Statistics from Altmetric.com
Correspondence to: Norio Watanabe, Department of Psychiatry and Cognitive Behavioral Medicine, Nagoya City University Graduate School of Medical Sciences, Mizuho-cho, Mizuho-ku Nagoya 467-8601, Japan;
Is psychotherapy an effective way to treat children and adolescents with depression?
Risk of response, defined as score below the threshold for diagnosis of depression on whichever scale the study used—“operationalized criteria” or “a validated depression severity measure”. Secondary outcomes were cost and safety of treatment.
Systematic review with meta-analysis.
Cochrane CENTRAL, MEDLINE, CINAHL, PsycINFO, EMBASE, PSYNDEX, LILACS, conference proceedings and hand searches of journals; performed by searching the Cochrane Collaboration Depression, Anxiety and Neurosis Registers on 17 December 2004. Additional supplementary search of CENTRAL, MEDLINE, PsycINFO and EMBASE was carried out, and from the selected papers, reference lists were examined and lead authors contacted for further data on other trials.
Study selection and analysis:
Individual or cluster randomised controlled trials of any psychotherapy (PT) versus no treatment, attention-placebo, waiting-list control, or treatment as usual, in adolescents (aged 6–18 years) with depression or dysthymia. Attention-placebo was defined as interventions that involved time spent with the patient but which did not focus on depressive symptoms (for example, psycho-drama or art therapy). Exclusions: psychotherapy that did not focus specifically on the individual—for example, family therapy, counselling. Standardised data extraction forms were used and methodological quality assessed using the Quality Rating Scale. Results from individual studies were combined and number needed to treat and relative risk of response (primary outcome) were calculated using a random effects model. Intracluster correlation coefficients were used, where available, to estimate effect size in cluster randomised trials. Subgroup analysis was planned a priori to assess the effect of psychotherapy type, control condition used, severity of depression, and age. Dropouts were included in intention-to-treat analysis (ITT).
Twenty seven studies involving 35 randomised comparisons (3 cluster) and 1744 patients, met the inclusion criteria. Psychotherapy (25 trials investigating cognitive behavioural therapy (CBT), 2 cognitive therapy (CT), 3 behavioural therapy (BT), 2 interpersonal therapy (IPT), 1 problem solving therapy (PST), 1 supportive therapy (ST), and 1 other) were compared to treatment as usual (TAU; used in 3 comparisons and included any non-study mental health care), attention-placebo (8 comparisons; including health education and psychodrama), waiting list control (18 comparisons), and no treatment (6 comparisons). Psychotherapy improved post-treatment response compared to control overall: 450/907 (49.6%) of the PT group had responded at study completion compared to 267/767 (34.8%) of the control groups (RR response: 1.39, 95% CI 1.18 to 1.65; p = 0.0001). Estimated NNT was 4.3. In analyses specifically looking at 1–6 month follow-up and 6–12 month follow-up, there were no differences between psychotherapy and the control conditions (p>0.15 for both). For subgroup analyses see online notes.
Psychotherapy improves post-treatment response for children and adolescents with depression compared to control, with most evidence for CBT. However, adding PT to usual treatment for depression does not improve response above usual treatment alone. Benefit is greatest for those in the 12–18 age group. Benefit over control is not sustained at longer-term follow-up.
Watanabe N, Hunot V, Omori IM, et al. Psychotherapy for depression among children and adolescents: a systematic review. Acta Psychiatr Scand 2007;116:84–95.
This rigorous review addressed the issue of the effectiveness of psychotherapy in the treatment of depression in children and adolescents. It is worth noting that the review includes a particularly large study on targeted depression prevention1 which excluded young people with depressive disorder, thus reducing the chance of showing an effect for psychotherapy.
The results provide evidence that cognitive behaviour therapy (CBT), interpersonal therapy (IPT) and possibly behaviour therapy (BT) are effective, at least for adolescents. The response rates on meta-analysis (49.6% psychotherapy vs 34.8% control) compare well with antidepressants.2 The psychotherapy studies have used self-rating scales of depressive symptoms. Studies of antidepressants have used clinician rated scales as primary outcomes and most have failed to show a difference on self-rating scales (using similar measures).2 Had clinician rating scales been used the results from psychotherapy might have been even better.
A difference in outcome at short-term follow-up (1–6 months) would have been expected with effective therapy, and failure to show this is of some concern, while the failure to show a difference at longer term follow-up (6–12 months) may well reflect a catch-up from spontaneous remission in the comparison group.
The authors rightly identify weaknesses in the current studies and point out the lack of information on cost-effectiveness and on the potential for harm. However, until more rigorous studies are done this review supports the continued use of CBT and IPT, and probably BT for depression in children and adolescents.
The challenge remains. Neither psychotherapy nor medication2 result in high rates of remission for depressive disorder. The quest for more effective therapies goes on. In the meantime, this review provides evidence to support the continued use of CBT and IPT in depression in adolescents, in line with the NICE guidelines.3 The issue is less clear for children.
web only references 11/2/49
Subgroup analyses of specific types of psychotherapy models showed significant post-treatment benefit of PT over control for BT (RR response: 6.76, 95% CI 1.45 to 31.40; p=0.01), IPT (RR response: 1.68, 95% CI 1.08 to 2.63; p=0.02), and CBT (RR response: 1.38, 95% CI 1.14 to 1.66; p=0.0009). At 1�6 month follow-up and 6�12 month follow-ups, there were no differences in response between different models of PT and control conditions. Subgroup analyses by control condition revealed significant post-treatment benefit of PT over attention-placebo (RR response: 1.48, 95% CI 1.12 to 1.96; p=0.006). This benefit was not evident at 1�6 month or 6�12 month follow-ups. There was a significant post-treatment benefit of PT over waiting list control (RR response: 2.00, 95% CI 1.34 to 2.98; p=0.0006), and also at 1�6 months (RR response: 1.98, 95% CI 1.27 to 3.07). There was no difference in response between PT (combined with TAU) and TAU alone, or PT and no treatment at any time point. PT improved post-treatment response in adolescents (RR response: 1.35, 95% CI 1.10 to 1.66; p=0.004), but not among children (6�12 years). PT improved response regardless of depression severity (mild to moderate or moderate to severe).
However, the analyses did not correct for the multiple subgroup analyses. This may have led to false-positive conclusions.
Most of the trials examined adolescents aged 12�18 years (81%), and were in people with mild to moderate depression (79%); 64% of participants were female. Follow-up ranged from between 1�6 months (14 studies) and 6�12 months (8 studies). Trials varied in methodological quality and diagnostic criteria used for diagnosing depression. Heterogeneity between studies was significant (p=0.04) and publication bias was significant (p <_0.001. _--="_--" end="end" desc="desc" dc1="dc1">
Files in this Data Supplement:
Source of funding: The Cochrane Child Health Field Bursay Scheme.
▸ Additional notes and a reference list are published online only at http://ebmh.bmj.com/content/vol11/issue2
Competing interests: None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.