U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Mental Health (UK). Social Anxiety Disorder: Recognition, Assessment and Treatment. Leicester (UK): British Psychological Society (UK); 2013. (NICE Clinical Guidelines, No. 159.)

Cover of Social Anxiety Disorder

Social Anxiety Disorder: Recognition, Assessment and Treatment.

Show details

7INTERVENTIONS FOR CHILDREN AND YOUNG PEOPLE

7.1. INTRODUCTION

In Chapter 5 the problems of case identification were discussed and the significant under-recognition of social anxiety disorder in children and young people was noted. This is a cause of considerable concern as social anxiety disorder usually starts in late childhood or early adolescence. As a consequence of under-recognition many children and young people with social anxiety disorder often only access treatment years after the onset of symptoms and a referral for early help from child and adolescent mental health services (CAMHS) is relatively rare. In addition, social anxiety disorder may evade identification in children and young people known to specialist CAMHS, its presence being overshadowed by more noticeable comorbid problems. Although effective interventions, in particular psychological interventions, have been identified for the treatment of social anxiety disorder, access to such interventions even for those in the care of CAMHS has been limited. In 2011 the English Department of Health launched an IAPT programme for children and young people26, which has some similarities to the IAPT adult programme (see Chapter 6) but is focused more on the transformation of the existing services rather than the training of a new cadre of psychological therapists. The initial focus of the child IAPT programme is on CBT interventions for depression and anxiety disorders and social learning-based programmes for parent training.

7.1.1. Pharmacological interventions

Pharmacological interventions to manage social anxiety disorder are used infrequently in CAMHS. In part this is because children and young people with social anxiety disorder are rarely treated in CAMHS (see Chapter 2) and because, as with all other anxiety disorders in children and young people, psychological interventions are accepted as first-line treatment for social anxiety disorder. However, if medication is used then it is usually with SSRIs, which are increasingly being prescribed in the management of other anxiety disorders, after non-response to psychological interventions, particularly where there is comorbid depression. All such prescribing is in the context of the MHRA statement27 regarding the balance of risks and benefits of using SSRIs in the treatment of depression in children and young people and caution in the prescribing of SSRIs is now widespread, particularly among GPs. Some potential prescribers are deterred by the concerns about the possible effects of SSRIs on the developing brain, some worry that suicidality and impulsivity may be precipitated in those without a previous history of this problem, while others are rather less concerned believing that the risk of precipitating self-harm is reduced if there is no comorbid depression, although the precise mechanism for the increase in suicidality with SSRIs in children and young people is not understood.

Children and young people with social anxiety disorder together with their parents and psychiatrist might decide against the option of an SSRI after reviewing their potential side effects, some of which might be particularly troublesome for people with social anxiety disorder. With respect to licensing considerations, while some SSRIs are licensed in those under 18 years of age (fluvoxamine for the treatment of OCD in children and young people aged over 8 years and sertraline for the treatment of OCD in children and young people aged over 6 years), none of the SSRIs are licensed for the treatment of social anxiety disorder for under 18s. Other SSRIs (for example, paroxetine) are not licensed for use in children in either the UK or USA for any conditions and do not feature in the BNF for children (Paediatric Formulary Committee, 2012–2013) as an unlicensed option.

Beta-adrenoceptor blocking drugs are sometimes considered as an option by psychiatrists in CAMHS and by GPs. In young people, these drugs could be seen as a safer option than SSRIs (once asthma has been excluded), although as can be seen in Chapter 6 the evidence in adults for their efficacy is limited. As in other anxiety disorders in children and young people, the doses of beta-blockers prescribed rarely have a significant impact on the impressive attempts of the body to protect itself in a situation of perceived threat. For this reason the results of beta-blocker use in social anxiety disorder are often disappointing, but nevertheless they continue to be tried periodically especially when a young person's preference is for a pharmacological option to help alleviate or fractionally reduce their symptoms.

Benzodiazepines are not used and, while the BNF for children (Paediatric Formulary Committee, 2012–2013) does indicate that antipsychotics have a possible place in the short term for ‘severe anxiety’, they do not feature in the current management of social anxiety disorder in CAMHS. Other agents described in Chapter 6 do not have evidence specifically targeted to children and young people.

7.1.2. Psychological interventions

A range of psychological interventions can be offered in CAMHs including CBT, systemic therapy (including family interventions), parenting interventions, counselling and psychodynamic therapy. The past 30 years have seen significant shifts in the provision of psychological interventions with the nature of the therapy moving away from psychodynamic to systemic approaches, to some extent, and more recently to cognitive behavioural interventions. There has been relatively little formal evaluation of these interventions until recent times but the last 20 years have seen a large expansion in RCT-based evidence particularly in the area of conduct disorder and oppositional defiant disorder.

To date various forms of CBT (individual, group or parent-delivered) are the only psychological interventions that have been evaluated within RCTs including children and young people with social anxiety disorder. Because of the high level of comorbidity between different anxiety disorders in children and young people, those with a principal primary social anxiety disorder have most commonly been included among groups of children and young people with other principal diagnoses (such as generalised anxiety disorder and separation anxiety disorder) in programmes that take a general cognitive behavioural approach to the treatment of anxiety disorders. In these programmes children and young people will be assisted in applying general cognitive and behavioural principles to the area that causes them greatest concern or impairment. Typically these studies have not included a sufficient number of participants to compare outcomes for children and young people with different principal anxiety diagnoses. Overall, they find that CBT is efficacious for anxiety disorders in children, but the evidence, reviewed by Cochrane (James et al., 2013), is insufficient to compare it with other therapies. Moreover, recent reports have suggested that children and young people with social anxiety disorder may have poorer outcomes (Hudson et al., 2010) or may not demonstrate equivalent gains beyond the end of treatment (Kerns et al., 2012) from these general interventions, compared with children and young people with other anxiety disorders.

Although there is variability in the particular procedures used in different manualised treatments for children and young people, the content of these interventions are broadly similar to adult CBT programmes, with most programmes (both for general anxiety and specifically for social anxiety disorder) involving exposure in vivo and cognitive restructuring. However, many of the programmes that have been developed specifically for children and young people with social anxiety disorder include social skills.

One other key factor that also distinguishes some programmes developed for children and young people from adult-oriented programmes is the involvement of parents to support treatment. The extent and manner of parental involvement varies across different treatments programmes. In some treatments parents are not included at all, at the other end of the scale treatment is delivered entirely via parents. The most common ways in which parents are involved in treatment are as follows:

  • Parent-education (for example, Beidel and colleagues [2000; 2007]): the parent is provided with information about the nature of social anxiety disorder and the focus of the programme in which their child is participating.
  • Parent-support (for example, March and colleagues [2009], Spence and colleagues [2000; 2011]): the parent attends sessions in parallel with the sessions for the child or young person. The sessions aim to teach parents to model, encourage and prompt the use of new skills, and manage socially anxious behaviour and avoidance, using instruction, discussion, modelling and role play.
  • Parent-led CBT (for example, Cartwright-Hatton and colleagues [2011], Lyneham and colleagues [2013], Rapee and colleagues [2006], Thirlwall and colleagues [2012]): this approach has been evaluated with pre-adolescents, either in a parent-group format or as a low-intensity treatment in which the parent is supported in working through a ‘self-help’ book. The child does not attend the treatment sessions at all, but the parent is taught skills for helping their child manage anxious thoughts and alter avoidant behaviour, given the opportunity to rehearse with a therapist and to problem solve difficulties that arise.
  • Therapeutic input for parents in their own right, for example, parent anxiety management (for example, Hudson and colleagues [2013]).

7.2. CLINICAL REVIEW PROTOCOL

A systematic review to identify RCTs of interventions for children and young people with social anxiety disorder was conducted. The review protocol, including the review questions, information about the databases searched, and the eligibility criteria used for this section of the guideline, is presented in Table 21. The first systematic evaluation of a programme to specifically target social anxiety disorder in children and young people was only published as recently as 2000 (Spence et al., 2000). This review therefore considers outcomes for children and young people with social anxiety disorder from both treatments aimed specifically at social anxiety disorder and generic anxiety treatments. From studies randomising people with social anxiety disorder and other mental disorders, only data for children and young people with social anxiety disorder, which were requested from authors when it was not available in published reports, were included in the analysis in this guideline. Only measures of social anxiety or subscales relating to social anxiety were analysed. Parts of these questions were addressed in Cochrane reviews, but the searches were up to 8 years old and all needed to be updated. Further details are included Appendix 6 (the complete search strategy and PRISMA28 chart), Appendix 16 (study characteristics), Appendix 17 (forest plots) and Appendix 19 (GRADE profiles). In the sections that follow, the number of participants reported is the number receiving treatment included in the analysis. Studies that were excluded from the analysis and reasons for exclusion are included in Appendix 25.

Table 21. Clinical review protocol for the interventions for children and young people.

Table 21

Clinical review protocol for the interventions for children and young people.

7.2.1. Extrapolation

The GDG took the view that with limited primary data of good quality (for example, RCTs) for children and young people with social anxiety disorder, it might be necessary to extrapolate from other populations.

For psychological interventions, the decision was made to extrapolate from the data for adult interventions for social anxiety disorder to mature adolescents. Extrapolation was performed on the basis that the extrapolated population shared common characteristics with the primary population (for example, some young people who have the cognitive and emotional capacity are able to describe their thoughts and feelings much like adults), where the harms were similar for the extrapolated dataset as for the primary dataset, and where the outcomes were comparable across trials. Extrapolated data were recognised as lower-quality evidence than data from studies on children and young people with social anxiety disorder.

7.3. OVERVIEW OF CLINICAL EVIDENCE

The search identified 23 RCTs that included children and young people with social anxiety disorder, including trials of interventions for all anxiety disorders that provided disaggregated data; four were unpublished and 19 were published in peer-reviewed journals between 1994 and 2012. Of these, 22 RCTs were included in at least one analysis; the remaining trial (BAER2005 [Baer & Garland, 2005]) merged groups for analysis and it was not possible to analyse the results of the trial (see Table 22 for a summary).

Table 22. Summary of results at post-treatment.

Table 22

Summary of results at post-treatment.

Meta-analyses were conducted for classes of interventions. For all classes, subgroup analyses were conducted to explore differences between members of the class (for example, different drugs or variations of a therapy). For each comparison, recovery (clinician-rated) and symptoms of anxiety were analysed. Symptom ratings by the child or young person and the parent were analysed separately. Analyses of secondary outcomes were not conducted to reduce the risk of spurious findings as the review includes many comparisons and very few studies.

The study characteristics can be found in Appendix 16, forest plots in Appendix 17 and GRADE evidence profiles in Appendix 19.

7.3.1. Study characteristics

Trials included between 15 and 322 participants at baseline (median 73), but many of these participants were not eligible for this review; that is, authors of several published studies that included children and young people with mixed anxiety disorders provided data for the subgroup of children with social anxiety disorder. Included trials randomised approximately 2,467 participants but only 1,194 are included in this review. Most of this difference results from the exclusion of participants who did not have social anxiety disorder and therefore were not eligible for inclusion rather than because of missing data.

Participants were on average (median of means) 11 years old, ranging from 4 to 21 years old. Approximately 77% were white. About half of the included participants were female (55%). Some participants were taking medication at baseline in two trials (HERBERT2009, RAPEE2006), and it was unclear in 11 studies if any participants were taking medication at baseline.

7.3.2. Risk of bias

All included trials were assessed for risk of bias (see Figure 9 and Appendix 20). Thirteen were at low risk for sequence generation and 11 of these were at low risk of bias for allocation concealment. Allocation concealment was unclear in ten trials, and one trial was at high risk of bias. Trials of psychological therapies were considered at high risk of bias for participant and provider blinding per se; three trials were at low risk of bias for blinding participants and providers, although the rate of side effects may make it difficult to maintain blinding in pharmacological trials. Most reported outcomes were self-rated, but assessor blinding was considered separately for all trials, and all were at low risk of bias (no assessor rated outcomes or assessors blind). For incomplete outcome data, 18 trials were at low risk of bias and four trials were at high risk of bias (for example, those that reported per protocol or completer analyses and those with very large amounts of missing data).

Figure 9. Risk of bias summary.

Figure 9

Risk of bias summary.

Selective outcome reporting and publication bias

Several methods were employed to minimise risk of selective outcome reporting and publication bias. All authors were contacted to request trial registrations and unpublished outcomes, and all authors of included studies, all stakeholders, and all pharmaceutical manufacturers were asked to provide unpublished trials. Nonetheless, most of the included studies were not registered. Only eight were at low risk of selective outcome reporting bias; 12 were unclear and one at high risk of bias.

7.4. PHARMACOLOGICAL INTERVENTIONS

7.4.1. Selective serotonin reuptake inhibitors (SSRIs) and serotonin and noradrenaline reuptake inhibitors (SNRIs)

Compared with placebo

Three studies compared an antidepressant with placebo in children with primary social anxiety disorder. One study (DINEEN-WAGNER2004) compared paroxetine (165 participants) with placebo; children (8 to 11 years; DINEEN-WAGNER2004a) received 29 mg daily and young people (12 to 17 years; DINEEN-WAGNER2004b) received 36 mg daily for 16 weeks. One study (MARCH2007) compared venlafaxine (137 participants) with placebo; children and young people (8 to 17 years) received 142 mg of venlafaxine daily for 16 weeks. One study (BEIDEL2007) compared fluoxetine (43 participants) with placebo; children and young people (7 to 17 years) received 30 mg daily for 12 weeks. The mean age of participants in the included studies was 12 to 14 years.

In two studies (BEIDEL2007, DINEEN-WAGNER2004) there was a small effect on clinician-rated recovery at post-treatment (RR = 0.85, 95% CI = 0.78 to 0.92) with no significant heterogeneity between drugs (I2 = 0%, Chi2 = 0.00, p = 0.96). In three studies (BEIDEL2007, DINEEN-WAGNER2004b, MARCH2007), there was a medium effect on self-rated symptoms of social anxiety disorder at post-treatment (SMD = −0.53, 95% CI = −0.69 to −0.36) with no heterogeneity between drugs (I2 = 0%, Chi2 = 1.41, p = 0.50). No controlled effects at follow-up were reported.

DINEEN-WAGNER2004 reported withdrawal from the study because of side effects with paroxetine, for which there was a difference between groups with very wide CIs (RR = 3.09, 95% CI = 0.19 to 50.43). Consistent with results for paroxetine in children and young people with depression (NCCMH, 2005), a GlaxoSmithKline investigation ([Redacted], 15 May 2003, unpublished) identified four ‘suicide-related’ events in the paroxetine group and none in the placebo group.

Compared with CBT

One of the SSRI studies also compared fluoxetine with CBT (BEIDEL2007). At post-treatment, there was a medium effect on recovery favouring CBT (RR = 0.59, 95% CI = 0.44 to 0.79), but the effect was not statistically significant for self-rated symptoms of social anxiety disorder (SMD = 0.15, 95% CI = −0.27 to 0.58).

Compared with placebo for selective mutism

One study (BLACK1994) compared fluoxetine with placebo for children (6 to 12 years) with selective mutism, which may be a specific form of social anxiety disorder. Six participants received 21 mg daily for 12 weeks. At post-treatment, there was a moderate effect on parent-rated symptoms of social anxiety (SMD = −0.74, 95% CI = −1.81 to 0.32).

7.5. PSYCHOLOGICAL INTERVENTIONS

7.5.1. Cognitive behavioural therapy

Compared with waitlist

Eight studies compared CBT with waitlist. These included: individual CBT as a specific treatment for social anxiety disorder (MELFSEN2011); individual CBT as a generic anxiety treatment (SPENCE2011); group CBT specifically for social anxiety disorder (GALLAGHER2004, SANCHEZ-GARCIA2009, SPENCE2000); and group CBT for mixed anxiety disorders (LAU2010, LYNEHAM2012, RAPEE2006). For studies of children and young people with multiple diagnoses, data for those with primary social anxiety disorder were included in the main analysis. A sensitivity analysis included participants with social anxiety disorder as either their primary or secondary diagnosis, thus adding social anxiety outcomes for participants whose social anxiety disorder was not their main problem (see Appendix 17). Treatment lasted 3 to 20 weeks and the group treatments had a mean of six to eight participants. The mean age of participants ranged from 9 to 14 years, and variation in participant age within studies was as great as the variation between them.

All studies reported clinician-rated recovery at post-treatment, and there was a medium effect (RR = 0.65, 95% CI = 0.50 to 0.85) with substantial heterogeneity between studies (I2 = 67%, Chi2 = 21.13, p = 0.004). The subgroups (specific [RR = 0.48, 95% CI = 0.31 to 0.77] and general CBT [RR = 0.65, 95% CI = 0.50 to 0.85]) were different (I2 = 80.2%, Chi2 = 5.04, p = 0.02), but each subgroup contained only four studies with no more than 40 events recorded (see Figure 10). No study reported controlled effects for clinician-rated recovery at follow-up.

Figure 10. Recovery for CBT compared with waitlist.

Figure 10

Recovery for CBT compared with waitlist.

In six studies (all but LAU2010 and SANCHEZ-GARCIA2009), there was a large effect on self-rated symptoms of social anxiety disorder at post-treatment (SMD = −1.20, 95% CI = −1.97 to -0.43), with considerable heterogeneity between studies (I2 = 84%, Chi2 = 44.38, p = 0.00001) but not between specific and general subgroups (I2 = 0%, Chi2 = 0.24, p = 0.62). One study of group CBT specifically for social anxiety disorder (SANCHEZ-GARCIA2009) reported a large effect on self-rated symptoms at follow-up (SMD = −3.08, 95% CI = −3.75 to −2.41).

In two studies (LYNEHAM2012, SPENCE2011), there was a small effect for parent-rated symptoms at post-treatment that was not statistically significant (SMD = −0.29, 95% CI = −0.96 to 0.38) with no heterogeneity. Controlled effects for parent-rated symptoms were not reported at follow-up.

Compared with psychological placebo

Four studies compared CBT with psychological placebo, and one of these had two intervention arms. These included: individual CBT as a specific treatment for social anxiety disorder (HERBERT2009); group CBT as a specific treatment for social anxiety disorder (BEIDEL2000, HERBERT2009, MASIA-WARNER2007); and group CBT for mixed anxiety disorders (HUDSON2009). For studies of children and young people with multiple diagnoses, data for those with primary social anxiety disorder were included in the main analysis. A sensitivity analysis included participants with social anxiety disorder as either their primary or secondary diagnosis (see Appendix 17). Treatment lasted 10 to 12 weeks and the group treatments had a mean of five or six participants per group. The mean age of participants in the included studies was 9 to 15 years, and variation in participant age within studies was as great as the variation between them.

Across all studies, there was a medium effect for clinician-rated recovery at post-treatment that was not statistically significant (RR = 0.72, 95% CI = 0.51 to 1.02), with considerable heterogeneity between studies (I2 = 79%, Chi2 = 19.47, p = 0.0006). A test for subgroup differences was not significant (I2 = 0%, Chi2 = 0.73, p = 0.39), but, again, the largest effect was for CBT designed specifically for social anxiety disorder (see Figure 11). At follow-up, there was a medium effect for recovery that was not statistically significant (RR = 0.79, 95% CI = 0.57 to 1.10), with substantial heterogeneity (I2 = 72%, Chi2 = 10.86, p = 0.01) and no significant differences between subgroups (I2 = 0%, Chi2 = 0.21, p = 0.65).

Figure 11. Recovery for CBT compared with psychological placebo.

Figure 11

Recovery for CBT compared with psychological placebo.

Across all studies, there was a medium effect for self-rated symptoms of social anxiety at post-treatment that was not statistically significant (SMD = −0.56, 95% CI = −1.16 to 0.04), with substantial heterogeneity between studies (I2 = 70%, Chi2 = 13.47, p = 0.009). Subgroups were not significantly different (I2 = 0%, Chi2 = 0.04, p = 0.85). The effect was similar at follow-up (SMD = −0.54, 95% CI = −1.21 to 0.13), with substantial heterogeneity between studies (I2 = 66%, Chi2 = 8.84, p = 0.03) and no significant subgroup differences (I2 = 0%, Chi2 = 0.57, p = 0.45).

In three studies (HERBERT2009, HUDSON2009, MASIA-WARNER2007), the effect was not statistically significant for parent-rated symptoms of social anxiety at post-treatment (SMD = 0.19, 95% CI = −0.18 to 0.56) with no significant heterogeneity between studies or subgroups (I2 = 0%, Chi2 = 0.13, p = 0.72). At follow-up, the effect was not statistically significant (SMD = 0.13, 95% CI = −0.82 to 1.09) with considerable heterogeneity between individual studies (I2 = 83%, Chi2 = 17.91, p = 0.0005), but not between subgroups (I2 = 0%, Chi2 = 0.02, p = 0.90).

Compared with pill placebo

One study (BEIDEL2007) compared CBT with pill placebo. At post-treatment, there was a moderate effect on recovery (RR = 0.51, 95% CI = 0.39 to 0.66) and a small effect for self-rated symptoms of social anxiety disorder that was not statistically significant (SMD = −0.22, 95% CI = −0.66 to 0.21). No controlled effects at follow-up were reported.

Compared with CBT plus parent anxiety management

In one study (HUDSON2012) comparing CBT versus CBT with an intervention to help parents manage their own anxiety, the effect was not statistically significant for recovery at post-treatment (RR = 1.31, 95% CI = 0.41 to 4.20) or at follow-up (RR = 1.23, 95% CI = 0.50 to 3.02). The effect was not statistically significant for self-rated symptoms of social anxiety disorder at post-treatment (SMD = 0.19, 95% CI = −0.48 to 0.87) or at follow-up (SMD = 0.58, 95% CI = −0.16 to 1.31). Similarly, the effect was not statistically significant for parent-rated symptoms of social anxiety disorder at post-treatment (SMD = −0.13, 95% CI = −0.81 to 0.56) or at follow-up (SMD = 0.23, 95% CI = −0.51 to 0.96).

Group CBT compared with group CBT plus individual CBT

One study (OLIVARES2008) compared three groups receiving: (1) group CBT with social skills training; (2) group CBT with 12 individual CBT sessions; and (3) group CBT with six individual sessions. The effect was not statistically significant for recovery at post-treatment comparing group CBT with the addition of 12 individual sessions (RR = 1.37, 95% CI = 0.82 to 2.29) or six individual sessions (RR = 1.20, 95% CI = 0.76 to 1.90). For self-rated symptoms of social anxiety disorder at post-treatment, there was a medium effect compared with the addition of 12 individual sessions that was not statistically significant (SMD = 0.50, 95% CI = −0.16 to 1.16) compared with the addition of six individual sessions (SMD = 0.18, 95% CI = −0.46 to 0.82). The same was true for self-rated symptoms at follow-up compared with the addition of 12 individual sessions (SMD = 0.55, 95% CI = −0.11 to 1.21) and compared with the addition of six individual sessions (SMD = 0.22, 95% CI = −0.42 to 0.86).

7.5.2. Cognitive behavioural therapy delivered via parents

Compared with waitlist

Three studies provided a CBT intervention that parents and carers were instructed to deliver to their children with some therapist support (CARTWRIGHT-HATTON2012, LYNEHAM2012, THIRLWALL2012); one study provided an intervention to be delivered by parents without therapist support (RAPEE2006). For studies of children with multiple diagnoses, data for children and young people with primary social anxiety disorder were included in the main analysis. A sensitivity analysis included participants with social anxiety disorder as either their primary or secondary diagnosis (see Appendix 17). Treatment lasted 10 to 16 weeks. The mean age of participants in the included studies was 7 to 10 years, and variation in participant age within studies was as great as the variation between them. For the supported interventions, parents received approximately 8 to 20 hours of therapist contact.

Across all studies, the effect was small for clinician-rated recovery at post-treatment (RR = 0.82, 95% CI = 0.64 to 1.06) with no significant heterogeneity between studies (I2 = 37%, Chi2 = 6.39, p = 0.17) nor between subgroups (I2 = 0%, Chi2 = 0.20, p = 0.65). The CIs increased at follow-up (RR = 0.72, 95% CI = 0.19 to 2.67) with considerable heterogeneity between studies (I2 = 80%, Chi2 = 5.02, p = 0.03) and no significant difference between subgroups (I2 = 50%, Chi2 = 2.00, p = 0.16).

In two studies (LYNEHAM2012, THIRLWALL2012), the effect was small for self-rated symptoms of social anxiety disorder at post-treatment, but not significant (SMD = −0.15, 95% CI = −1.03 to 0.73), with no significant heterogeneity (I2 = 43%, Chi2 = 3.52, p = 0.17).

In the three studies with therapist support (CARTWRIGHT-HATTON2012, LYNEHAM2012, THIRLWALL2012), there was a small effect for parent-rated symptoms of social anxiety disorder at post-treatment that was not significant (SMD = −0.38, 95% CI = −0.96 to 0.19), with no heterogeneity (I2 = 0%, Chi2 = 1.64, p = 0.65). Only one study reported controlled effects for parent-rated symptoms at follow-up (CARTWRIGHT-HATTON2012), and the effect was larger but not significant with wide CIs (SMD = −0.72, 95% CI = −1.80 to 0.35).

Figure 12. Recovery for CBT via parents compared with waitlist.

Figure 12

Recovery for CBT via parents compared with waitlist.

Compared with self-help

One study (LYNEHAM2012) compared a group cognitive behavioural intervention with a self-help book for parents delivered with therapist support over 16 weeks. Participants were 6 to 13 years and received approximately 20 hours of contact in the CBT group and 8 hours in the self-help group. There was a small effect between the interventions on self-rated symptoms of social anxiety disorder at post-treatment that was not statistically significant (SMD = −0.26, 95% CI = −1.32 to 0.79) and at follow-up (SMD = −0.50, 95% CI = −1.77 to 0.77). Similarly, there was a small effect on parent-rated symptoms of social anxiety disorder that was not statistically significant at post-treatment (SMD = 0.20, 95% CI = −0.85 to 1.25) and at follow-up (SMD = −0.07, 95% CI = −1.25 to 1.12).

7.5.3. Self-help and supported self-help

Two studies compared internet self-help interventions for children and young people with any anxiety disorder versus waitlist (MARCH2009, SPENCE2011). Interventions were delivered to children and young people with parental involvement. Participants in one study (MARCH2009) were aged 7 to 12 years and participants in the other were 12 to 18 years (SPENCE2011). A third study used an internet intervention aimed specifically at young people (15 to 21 years) with social anxiety disorder (TILLFORS2011). For studies of children and young people with multiple diagnoses, data for those with primary social anxiety disorder were included in the main analysis. A sensitivity analysis included participants with social anxiety disorder as either their primary or secondary diagnosis (see Appendix 17). Treatment lasted 9 to 10 weeks. Therapists spent approximately 2 hours supporting parents in one study (MARCH2009) and the amount of contact was unclear in SPENCE2011 and TILLFORS2011.

In two (MARCH2009, SPENCE2011), the effect was small and not statistically significant for clinician-rated recovery at post-treatment (RR = 0.85, 95% CI = 0.62 to 1.15), with no significant heterogeneity (I2 = 24%, Chi2 = 1.31, p = 0.25). The studies did not report controlled effects at follow-up.

Across all three studies, there was a medium effect for self-rated symptoms of social anxiety disorder at post-treatment that was not statistically significant (SMD = −0.47, 95% CI = −1.71 to 0.78), with considerable heterogeneity between studies (I2 = 81%, Chi2 = 10.54, p = 0.005). There was a significant difference between the generic anxiety treatments and the study using an intervention specifically designed for social anxiety disorder (I2 = 82%, Chi2 = 5.63, p = 0.02). No controlled effects at follow-up were reported.

In two studies (MARCH2009, SPENCE2011) there was a small effect for parent-rated symptoms at post-treatment that was not statistically significant (SMD = −0.33, 95% CI = −0.94 to 0.27), with no heterogeneity (I2 = 0%, Chi2 = 0.00, p = 0.97).

One study compared individual CBT with an internet-delivered self-help intervention supported by a therapist (SPENCE2011) over 10 weeks. Participants were aged 12 to 18 years and received approximately 10 hours of contact in the CBT group and 2 hours in the self-help group. There was no statistical difference between the interventions on self-rated (SMD = 0.13, 95% CI = −0.64 to 0.90) or parent-rated symptoms of social anxiety disorder at post-treatment (SMD = 0.21, 95% CI = −0.57 to 1.00). No controlled effects at follow-up were reported.

7.6. CLINICAL EVIDENCE SUMMARY

7.6.1. Pharmacological interventions

A systematic search identified few studies of pharmacological interventions for children and young people with social anxiety disorder. There was some evidence of a small increase in recovery and a moderate reduction in symptoms of social anxiety disorder with two SSRIs (fluoxetine and paroxetine), but these were from a few relatively small studies; bias and publication bias may have also affected the results. The GDG referred to studies of pharmacological interventions for depression in children and young people (NCCMH, 2005), which demonstrate that pharmacological therapy (in particular the SSRIs, with the possible exception of fluoxetine) may be associated with serious adverse events, including increased suicide.

In the one trial comparing drugs (fluoxetine) and group CBT there was a suggestion that group CBT may be more effective in prompting recovery.

7.6.2. Psychological interventions

There is limited evidence that psychological interventions may be efficacious for children and young people with social anxiety disorder, but small sample sizes require caution to be exercised when coming to any conclusions about which specific interventions are most effective. Psychological interventions that include group CBT, exposure and opportunities to practice and receive feedback may have performed better than others. Group CBT specifically for social anxiety disorder may be more effective than group CBT for all anxiety disorders, but there have been no direct comparisons of outcomes following general anxiety and social anxiety-specific treatments.

For younger children, there is some evidence that CBT delivered by parents who received specific training in the intervention can reduce symptoms of social anxiety disorder and help children recover.

7.6.3. Combined interventions

There were no trials of combined psychological and pharmacological interventions for children and young people with social anxiety disorder.

7.7. HEALTH ECONOMIC EVIDENCE

7.7.1. Systematic literature review

No studies assessing the cost effectiveness of children and young people with social anxiety disorder were identified by the systematic search of the economic literature undertaken for this guideline. Details on the methods used for the systematic search of the economic literature are described in Chapter 3.

7.8. FROM EVIDENCE TO RECOMMENDATIONS

The evidence identified in the review is limited and although generally rated at low risk of bias, the size of the dataset and considerable variation in the nature of the interventions and the different populations included in the trials required caution to be exercised when generating recommendations.

The GDG considered that recovery from social anxiety disorder was the most important clinical outcome and that, for pharmacological interventions, side effects were an especially important concern in children and young people because of the potential increased risk of harm with side effects in this age group. Given the limited dataset, the absence of any licence for the use of drugs in social anxiety disorder, and potential harms, the GDG decided that drugs should not be routinely offered for the treatment of social anxiety disorder in children and young people. Drawing on the evidence for physical interventions reviewed in Chapter 6, the GDG decided also not to recommend the use of such interventions (for example, botulinum toxin) for children and young people.

Although the data for psychological interventions were also limited, there was a relatively more substantial and effective set of interventions that did not carry the same potential harms as drugs. The GDG judged that CBT (individual or group) focused on social anxiety disorder was the most promising intervention for children and young people. The GDG wished to emphasise that for younger children, the therapist should consider involving the parents or carers to help deliver the intervention effectively; some trials suggest there are positive effects when parents are offered training to facilitate this. Although the data were limited, these results were consistent with evidence for adults in that CBT was the most efficacious intervention. The GDG was also of the view that the underlying mechanisms of change were also similar. Given these factors the GDG decided that for older adolescents (this typically could include young people aged 15 years and older but would vary with developmental and emotional maturity), consideration should also be given to offering them psychological interventions recommended for adults (see Chapter 6).

As with the delivery of adult psychological interventions, the GDG was concerned that psychological interventions should be delivered properly and the outcomes effectively monitored and therefore they decided to adopt the same recommendation as was developed for adults, adjusting the outcome measures to be appropriate for children and young people. In addition, the GDG was concerned that children and young people would have less control over the home, social and educational environment and decided on the basis of their expert knowledge that those delivering interventions should take care to ensure that wider environmental concerns were taken into consideration when developing and implementing treatment plans. The GDG also judged that offering interventions for comorbid substance misuse as recommended for adults was relevant to young people.

7.9. RECOMMENDATIONS

7.9.1. Treatment principles

7.9.1.1.

All interventions for children and young people with social anxiety disorder should be delivered by competent practitioners. Psychological 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 sessional outcome measures, for example:
    –

    the LSAS – child version or the SPAI-C, and the SPIN or LSAS for young people

    –

    the MASC, RCADS, SCAS or SCARED for children

  • engage in monitoring and evaluation of treatment adherence and practitioner competence – for example, by using video and audio tapes, and external audit and scrutiny if appropriate.
7.9.1.2.

Be aware of the impact of the home, school and wider social environments on the maintenance and treatment of social anxiety disorder. Maintain a focus on the child or young person's emotional, educational and social needs and work with parents, teachers, other adults and the child or young person's peers to create an environment that supports the achievement of the agreed goals of treatment.

7.9.1.3.

For people (including young people) with social anxiety disorder who misuse substances, be aware that alcohol or drug misuse is often an attempt to reduce anxiety in social situations and should not preclude treatment for social anxiety disorder. Assess the nature of the substance misuse to determine if it is primarily a consequence of social anxiety disorder and:

  • offer a brief intervention for hazardous alcohol or drug misuse (see Alcohol-Use Disorders [NICE clinical guideline 115]29 or Drug Misuse [NICE clinical guideline 51]30)
  • for harmful or dependent alcohol or drug misuse consider referral to a specialist alcohol or drug misuse service31.

7.9.2. Treatment for children and young people with social anxiety disorder

7.9.2.1.

Offer individual or group CBT focused on social anxiety (see recommendations 7.9.3.1 and 7.9.3.2) to children and young people with social anxiety disorder. Consider involving parents or carers to ensure the effective delivery of the intervention, particularly in young children.

7.9.3. Delivering psychological interventions for children and young people

7.9.3.1.

Individual CBT should consist of the following, taking into account the child or young person's cognitive and emotional maturity:

  • 8 − 12 sessions of 45 minutes' duration
  • psychoeducation, exposure to feared or avoided social situations, training in social skills and opportunities to rehearse skills in social situations
  • psychoeducation and skills training for parents, particularly of young children, to promote and reinforce the child's exposure to feared or avoided social situations and development of skills.
7.9.3.2.

Group CBT should consist of the following, taking into account the child or young person's cognitive and emotional maturity:

  • 8−12 sessions of 90 minutes' duration with groups of children or young people of the same age range
  • psychoeducation, exposure to feared or avoided social situations, training in social skills and opportunities to rehearse skills in social situations
  • psychoeducation and skills training for parents, particularly of young children, to promote and reinforce the child's exposure to feared or avoided social situations and development of skills.
7.9.3.3.

Consider psychological interventions that were developed for adults (see Section 6.13) for young people (typically aged 15 years and older) who have the cognitive and emotional capacity to undertake a treatment developed for adults.

7.9.4. Interventions that are not recommended

7.9.4.1.

Do not routinely offer pharmacological interventions to treat social anxiety disorder in children and young people.

7.9.5. Research recommendations

7.9.5.1.

What is the clinical and cost effectiveness of specific CBT for children and young people with social anxiety disorder compared with generic anxiety-focused CBT? (See Appendix 9 for further details.)

7.9.5.2.

What is the best way of involving parents in the treatment of children and young people (at different stages of development) with social anxiety disorder? (See Appendix 9 for further details.)

7.9.5.3.

What is the clinical and cost effectiveness of individual and group CBT for children and young people with social anxiety disorder? (See Appendix 9 for further details.)

Footnotes

26
27
28

Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

29

NICE, 2011.

30
31

This recommendation also appears in Chapter 6 regarding interventions for adults.

Copyright © The British Psychological Society & The Royal College of Psychiatrists, 2013.

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.

Bookshelf ID: NBK327663
Contents

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (4.7M)

In this Page

Other titles in this collection

Related NICE guidance and evidence

Recent Activity

ClearTurn OffTurn On

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...