off (TTO) or SG, in a representative sample of the UK population (NICE, 2008b). Therefore, based on these recommendations, the EQ-5D utility scores estimated by Sapin and colleagues (2004) were considered to be the most suitable for calculating QALYs in the guideline economic models. Although these utility scores were based on a cohort of French primary care patients, which may limit their applicability to the UK setting, preference values assigned to health states were elicited from the UK population sample. Furthermore, utility scores were stratified according to disease severity and clinical response, which is useful for modelling health states in cost-utility analyses. Full details of utility scores used in the model are presented in Table 62.

Estimation of quality-adjusted life years

By applying the utility scores estimated by Sapin and colleagues (2004), the QALY profiles over 15 months were estimated when patients entered the three end health states in the model (no remission; no relapse; relapse) based on the following assumptions for patients who completed treatment:

  • No remission: a linear increase from baseline utility score (0.33 or 0.15) to the ‘no response’ health state (0.58) over the initial 3-month treatment period; decreasing immediately back to their baseline utility over the remaining 12 months.
  • No relapse: a linear increase from baseline utility to the ‘response with remission’ health state (0.85) over the initial 3-month treatment period; remaining in the ‘response with remission’ health state for the following 12 months.
  • Relapse: a linear increase from baseline utility to the ‘response with remission’ health state over the initial 3-month treatment period; followed by a linear deterioration back to baseline utility over the remaining 12 months.

For patients who did not complete their initial treatment, the following assumptions were used:

  • No response: patient remains at baseline utility (0.33 or 0.15) over 15 months.
  • Relapse: a linear increase from baseline utility to the ‘response – no remission’ health state (0.72) over 3 months; followed by linear decrease back to baseline utility over the remaining 12 months.
  • No relapse: a linear increase from baseline utility to the ‘response – no remission’ health state over 3 months; followed by linear increase to ‘response with remission’ health state over the remaining 12 months.

Incremental cost effectiveness of COMB versus antidepressant treatment

The incremental cost effectiveness of COMB compared with antidepressant treatment for patients with moderate or severe depression was evaluated by assessing the difference in costs and effectiveness of each therapy. The ICERs were calculated as the difference in the expected healthcare costs divided by the difference in the overall effectiveness of the two strategies.

Sensitivity analyses

Deterministic sensitivity analysis

Given the considerable uncertainty around some of the input parameters used in the base case model and ambiguity surrounding any policy implications of point estimates, one-way sensitivity analysis was undertaken. This involved varying a single parameter between its plausible minimum and maximum values while maintaining all remaining parameters in the model at their base case value. Uncertainty around the various transition probabilities and quality-of-life weights, as well as the cost implications of different levels of resource use involved in patient clinical management, were explored.

Probabilistic sensitivity analysis

To demonstrate the joint uncertainty between the different parameters, probabilistic analysis is required. Using the mean point estimates and their 95% CIs, appropriate distributions were assigned for each parameter estimate. For example, lognormal distributions were applied to relative risk estimates, gamma distributions to cost estimates and beta distributions to utility estimates and absolute rates. For cost estimates that did not have 95% CIs, a standard error based on 30% of the mean estimate was applied to reflect any potential uncertainty around these estimates. Effectiveness and cost estimates were then recalculated 10,000 times using Monte Carlo simulation. Whether an intervention is cost effective or not depends on how decision makers value the additional health gain achieved by the therapy. The probability that COMB therapy is cost effective compared with AD treatment as a function of decision makers’ maximum willingness-to-pay for an additional successfully treated patient or QALY was illustrated by CEACs (Briggs, 2000).

8.9.6. Results

Clinical outcomes

The systematic review of the clinical evidence showed that the probability of not completing the initial 3-month therapy was higher for AD than for COMB (RR = 0.80, 95% CI 0.65, 1.01) while the probability of not achieving remission following therapy was also lower in the COMB group (RR = 0.76, 95% CI 0.55, 1.03). The two follow-up studies suggested that there is a lower risk of relapse in the COMB therapy arm (RR = 0.68, 95% CI 0.38, 1.24) over a 12-month follow-up period although this was not statistically significant (p = 0.21).

Quality-adjusted life years

The decision model for patients with moderate depression resulted in an average of 0.67 QALYs per patient in the COMB therapy group and 0.58 QALYs per patient in the AD group. The decision model for patients with severe depression resulted in an average of 0.53 QALYs per patient in the COMB therapy group and 0.42 QALYs in the AD group. Therefore, the average gain in QALYs over 15 months for COMB therapy was 0.09 per patient with moderate depression and 0.11 per patient with severe depression.

Costs and cost effectiveness

The full cost of a 3-month course of antidepressant treatment plus 6-month maintenance therapy was £270 for patients with moderate depression and £599 for patients with severe depression. The full cost of 3-month COMB therapy, including a full course of CBT, plus 6 months’ maintenance therapy was £1314 for patients with moderate depression and £1643 for patients with severe depression. The expected subsequent health and social care cost over 15 months for patients who did not complete their initial therapy was £1638 for both moderate and severe patients. The expected subsequent health and social care cost over 15 months for patients who did not respond to therapy and achieve remission was £1404 for both patient groups. The expected subsequent health and social care cost of relapse while in remission was £702 for both patient groups.

Incremental cost effectiveness of COMB versus antidepressant treatment

Overall, COMB therapy was estimated to be significantly more effective and more costly than antidepressant treatment for patients with both moderate and severe depression. On average, the strategy of COMB therapy was £624 more costly per patient with moderate depression and £653 more costly per patient with severe depression. The resulting base case ICERs were £7,052 per QALY gained for moderate depression and £5,558 per QALY gained for severe depression.

Sensitivity analyses

Deterministic sensitivity analysis

The parameter values used in the sensitivity analyses and the resultant ICERs are presented in Table 63. The results of the deterministic sensitivity analysis indicated that the results were fairly robust when single parameters were varied over their uncertainty ranges. The cost-effectiveness estimates were most sensitive to: (1) the relative risk of no remission following therapy completion and; (2) the relative risk of relapse while in remission. This is explained by the high uncertainty around the relative risk estimate of no remission and to a lesser extent around the relative risk of relapse for COMB versus AD. Other factors had a much lesser role in the variation of the results.

Table 63. Deterministic sensitivity analysis.

Table 63

Deterministic sensitivity analysis.

Probabilistic sensitivity analysis

In order to present the results of the probabilistic sensitivity analysis, CEACs were constructed (see Figure 7). The CEAC indicates the probability of COMB therapy being cost effective for a range of threshold values. The threshold value represents the maximum a decision maker would be willing to pay for a unit of effect, in this case a QALY.

Figure 7. CEACs of COMB therapy versus AD for patients with moderate and severe depression.

Figure 7

CEACs of COMB therapy versus AD for patients with moderate and severe depression.

Current NICE guidance sets a threshold range of £20,000 to £30,000 per QALY (NICE, 2008a). Within this threshold range, the probability of COMB therapy being cost effective for patients with moderate depression was 86 to 90% and for patients with severe depression was 88 to 92%.

8.9.7. Discussion

In this economic evaluation, CBT was chosen as the psychological therapy and citalopram as the antidepressant drug being compared. An updated cost-effectiveness model was constructed to investigate the difference in clinical outcomes and direct health and social care costs between the different strategies. The updated clinical evidence review indicated that CBT alone may be more costly yet less clinically effective than antidepressant treatment and so it was excluded from the final model. As combination therapy is both more effective and more costly than antidepressant treatment, these strategies were compared in a formal cost-effectiveness analysis.

Two separate analyses were conducted for patients with moderate and severe depression. The difference in costs between combination therapy and antidepressant treatment was slightly higher for patients with severe depression, while the difference in QALY gains was also slightly higher. The cost results for patients with both moderate and severe depression suggest that although the initial treatment cost of combination therapy is substantially higher, these costs were partially offset by savings due to lower subsequent treatment costs. Overall, the results of the analysis indicate that combination therapy is likely to be a cost-effective first-line treatment for both moderate and severe depression.

Limitations of the analysis

The clinical effectiveness estimates used in the analyses were based on efficacy data obtained from RCTs, resulting in possible over estimates of successful outcomes for both treatment options provided within the NHS setting. However, this is unlikely to significantly influence the relative effectiveness of the two treatment options.

Another issue concerns the time horizon used for the analysis. A 15-month time horizon was used, with remission rates applied at the end of the initial 3 months of treatment and relapse rates applied during the 12-month follow-up period. One study in the clinical evidence review indicated lower relapse rates with combination therapy versus antidepressant treatment for up to 6 years after treatment (Fava et al., 2004). This suggests that the relative cost effectiveness of combination therapy versus anti-depressant treatment may be underestimated when based on a short time horizon. It would have been preferable to evaluate the two strategies over a longer follow-up period, but the lack of direct clinical evidence beyond 15 months precluded this.

Depression incurs significant non-healthcare costs, such as social service costs, direct costs to patients and their families and lost productivity costs because of morbidity or premature mortality (Thomas & Morris, 2003; McCrone et al., 2008). As this analysis was conducted from the health service and PSS perspective, as per NICE guidance, such non-healthcare costs were not considered. It is likely that the inclusion of these costs would have further increased the probability of combination therapy being cost effective compared with antidepressant treatment.

8.10. FROM EVIDENCE TO RECOMMENDATIONS

This section synthesises the evidence from the clinical summaries of all the psychological interventions reviewed in this chapter and the health economic evidence. This is because some key recommendations about psychological therapies are common to all types of interventions and also because a number of the recommendations draw on evidence from several different reviews. Overall, the evidence indicates that psychological interventions have a beneficial effect in the treatment of people with depression and they do not have an increased risk for discontinuation when compared with antidepressants. However, the evidence suggests that there are differences in the evidence base for the effectiveness among the psychological interventions reviewed in this chapter, and this is the focus of this section.

8.10.1. Cognitive behavioural therapies

With 46 studies, cognitive behavioural therapies have the largest evidence base. Within this group of studies, the largest dataset is that which compares individual CBT with antidepressants and which shows broad equivalence of effect across the range of severity. The clinical effectiveness data also points to a clear advantage of combination treatment over antidepressants alone. This is supported by the outcome of the health economics model, which suggested that combination treatment is cost effective not just for severe depression but also for moderate depression, and as a result the recommendations from the previous guideline were changed. The outcome of the model does not support the simple adoption of combination treatment as the first choice, but as potentially the most cost-effective option because of its greater benefit despite the increased cost. The GDG took the view that for patients with moderate depression a number of options, including antidepressants alone and CBT alone (CBT alone was found to be better than antidepressants alone when both were compared with combined treatment), should be available. This should then allow for a discussion between patient and clinician in which a number of factors are taken into account, including the demands of adhering to the various treatment options and experience of past treatment, when determining treatment choice.

A more limited (because the evidence relates primarily to mild depression) dataset was examined for group CBT. The group CBT approach (based on the Coping with Depression model) showed evidence of benefit at post-treatment and at follow-up over a waitlist control. There was no clear evidence for the effectiveness of other models of group CBT.

Given the relapsing and remitting nature of depression, the GDG looked closely at the evidence for relapse. The most important evidence came from two sources: the studies comparing CBT with antidepressants, which showed a reduced relapse rate for CBT in the follow-up of individual trials; and the data from psychological interventions specifically designed to reduce relapse. The provision of individual CBT is therefore one option when there are concerns over the risk of relapse (an almost ever-present concern with people who have had more than two episodes of depression) and should be considered along with the evidence reviewed for pharmacological interventions and relapse prevention (see Chapter 12). Of the treatments specifically designed to reduce relapse, group-based mindfulness-based cognitive therapy has the strongest evidence base with evidence that it is likely to be effective in people who have experienced three or more depressive episodes.

The GDG also reviewed the relative effectiveness of CBT against a range of other psychological interventions; the detailed outcomes of these reviews are set out in the sections for these interventions below. In brief, the GDG found evidence for some other interventions including IPT and couples therapy that suggested, in some comparisons, broadly similar effects to CBT and, to a lesser extent, for behavioural activation. The GDG did not consider the evidence for counselling, short-term psychodynamic psychotherapy or problem solving therapy to be as strong as that for other interventions (see below).

In making recommendations on CBT, the GDG were conscious of feedback from stakeholders from the previous guideline (NICE, 2004a) and of their experiences of providing or receiving psychological treatments. This led the group to specify in greater detail than previously the ways in which all psychological treatments in this guideline (including CBT) should be provided. It also led the GDG, after considering the evidence, to remove the previous recommendation about the provision of brief CBT because the GDG did not think that the rather limited evidence for brief CBT justified such a specific recommendation. There was concern that this recommendation had led to an unnecessary restriction on the number of sessions of psychological intervention being made available. Instead, the GDG elected to recommend that the duration of treatment should be in line with that found in the majority of trials but also suggested that the target in treatment should be remission and, should that be achieved after fewer than the recommended sessions, then treatment need not be continued beyond that point.

The GDG also took into account the evidence on the delivery of effective psychological interventions reviewed in Chapter 6 and used this to develop a number of recommendations on the need to adhere to, as far as possible, the treatments set out in the trials, as well as the need for routine outcome monitoring and the use of appropriate training and supervision. (Note that this approach has been adopted for all recommendations for psychological therapies in this guideline.)

The GDG felt it was important to locate all psychological interventions in proper relation to each other, having considered the evidence on clinical and cost effectiveness. This meant developing recommendations that locate all psychological interventions at the appropriate place in the stepped-care model. Low-intensity interventions are clinically and cost effective for subthreshold depressive symptoms and mild depression, and therefore are to be preferred over individual and group CBT (and other high-intensity psychological interventions) as the initial treatment for subthreshold depressive symptoms and mild depression. Group CBT is an effective treatment for mild depression but, given the duration of the group and the staffing of such groups, it was viewed on cost-minimisation grounds as less cost effective than low-intensity interventions but more cost effective than individual CBT, and so was placed between them in the stepped-care model.

8.10.2. Behavioural activation

There has been renewed interest in behavioural activation as a treatment for depression and a number of new studies were identified for the review in this guideline. It is also a component part of cognitive behavioural interventions for depression and one of the first important trials of behavioural activation was a deconstruction study (Jacobson1999). No direct evidence on the cost effectiveness of behavioural activation as a high- or low-intensity intervention was identified in the systematic review of the health economics literature, although it should be noted that the duration and frequency of high-intensity behavioural activation is identical to that of high-intensity CBT. It was also not possible to evaluate the cost effectiveness of behavioural activation in an economic model because of the limited clinical evidence available. However, consideration was given to the emergence of new evidence, including data on comparisons with placebo, antidepressants, CBT and usual care, all of which were positive for behavioural activation (that is, there was no evidence of the superiority of these other interventions). Note was also taken of the evidence for the effectiveness of behavioural activation in low-intensity interventions. The GDG decided that although the evidence was not sufficiently robust to recommend behavioural activation as a direct alternative individual treatment option to CBT or IPT, it could be considered as an option. The GDG did, however, decide that healthcare professionals should be made aware of the more limited evidence base for behavioural activation when compared with CBT, IPT and couples therapy (see below).

8.10.3. Problem solving therapy

Problem solving therapy was recommended as a separate intervention in the last guideline. No new studies were identified, leaving a limited dataset based only on two studies with much of the evidence for effectiveness being dependent on one study (Mynors-Wallis1995). In light of the improved evidence for a range of low-intensity interventions that have emerged since the last guideline, the GDG decided not to recommend problem solving therapy as a separate intervention in this guideline. However, the GDG expects that it will continue to be one of the component parts of the low-intensity interventions offered for the treatment of depression (Richards & Whyte, 2008). In the health economics literature, one study was found that suggested that problem solving therapy delivered by community mental health nurses was not cost effective compared with usual GP care in patients with new episodes of anxiety, depression and life difficulties (Kendrick et al., 2006a).

8.10.4. Couples therapy

In the review for this guideline, a number of additional studies were included and one from the previous guideline was excluded. The evidence base for couples therapy is relatively modest with just six studies, but there are indications of a beneficial effect in couples with depression (in particular, those who adopted a behavioural approach to treatment) when compared with waitlist control, and evidence of similar outcomes for couples therapy when compared with individual CBT and IPT (although the evidence in comparison with IPT is more uncertain). As a result of the increased evidence identified in this guideline, couples therapy (based on a behavioural model) is recommended; however, the GDG did not consider it appropriate to offer it as a direct alternative to CBT or IPT, but rather decided that it should be focused on patients in established relationships where the relationship may play a role in the development, maintenance or resolution of depression, because these issues were typical of the patients who entered the trials reviewed in this guideline. Only one study was identified in the health economics literature, which suggested that couples therapy may be a cost-effective treatment compared with antidepressant medication for patients with major depressive disorder (Leff et al., 2000), but this study was excluded from the clinical evidence analyses because of its high attrition rate.

8.10.5. Interpersonal therapy

The evidence for the effectiveness of IPT reviewed in this guideline confirms the picture in the previous guideline of IPT as an effective treatment for depression. However, the dataset is not as large as that for CBT, nor is the evidence for the range of applications for IPT as strong or as wide ranging (for example the evidence on group- or individual-based approaches to relapse prevention). There was also no good economic evidence for the effectiveness of IPT and it was also not possible to evaluate the cost effectiveness of IPT in an economic model because of the limited clinical evidence available. Therefore, the GDG did not develop recommendations for IPT that were as broad in scope as for CBT (for example, the use of combination CBT and antidepressant drugs as the initial treatment for severe depression), but for many patients with mild to moderate depression IPT is an appropriate alternative to CBT.

8.10.6. Counselling

The evidence base for counselling identified in the previous guideline was small (only three studies – one of which, Ward2000, did not meet current inclusion criteria). Of the new studies identified, only one provided relevant data on an important comparison relevant to the effectiveness of counselling (WATSON2003), while one did not meet inclusion criteria and two other studies compared different forms of counselling (GOLDMAN2006, GREENBERG1998). An inconsistent picture of the effectiveness of counselling emerges from the review, with one trial having poorer outcomes against usual care (Simpson2003) and one against antidepressants (Bedi2000), but no difference identifiable in the two comparisons with CBT (Ward2000, WATSON2003). Two studies identified in the health economics literature suggested no advantage, in terms of cost effectiveness, of counselling compared with either usual GP care or antidepressant treatment in adults with depression (Friedli et al., 2000; Miller et al., 2003). Furthermore, a review of the practice-based evidence did not provide clear evidence of a benefit for counselling in depression (for example, Stiles et al., 2008; Marriott & Kellett, 2009). The evidence base for counselling in contrast to that for both CBT and IPT lacked data on both long-term follow-up and relapse prevention. The previous guideline recommended counselling in mild to moderate depression, but in light of the increased evidence for a range of low-intensity interventions and group CBT for mild to moderate depression, the GDG decided not to support the same recommendation for counselling in this guideline update. Nevertheless, the GDG thought that counselling may be considered for people with mild to moderate depression who have declined an antidepressant, CBT, IPT, behavioural activation or behavioural couples therapy, but felt that the limited evidence should be drawn to the attention of the healthcare professional. There was considerable discussion of this recommendation in the GDG, which took into account not only the limited evidence for counselling but also the increased evidence for other interventions, such as CBT, IPT, behavioural activation and behavioural couples therapy.

8.10.7. Short-term psychodynamic psychotherapy

A number of new studies were identified for short-term psychodynamic psychotherapy as a treatment for depression, taking the total number of included studies to ten. The comparators were very varied and so significantly limited the amount of meta-analysis that was possible. Nevertheless, from a review of these studies it was not possible to demonstrate a consistent picture of any clinically important benefit for short-term psychodynamic psychotherapy in depression. For example, the two comparisons with antidepressants revealed directly contradictory results (DEKKER2008, SALMINEN2008) and some of the largest effects were obtained in dysthymic or subthreshold populations (MAINA2005). Two studies identified in the health economics literature suggested no advantage, in terms of cost effectiveness, of short-term psychodynamic psychotherapy compared with usual care for primary care patients with depression (Simpson et al., 2003) or high utilisers of psychiatric services with a significant number of patients with a diagnosis of depression (Guthrie et al., 1999). The previous guideline recommended psychodynamic psychotherapy for complex comorbidities, but the current dataset offered no clear evidence for the effectiveness of short-term psychodynamic psychotherapy for complex comorbidities. As a result of the limited evidence for short-term psychodynamic psychotherapy for depression with or without complex comorbidities, the GDG did not feel able to endorse the recommendation in the previous guideline and developed a more specific recommendation for this update. Results from the KOOL2003 study, which included a subgroup analysis of those with a personality disorder, lacked the power to inform a decision on the use of short-term psychodynamic psychotherapy with comorbidities such as personality disorder. As with the evidence base for counselling, the short-term psychodynamic psychotherapy evidence base lacked data on both long-term follow-up and relapse prevention. Nevertheless, the GDG took the view that short-term psychodynamic psychotherapy may be considered for people with mild to moderate depression who have declined an antidepressant, CBT, IPT, behavioural activation or behavioural couples therapy, but that the limited evidence should be drawn to the attention of the healthcare professional. There was considerable discussion of this recommendation in the GDG which took into account not only the limited evidence for short-term psychodynamic psychotherapy but also the increased evidence for other interventions such as CBT, IPT, behavioural activation and behavioural couples therapy.

8.11. RECOMMENDATIONS

8.11.1. Effective delivery of interventions for depression

8.11.1.1.

All interventions for depression should be delivered by competent practitioners. Psychological and psychosocial interventions should be based on the relevant treatment manual(s), which should guide the structure and duration of the intervention. Practitioners should consider using competence frameworks developed from the relevant treatment manual(s) and for all interventions should:

  • receive regular high-quality supervision
  • use routine outcome measures and ensure that the person with depression is involved in reviewing the efficacy of the treatment
  • engage in monitoring and evaluation of treatment adherence and practitioner competence - for example, by using video and audio tapes, and external audit and scrutiny where appropriate.

8.11.2. Group cognitive behavioural therapy

8.11.2.1.

Consider group-based CBT for people with persistent subthreshold depressive symptoms or mild to moderate depression who decline low-intensity psychosocial interventions (see 7.5.1.1).

8.11.2.2.

Group-based CBT for people with persistent subthreshold depressive symptoms or mild to moderate depression should:

  • be based on a structured model such as ‘Coping with Depression’
  • be delivered by two trained and competent practitioners
  • consist of ten to 12 meetings of eight to ten participants
  • normally take place over 12 to 16 weeks, including follow-up.

8.11.3. Treatment options

8.11.3.1.

For people with persistent subthreshold depressive symptoms or mild to moderate depression who have not benefited from a low-intensity psychosocial intervention, discuss the relative merits of different interventions with the person and provide:

  • an antidepressant (normally a selective serotonin reuptake inhibitor [SSRI]) or
  • a high-intensity psychological intervention, normally one of the following options:
    -

    CBT

    -

    interpersonal therapy (IPT)

    -

    behavioural activation (but note that the evidence is less robust than for CBT or IPT)

    -

    behavioural couples therapy for people who have a regular partner and where the relationship may contribute to the development or maintenance of depression, or where involving the partner is considered to be of potential therapeutic benefit70.

8.11.3.2.

For people with moderate or severe depression, provide a combination of antidepressant medication and a high-intensity psychological intervention (CBT or IPT)71.

8.11.3.3.

The choice of intervention should be influenced by the:

  • duration of the episode of depression and the trajectory of symptoms
  • previous course of depression and response to treatment
  • likelihood of adherence to treatment and any potential adverse effects
  • person’s treatment preference and priorities72.
8.11.3.4.

For people with depression who decline an antidepressant, CBT, IPT, behavioural activation and behavioural couples therapy, consider:

  • counselling for people with persistent subthreshold depressive symptoms or mild to moderate depression
  • short-term psychodynamic psychotherapy for people with mild to moderate depression.

Discuss with the person the uncertainty of the effectiveness of counselling and psychodynamic psychotherapy in treating depression.

8.11.4. Delivering high-intensity psychological interventions

8.11.4.1.

For all high-intensity psychological interventions, the duration of treatment should normally be within the limits indicated in this guideline. As the aim of treatment is to obtain significant improvement or remission the duration of treatment may be:

  • reduced if remission has been achieved
  • increased if progress is being made, and there is agreement between the practitioner and the person with depression that further sessions would be beneficial (for example, if there is a comorbid personality disorder or significant psychosocial factors that impact on the person’s ability to benefit from treatment).
8.11.4.2.

For all people with depression having individual CBT, the duration of treatment should typically be in the range of 16 to 20 sessions over 3 to 4 months. Also consider providing:

  • two sessions per week for the first 2 to 3 weeks of treatment for people with moderate or severe depression
  • follow-up sessions typically consisting of three to four sessions over the following 3 to 6 months for all people with depression.
8.11.4.3.

For all people with depression having IPT, the duration of treatment should typically be in the range of 16 to 20 sessions over 3 to 4 months. For people with severe depression, consider providing two sessions per week for the first 2 to 3 weeks of treatment.

8.11.4.4.

For all people with depression having behavioural activation, the duration of treatment should typically be in the range of 16 to 20 sessions over 3 to 4 months. Also consider providing:

  • two sessions per week for the first 3 to 4 weeks of treatment for people with moderate or severe depression
  • follow-up sessions typically consisting of three to four sessions over the following 3 to 6 months for all people with depression.
8.11.4.5.

Behavioural couples therapy for depression should normally be based on behavioural principles, and an adequate course of therapy should be 15 to 20 sessions over 5 to 6 months.

8.11.5. Delivering counselling

8.11.5.1.

For all people with persistent subthreshold depressive symptoms or mild to moderate depression having counselling, the duration of treatment should typically be in the range of six to ten sessions over 8 to 12 weeks.

8.11.6. Delivering short-term psychodynamic psychotherapy

8.11.6.1.

For all people with mild to moderate depression having short-term psychodynamic psychotherapy, the duration of treatment should typically be in the range of 16 to 20 sessions over 4 to 6 months.

8.11.7. Combined psychological and drug treatment

8.11.7.1.

For a person whose depression has not responded to either pharmacological or psychological interventions, consider combining antidepressant medication with CBT73.

8.11.8. Psychological interventions for relapse prevention

8.11.8.1.

People with depression who are considered to be at significant risk of relapse (including those who have relapsed despite antidepressant treatment or who are unable or choose not to continue antidepressant treatment) or who have residual symptoms, should be offered the following psychological interventions:

  • individual CBT for people who have relapsed despite antidepressant medication and for people with a significant history of depression and residual symptoms despite treatment
  • mindfulness-based cognitive therapy for people who are currently well but have experienced three or more previous episodes of depression.

8.11.9. Delivering psychological interventions for relapse prevention

8.11.9.1.

For all people with depression who are having individual CBT for relapse prevention, the duration of treatment should typically be in the range of 16 to 20 sessions over 3 to 4 months. If the duration of treatment needs to be extended to achieve remission it should:

  • consist of two sessions per week for the first 2 to 3 weeks of treatment
  • include additional follow-up sessions, typically consisting of four to six sessions over the following 6 months.
8.11.9.2.

Mindfulness-based cognitive therapy should normally be delivered in groups of eight to 15 participants and consist of weekly 2-hour meetings over 8 weeks and four follow-up sessions in the 12 months after the end of treatment.

8.12. RESEARCH RECOMMENDATIONS

8.12.1.1.

The efficacy of short-term psychodynamic psychotherapy compared with CBT and antidepressants in the treatment of moderate to severe depression.

In well-defined depression of moderate to severe severity, what is the efficacy of short-term psychodynamic psychotherapy compared with CBT and antidepressants?

Why this is important

Psychological treatments are an important therapeutic option for people with depression. CBT has the best evidence base for efficacy but it is not effective for everyone. The availability of alternatives drawing from a different theoretical model is therefore important. Psychotherapy based on psychodynamic principles has historically been provided in the NHS but provision is patchy and a good evidence base is lacking. It is therefore important to establish whether short-term psychodynamic psychotherapy is an effective alternative to CBT and one that should be provided. The results of this study will have important implications for the provision of psychological treatment in the NHS.

This question should be answered using a randomised controlled trial design that reports short- and medium-term outcomes (including cost-effectiveness outcomes) of at least 18 months’ duration. Particular attention should be paid to the reproducibility of the treatment model and training and supervision of those providing interventions to ensure that the treatments are both robust and generalisable. The outcomes chosen should reflect both observer- and patient-rated assessments of improvement and an assessment of the acceptability of the treatment options. The study needs to be large enough to determine the presence or absence of clinically important effects using a non-inferiority design, and mediators and moderators of response should be investigated.

8.12.1.2.

The cost effectiveness of combined antidepressants and CBT compared with sequenced treatment for moderate to severe depression

What is the cost effectiveness of combined antidepressants and CBT compared with sequenced medication followed by CBT and vice versa for moderate to severe depression?

Why this is important

There is a reasonable evidence base for the superior effectiveness of combined anti-depressants and CBT over either treatment alone in moderate to severe depression. However the practicality, acceptability and cost effectiveness of combined treatment over a sequenced approach is less well-established. The answer has important practical implications for service delivery and resource implications for the NHS.

This question should be answered using a randomised controlled trial design in which people with moderate to severe depression receive either combined treatment from the outset, or single modality treatment with the addition of the other modality if there is inadequate response to initial treatment. The outcomes chosen should reflect both observer and patient-rated assessments for acute and medium-term outcomes to at least 6 months, and an assessment of the acceptability and burden of the treatment options. The study needs to be large enough to determine the presence or absence of clinically important effects using a non-inferiority design together with robust health economic measures.

8.12.1.3.

The efficacy of CBT compared with antidepressants and placebo for persistent subthreshold depressive symptoms

What is the efficacy of CBT compared with antidepressants and placebo for persistent subthreshold depressive symptoms?

Why this is important

Persistent subthreshold depressive symptoms are increasingly recognised as affecting a considerable number of people and causing significant suffering, but the best way to treat them is not known. There are studies of the efficacy of antidepressants for dysthymia (persistent subthreshold depressive symptoms that have lasted for at least 2 years) but there is a lack of evidence for CBT. Subthreshold depressive symptoms of recent onset tend to improve but how long practitioners should wait before offering medication or psychological treatment is not known. This research recommendation is aimed at informing the treatment options available for this group of people with subthreshold depressive symptoms that persist despite low-intensity interventions.

This question should be answered using a randomised controlled trial design that reports short- and medium-term outcomes (including cost-effectiveness outcomes) of at least 6 months’ duration. A careful definition of persistence should be used which needs to include duration of symptoms and consideration of failure of low-intensity interventions and does not necessarily imply a full diagnosis of dysthymia. The outcomes chosen should reflect both observer and patient-rated assessments of improvement, and an assessment of the acceptability of the treatment options. The study needs to be large enough to determine the presence or absence of clinically important effects using a non-inferiority design, and mediators and moderators of response should be investigated.

8.12.1.4.

The efficacy of counselling compared with low-intensity cognitive behavioural interventions and treatment as usual in the treatment of persistent subthreshold depressive symptoms and mild depression

In persistent subthreshold depressive symptoms and mild depression, what is the efficacy of counselling compared with low-intensity cognitive behavioural interventions?

Why this is important

Psychological treatments are an important therapeutic option for people with subthreshold symptoms and mild depression. Low-intensity cognitive behavioural interventions have the best evidence base for efficacy but the evidence is limited and longer-term outcomes are uncertain, as are the outcomes for counselling. It is therefore important to establish whether either of these interventions is an effective alternative to treatment as usual and should be provided in the NHS. The results of this study will have important implications for the provision of psychological treatment in the NHS.

This question should be answered using a randomised controlled trial design which reports short-term and medium-term outcomes (including cost-effectiveness outcomes) of at least 18 months’ duration. Particular attention should be paid to the reproducibility of the treatment model and training and supervision of those providing interventions in order to ensure that the treatments are both robust and generalisable. The outcomes chosen should reflect both observer and patient-rated assessments of improvement and an assessment of the acceptability of the treatment options. The study needs to be large enough to determine the presence or absence of clinically important effects using a non-inferiority design, and mediators and moderators of response should be investigated.

8.12.1.5.

The efficacy of behavioural activation compared with CBT and antidepressants in the treatment of moderate to severe depression

In well-defined depression of moderate to severe severity, what is the efficacy of behavioural activation compared with CBT and antidepressants?

Why this is important

Psychological treatments are an important therapeutic option for people with depression. Behavioural activation is a promising treatment but does not have the substantial evidence base that CBT has. The availability of alternatives drawing from a different theoretical model is important because outcomes are modest even with the best supported treatments. It is therefore important to establish whether behavioural activation is an effective alternative to CBT and one that should be provided. The results of this study will have important implications for the provision of psychological treatment in the NHS.

This question should be answered using a randomised controlled trial design which reports short-term and medium-term outcomes (including cost-effectiveness outcomes) of at least 18 months’ duration. Particular attention should be paid to the reproducibility of the treatment model and training and supervision of those providing interventions in order to ensure that the treatments are both robust and generalisable. The outcomes chosen should reflect both observer and patient-rated assessments of improvement, and an assessment of the acceptability of the treatment options. The study needs to be large enough to determine the presence or absence of clinically important effects using a non-inferiority design, and mediators and moderators of response should be investigated.

8.12.1.6.

The efficacy and cost effectiveness of cognitive behavioural therapy, interpersonal therapy and antidepressants in prevention of relapse in people with moderate to severe recurrent depression

In people with moderate to severe recurrent depression, what is the relative efficacy of CBT, IPT and antidepressants in preventing relapse?

Why this is important

Psychological and pharmacological treatments are important therapeutic options for people with depression, but evidence on the prevention of relapse (especially for psychological interventions) is limited. All of these treatments have shown promise in reducing relapse but the relapse rate remains high. New developments in the style and delivery of CBT and IPT show some promise in reducing relapse but need to be tested in a large-scale trial. The results of this study will have important implications for the provision of psychological treatment in the NHS.

This question should be answered using a randomised controlled trial design which reports short-term and medium-term outcomes (including cost-effectiveness outcomes) of at least 24 months’ duration. Particular attention should be paid to the development and evaluation of CBT, IPT and medication interventions tailored specifically to prevent relapse, including the nature and duration of the intervention. The outcomes chosen should reflect both observer and patient-rated assessments of improvement and an assessment of the acceptability of the treatment options. The study needs to be large enough to determine the presence or absence of clinically important effects using a non-inferiority design, and mediators (including the focus of the interventions) and moderators (including the severity of the depression) of response should be investigated.

Five out of six of the included studies of couples therapy were based on a behavioural model.

Here and elsewhere in the guideline, each study considered for review is referred to by a ‘study ID’ made up of first author and publication date (unless a study is in press or only submitted for publication, when first author only is used). Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18 and references for studies for the update are in Appendix 17b.

Follow-up to Fava1994 (study in the previous guideline).

For reasons of brevity this analysis is not included in the summary evidence table, but can be found in Appendix 16b and 19b. The study characteristics are in Appendix 17b.

Ibid.

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

The review team contacted the authors of the study but did not receive the data.

Cognitive therapy based on Beck and colleagues (1979) and Emery (1981) and compared with behaviour therapy based on Lewinsohn (1975).

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

Note that this study was excluded from the analysis of clinical effectiveness as only 50% might have met diagnostic criteria for depression.

This is the economic analysis of Bedi2000.

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

Each study considered for review is referred to by a study ID in capital letters (primary author and date of study publication, except where a study is in press or only submitted for publication, then a date is not used).

This recommendation also appears in the chapter on pharmacological interventions.

Ibid.

Ibid.

This recommendation can also be found in the chapter on pharmacological interventions.

Footnotes

54

Five out of six of the included studies of couples therapy were based on a behavioural model.

55

Here and elsewhere in the guideline, each study considered for review is referred to by a ‘study ID’ made up of first author and publication date (unless a study is in press or only submitted for publication, when first author only is used). Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18 and references for studies for the update are in Appendix 17b.

56

Follow-up to Fava1994 (study in the previous guideline).

57

For reasons of brevity this analysis is not included in the summary evidence table, but can be found in Appendix 16b and 19b. The study characteristics are in Appendix 17b.

58

Ibid.

59

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

60

The review team contacted the authors of the study but did not receive the data.

61

Cognitive therapy based on Beck and colleagues (1979) and Emery (1981) and compared with behaviour therapy based on Lewinsohn (1975).

62

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

63

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

64

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

65

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

66

Note that this study was excluded from the analysis of clinical effectiveness as only 50% might have met diagnostic criteria for depression.

67

This is the economic analysis of Bedi2000.

68

Study IDs in title case refer to studies included in the previous guideline and study IDs in capital letters refer to studies found and included in this guideline update. References for studies from the previous guideline are in Appendix 18.

69

Each study considered for review is referred to by a study ID in capital letters (primary author and date of study publication, except where a study is in press or only submitted for publication, then a date is not used).

70

This recommendation also appears in the chapter on pharmacological interventions.

71

Ibid.

72

Ibid.

73

This recommendation can also be found in the chapter on pharmacological interventions.

Copyright © The British Psychological Society & The Royal College of Psychiatrists, 2010.

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Enquiries in this regard should be directed to the British Psychological Society.

Cover of Depression
Depression: The Treatment and Management of Depression in Adults (Updated Edition).
NICE Clinical Guidelines, No. 90.
National Collaborating Centre for Mental Health (UK).
Leicester (UK): British Psychological Society; 2010.

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...