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

National Collaborating Centre for Mental Health (UK). Post-Traumatic Stress Disorder: The Management of PTSD in Adults and Children in Primary and Secondary Care. Leicester (UK): Gaskell; 2005. (NICE Clinical Guidelines, No. 26.)

Cover of Post-Traumatic Stress Disorder

Post-Traumatic Stress Disorder: The Management of PTSD in Adults and Children in Primary and Secondary Care.

Show details

9Children and young people with PTSD

9.1. Introduction

When the diagnosis of PTSD was first formulated in 1980 (American Psychiatric Association, 1980) it was initially believed that it would not be relevant to children and young people. This was soon demonstrated to be false and it is now accepted that children and young people can develop PTSD following traumatic events. For the purposes of this guideline, ‘children’ refers to ages 2–12 years and ‘young people’ (adolescents) to ages 13–18 years.

9.2. Developmental differences

9.2.1. Post-traumatic stress reactions in children

The broad categories of PTSD symptoms (re-experiencing, avoidance/numbing and increased arousal) are present in children as well as in adults. The requirements of DSM criteria for the diagnosis of PTSD in children are that children must exhibit at least one re-experiencing symptom, three avoidance/numbing symptoms and two increased arousal symptoms. From the age of 8–10 years, following traumatic events, children display reactions closely similar to those manifested by adults. Below 8 years of age, and in particular below the age of 5 years, there is less agreement as to the range and severity of the reactions. Scheeringa et al (1995) have suggested an alternative set of criteria for the diagnosis of PTSD in children, placing more emphasis on regressive behaviours and new fears, but these have yet to be fully validated.

Traumatic reactions in children have been less extensively studied than in adults and there are few naturalistic, longitudinal studies mapping the natural history of these reactions. It has long been recognised (Eth, 2001) that it is much more difficult to elicit evidence of emotional numbing in young children. Other items indicating avoidance reactions in children simply are not relevant, thereby making it difficult for children to meet DSM criteria for that part of the diagnostic algorithm (although this does not apply to the ICD diagnosis).

In general, it is agreed that children display a wide range of stress reactions. To some extent these vary with age, with younger children displaying more overt aggression and destructiveness. They may also show more repetitive play about the traumatic event, and this may even be reflected in repetitive drawing.

9.2.2. Family influences

As with other anxiety disorders, children’s reactions are influenced by parental reactions. In addition to modelling on their parents’ reactions (social influence) there are probably also inherited dispositions to react adversely to traumatic events (genetic influence). This has not been adequately studied in relation to PTSD in children.

What is clinically described and widely accepted is that children are very sensitive to their parents’ reactions – both to the event itself and to talking about it afterwards. Children often say that they choose not to discuss a traumatic event and/or their reactions to it with their parents, as they do not wish to upset the parents further. This is, in part, one of the reasons for the finding that even more than with other anxiety disorders, parents grossly underestimate the degree of stress reactions experienced by their children. Thus, one cannot rely solely on parental report when making diagnoses or estimating prevalence. A study by McFarlane (1987) suggested that in an Australian bush fire, the children’s reactions to the event were fully accounted for by the mothers’ own mental health, rather than by the exposure to the fire. However, as mothers had rated both their own adjustment and that of their children, this finding was suspect. Subsequent studies (for example Smith et al, 2001) have found that direct exposure is usually a stronger determinant of child reaction, with maternal reactions being important modifying influences. It is important therefore always to consider the nature and extent of a child’s exposure to a traumatic event.

9.2.3. Multiple versus single trauma

Many children presenting with symptoms of PTSD may have been subjected to multiple traumas such as childhood sexual abuse or domestic violence. The most common form of multiple trauma for children that has been studied and investigated is childhood sexual abuse, which often occurs in secret and is repeated over a long period. The traumatic reactions associated with such multiple trauma can be usefully construed as similar to those that follow from single traumas, although issues of abuse of power, loss of trust and so on do make them different. Although there is evidence that the social circumstances and events surrounding multiple traumas for children may have consequences for their future management (Ramchandani & Jones, 2003), the evidence does not support the idea that multiple traumas are associated with significantly different outcomes or that the treatment required for PTSD is significantly different when compared with single traumas.

9.3. Incidence, prevalence and natural history

9.3.1. Prevalence

Most epidemiological studies have been of older young people and adults. Giaconia et al (1995) reported a lifetime prevalence of 6% in a community sample of older young people. Kessler et al (1995) reported a lifetime prevalence of 10% using data collected from older young people and adults in the US National Comorbidity Survey. In contrast, the British National Survey of Mental Health of over 10 000 children and young people (Meltzer et al, 2000) reported that 0.4% of children aged 11–15 years were diagnosed with PTSD, with girls showing twice the rate of boys. Below the age of 10 years, PTSD was scarcely registered. This lower rate is, of course, a point prevalence estimate and is bound to be lower than a lifetime prevalence estimate. Moreover, the screening instrument employed was not specifically developed to screen for PTSD.

9.3.2. Incidence

Estimates of the incidence of PTSD are more frequently reported after various natural and other disasters. Rates vary enormously, partly as a result of different methodologies and partly as a result of different types of traumatic event. In various studies of the effects of road traffic accidents (not resulting in an overnight stay in hospital) rates of 25–30% are reported. The study of 200 young survivors of the sinking of the cruise ship Jupiter (Yule et al, 2000) reported an incidence of PTSD of 51%. Most cases manifested within the first few weeks, with delayed onset being rare. Other disorders such as anxiety and depression were common as well. Studies of the mental health of child refugees from war-torn countries find the incidence to be close to 67% (W. Yule, personal communication, 2004). Therefore, significantly increased demands may be made at all levels of primary and secondary child and adolescent mental health services following traumatic events.

The implication of this for the NHS is that while the numbers of children and young people experiencing PTSD at any one point in time may be approaching 1% and represents a significant level of morbidity in any community, by way of comparison, in adults PTSD has a point prevalence of 1.5–3%, and schizophrenia in adults has a prevalence of 1%.

9.3.3. Natural history

The follow-up study of young people who survived the sinking of the Jupiter found that 15% still met criteria for PTSD 5–7 years after the event. More recently, a 33-year follow-up of the children who survived the Aberfan landslide disaster found that 29% of those traced and interviewed still met criteria for PTSD (Morgan et al, 2003). In other words, in the absence of effective therapy, the long-term effects of life-threatening, traumatic events in childhood can be severe.

9.4. Diagnostic and assessment measures

9.4.1. Children over 7 years old

More is known about screening, assessment and diagnosis in children over the age of 7 years because above that age many children can read independently and can complete self-rating scales. It is much more time-consuming and expensive to conduct standardised clinical interviews with both parent and child to establish a diagnosis in large groups of children.

9.4.1.1. Self-completed PTSD scales

The most widely used self-report scales in research and clinical settings are the Children’s Impact of Event Scale, the Child Post Traumatic Stress Reaction Index and the Child PTSD Symptom Scale. For a detailed recent review of self-completed scales, see Ohan et al (2002).

The Children’s Impact of Event Scale was developed from the widely used adult self-report PTSD measure, the Impact of Event Scale (Horowitz et al, 1979). The adult version has been used with children and young people in its original 15-item version. Following two large principal component analyses, a briefer eight-item version was developed for children (Yule, 1997) and subsequently expanded to a 13-item version to include five items attempting to measure arousal (see http://www.childrenandwar.org).

The Child Post Traumatic Stress Reaction Index (CPTS–RI) was originally rated following interview by a clinician with the carer and sometimes the child. More recently it has been modified to be a self-report instrument (Pynoos et al, 1987; Pynoos, 2002).

The Child PTSD Symptom Scale (CPSS; Foa et al, 2001) is a 17-item scale used both in initial diagnosis and in monitoring progress. It contains a brief functional impairment rating.

9.4.1.2. Structured interviews for PTSD in children and young people

Structured interviews for children are not well developed; three of the more commonly used scales are described below.

  • The Clinician-Administered PTSD Scale for Children and Adolescents for DSM–IV (CAPS–CA; Nader et al, 2002) is modelled on the adult CAPS and is widely regarded as the gold standard measure to diagnose DSM–IV PTSD in children.
  • The Anxiety Disorders Interview Schedule for Children for DSM–IV (ADIS–C; Silverman & Albano, 1996) can be used to diagnose a range of anxiety disorders and has a specific module for PTSD symptoms.
  • The Schedule for Affective Disorders and Schizophrenia for School Age Children (K–SADS; Kaufman et al, 1997) can be used to diagnose a range of anxiety disorders and has specific supplementary questions for measuring PTSD symptoms. Both the parent and the child are interviewed.
Trauma-specific PTSD measures

The Children’s Impact of Traumatic Events Scale – Revised (CITES–R; Wolfe et al, 1991) is a measure of PTSD symptoms arising from sexual abuse and measures aspects such as social reactions to disclosure, eroticism and abuse-related attributions in addition to non-trauma-specific PTSD symptoms.

9.4.2. Children aged 7 years or younger

No consensus has emerged as to how to measure PTSD symptoms in children aged 7 years or younger. In the recent past a range of scales measuring behavioural problems have been adopted such as the Child Behaviour Checklist (CBCL; Achenbach & Edelbrock, 1983) and, for children who have suffered sexual abuse, the Child Sexual Behaviour Inventory (CSBI; Friedrich et al, 1992).

9.4.2.1. Measures of outcome for children within this review

Given the lack of consensus about the measurement of PTSD for younger children, a range of child-specific measures were included in this review (Table 9.1).

Table 9.1. Measures of PTSD in children.

Table 9.1

Measures of PTSD in children.

In addition to PTSD scales, a range of child measures of depression, anxiety and quality of life were included within the review.

9.4.2.2. Measures of exposure to traumatic events

The structured interviews indicate the most likely adverse life events that may result in PTSD in children and young people, but they do not constitute formal measures. General practitioners, paediatricians and child mental health workers who see a child presenting with a sudden change in sleep pattern, nightmares and jumpiness should enquire about intrusive images and then ask whether the child has experienced any threatening life event such as a bad accident, natural disaster, or physical or sexual abuse.

9.4.2.3. Measures of process and related aspects

Increasing attention is being paid to cognitive factors such as the way in which children attribute blame for an event or the extent to which they erroneously believe that they might have died in the accident. The effective social support that is available to the child is also likely to be a key determinant of whether the child continues to respond adversely (Joseph et al, 1993). Standard measures of these aspects are still being developed.

9.5. Psychological interventions

Early intervention would be attractive if it could be shown that it prevented later development of PTSD or other disorders, but, as with adult studies, there have been few properly controlled trials of any early intervention. The only one known is that of Stallard et al (2005), which is discussed below.

Ramchandani & Jones (2003) reported a systematic review of RCTs treating a range of psychological symptoms in sexually abused children. They identified 12 RCTs: three investigating group CBT; six investigating individual CBT; one of adding group therapy to a family therapy intervention; and two comparing individual (non-CBT) therapy with group therapy. However, the dependent (outcome) measures were very varied, and only five studies looked at recognised, specific measures of PTSD.

9.5.1. Studies included

The inclusion criteria that 70% of participants within a study have a PTSD diagnosis was not applied for the review of children and young people because, as discussed above, diagnosis of child and adolescent PTSD is still evolving and is relatively undeveloped for younger children. Otherwise the inclusion criteria were identical to those for adults (see Chapter 4). However, all studies had to include a measure of the child’s PTSD symptoms, although as discussed above (see section 9.4.5) a wider range of measures was deemed more acceptable than for adult PTSD scales.

From the main search for RCTs (see Appendix 6), 11 studies of psychological interventions were identified by the Guideline Development Group as meeting the inclusion criteria: CELANO 1996, COHEN 1996, COHEN 1997 (COHEN 1997 is a follow-up study to COHEN 1996), COHEN 1998, COHEN 2004, DEBLINGER 1999, JABERGHADERI1, KING 2000, STALLARD, STEIN 2003 and TROWELL 2002. References given in shortened format and summary characteristics of individual included trials are given in Appendix 14.

9.5.1.1. Interventions considered

Broadly, four different psychological interventions were covered within the included studies: one early intervention treatment (debriefing), cognitive–behavioural therapy, eye movement desensitisation and reprocessing (EMDR) and supportive therapy (fuller definitions of these treatment classifications, as they apply to adults, are given in Chapter 5).

Within these broad categories of treatments there was considerable variation in how treatments were delivered, with many studies allowing for some part of treatment being delivered to the caretaker as well as the child, either individually or in sessions for both the child and the caretaker. Two studies involved treatment arms that consisted of treatments delivered to groups of children (STEIN 2003, TROWELL 2002). Given the many different formats in which the four treatments were delivered, it was not possible to combine many of the studies for the purpose of this review. Table 9.2 provides a summary of the range of interventions and delivery formats (treatment delivered to both the caretaker/parent and the child, to the child individually, etc.) for which eligible studies were available.

Table 9.2. Interventions for which eligible studies were available.

Table 9.2

Interventions for which eligible studies were available.

9.5.1.2. Populations – childhood sexual abuse and other traumas

Nine of the studies related to childhood sexual abuse and were analysed separately from the remaining studies, which covered a range of traumatic events including witnessing violence, natural disaster, war and burns.

9.5.2. Childhood sexual abuse

Trials comparing different forms of CBT with waiting list or supportive therapy

Seven of the trials covering childhood sexual abuse compared different forms of CBT therapies against waiting list, supportive therapy or another mode of delivering CBT (CELANO 1996, COHEN 1996, COHEN 1997COHEN 1998, COHEN 2004, DEBLINGER 1999, KING 2000).

9.5.2.1. Child and carer CBT versus supportive therapy

Children over 7 years old

One large study (COHEN 2004) compared trauma-focused CBT with supportive therapy for children aged 8–14 years and a parent. Each treatment arm consisted of 12 weekly sessions of 45 min for the child individually and 45 min for the parent, although three of the weekly sessions involved 30 min of joint parent–child therapy. Supportive therapy was child- and parent-centred, allowing the child or parent to guide the structure and content of the treatment, supplemented by the provision of written psychoeducational information. Trauma-focused therapy worked on expression of feelings, coping skills and gradual exposure, whereby the children were assisted in developing their own trauma narrative as well as some psychoeducation. An earlier smaller study by COHEN 1998 of CBT versus supportive therapy for children aged 7–14 years compared similar interventions but used a behavioural-based measure to assess PTSD symptoms. Each treatment arm consisted of 12 individual treatment sessions of 90 min (45 min for the child individually and 45 min for the parent individually). Another study (CELANO 1996) compared developmentally appropriate cognitive–behavioural techniques and metaphoric techniques with supportive therapy for girls aged 8–13 years and their carers. Each treatment arm consisted of eight 1-hour weekly sessions. Sessions were split with 30 min of treatment each for the child and the carer individually, although two or three sessions included some joint work. For this review the COHEN 1998, COHEN 2004 and CELANO 1996 studies were combined.

There was limited evidence that CBT for children over 7 years old and their carers was better than supportive therapy in reducing the severity of PTSD symptoms post-treatment. For the other outcome measures post-treatment, the evidence was either inconclusive (child-rated depression, parent-rated internalising and externalising behaviours, sexualised behaviours and likelihood of leaving the study early) or indicated that there was unlikely to be a clinically important difference (child-rated anxiety). Unfortunately we do not know how sustained these improvements are because no follow-up data are currently available.

There is limited evidence favouring CBT for children over 7 years old and their parents/carers over supportive therapy on reducing the severity of PTSD symptoms (k=2; n=212; SMD=−0.55, 95% CI −0.83 to −0.28). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between CBT for children over 7 years old and their parents/carers and supportive therapy on reducing self-rated post-treatment depression symptoms as measured by the Child Depression Inventory (k=2; n=232; SMD=−0.44, 95% CI −0.7 to −0.18). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between CBT for children over 7 years old and their parents/carers and supportive therapy on reducing the likelihood of leaving the study early for any reason (k=2; n=276; RR=1.18, 95% CI 0.77 to 1.82). [I]

Children under 7 years old

The study by COHEN 1996 and its follow-up study (COHEN 1997) compared CBT with supportive therapy for children under 7 years old. Each treatment arm consisted of 12 weekly sessions of 50 min for the parent individually and 30–40 min for the child. Measurement of PTSD symptoms in very young children is still evolving, and for this study there was no direct measure of PTSD but rather a range of behavioural measures such as CBCL and CSBI. There is limited evidence favouring CBT over supportive therapy for reducing parent-rated externalising symptoms and sexualised behaviour both post-treatment and at 1-year follow-up.

There is limited evidence favouring CBT for children under 7 years old and their parents/carers over supportive therapy on reducing parent-rated externalising behaviours post-treatment as measured by the CBCL (k=1; n=67; SMD=−0.79, 95% CI −1.29 to −0.28). [I]

There is limited evidence favouring CBT for children under 7 years old and their parents/carers over supportive therapy on reducing parent-rated externalising behaviours at 1-year follow-up as measured by the CBCL (k=1; n=43; SMD=−0.53, 95% CI −1.17 to 0.11). [I]

9.5.2.2. Child only, mother only and mother and child CBT versus community care

One study (DEBLINGER 1999) compared CBT treatments with community care. Community care consisted of support from child protection workers and victim witness advocates and encouragement to seek therapists within the local community, and was treated as an active intervention for this review. There were three CBT treatment arms comprising CBT for the child only, CBT for the mother only and CBT for child and mother. Individual treatment consisted of 12 treatment sessions of 45 min. The joint treatment condition comprised 12 sessions of 90 min with individual sessions for the child and mother and some joint sessions. Children ranged in age from 7 years to 13 years.

CBT for the child versus community care

On the clinician-rated K–SADS–E, there is limited evidence favouring CBT for the child only over community care on reducing the severity of PTSD symptoms both immediately post-treatment and 6 months post-treatment. However, by the 2-year follow-up the difference between the groups had diminished. There are similar results for the CDI (self-report) measure of depression, and the advantage of CBT for child only is present at the end of treatment and at the 6-month and 2-year post-treatment evaluations. Given the nature of the community care, evidence on tolerability (leaving the study early) is difficult to interpret.

There is limited evidence favouring CBT for the child only over community care on reducing the severity of PTSD symptoms post-treatment as measured by K–SADS–E (clinician-rated measure) (k=1; n=35; SMD=−0.96, 95% CI −1.68 to −0.24). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between CBT for the child only and community care on reducing the severity of PTSD symptoms at 2 years post-treatment as measured by K–SADS–E (clinician-rated measure) (k=1; n=35; SMD=−0.45, 95% CI −1.14 to 0.24). [I]

CBT for the mother only versus community care

Cognitive–behavioural therapy with the mother alone does not appear to have any advantage over community treatment as far as the severity of PTSD symptoms as measured by the K–SADS–E is concerned, either immediately post-treatment or at 6 months. However, somewhat surprisingly, at 2 years’ follow-up the reduction in the severity of PTSD symptoms was clinically significant. For children’s self-report depression symptoms the pattern is similar, with clinically significant effects occurring at 2 years’ follow-up but not for earlier assessments.

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between CBT for the mother only and community care on reducing the severity of PTSD symptoms post-treatment as measured by K–SADS–E (clinician-rated measure) (k=1; n=34; SMD=−0.43, 95% CI −1.13 to 0.26). [I]

There is limited evidence favouring CBT for the mother only over community care on reducing the severity of PTSD symptoms at 2 years post-treatment as measured by K–SADS–E (clinician-rated measure) (k=1; n=34; SMD=−0.77, 95% CI −1.48 to −0.06). [I]

Mother and child CBT versus community care

On the clinician-rated K–SADS–E there is limited evidence favouring CBT for the child and mother over community care on reducing the severity of PTSD symptoms both immediately post-treatment and at 6 months’ follow-up, and this improvement was quite well sustained at 2 years’ follow-up. The results for depression (child-rated) showed clinically significant improvement at the post-treatment and 2-year follow-up assessments but did not reach the threshold for clinical significance at the intervening assessments.

There is limited evidence suggesting a difference favouring mother and child CBT over community care on reducing the severity of PTSD symptoms post-treatment as measured by K–SADS–E (clinician-rated measure) (k=1; n=33; SMD=−0.86, 95% CI −1.58 to −0.13). [I]

There is limited evidence suggesting a difference favouring mother and child CBT over community care on reducing the severity of PTSD symptoms at 2 years post-treatment as measured by K–SADS–E (clinician-rated measure) (k=1; n=33; SMD=−0.64, 95% CI −1.35 to 0.07). [I]

9.5.2.3. Child only and child and mother CBT versus waiting list

One study (KING 2000) compared CBT intervention for the child individually with CBT for the child and mother jointly. The interventions consisted of 20 weekly sessions of 50 min; however, the child and mother joint intervention arm consisted of a further 20 weekly 50 min sessions of training for the parents in child behaviour management skills. Children ranged in age from 5 years to 17 years.

Individual child CBT versus waiting list

There was limited evidence of a clinically important improvement for PTSD severity, although this was not sustained at 3 months’ follow-up. Effect sizes for depression and anxiety did not reach the threshold for clinical importance.

There is limited evidence suggesting a difference favouring child CBT over waiting list on reducing the severity of PTSD symptoms post-treatment as measured by the ADIS–C (clinician-rated measure) (k=1; n=24; SMD=−1.05, 95% CI −1.92 to −0.19). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between child CBT and waiting list on reducing depression symptoms as measured by the CDI (k=1; n=24; SMD=−0.29, 95% CI −1.1 to 0.51). [I]

Child and mother CBT versus waiting list

The results were similar to those for individual child CBT versus waiting list, with limited evidence of a clinically important improvement for PTSD severity, although this was not sustained at 3 months’ follow-up. Effect sizes for depression and anxiety did not reach the threshold for clinical importance.

There is limited evidence suggesting a difference favouring child and mother CBT over waiting list on reducing the severity of PTSD symptoms post-treatment as measured by the ADIS–C (clinician-rated measure) (k=1; n=24; SMD=−1.19, 95% CI −2.08 to −0.31). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between child and mother CBT and waiting list on reducing depression symptoms post-treatment as measured by the CDI (k=1; n=24; SMD=−0.28, 95% CI −1.09 to 0.52). [I]

Individual child CBT versus child and mother CBT

Direct comparison of individual child CBT versus CBT for the child and mother did not yield evidence of clinically important differences between the two treatment conditions for PTSD severity, depression, anxiety or tolerability (leaving the study early).

9.5.2.4. Individual versus group psychotherapy for children

One study (TROWELL 2002) compared group psychotherapy with individual psychotherapy for sexually abused girls aged 6–14 years. Individual psychotherapy entailed up to 30 weekly sessions, compared with up to 18 sessions for those completing group psychotherapy. Unfortunately data were not available for total PTSD symptoms (arousal symptoms data were not reported). There was limited evidence that individual therapy was better than the group delivery in terms of reducing re-experiencing and avoidance symptoms at 12 months and 24 months post-therapy, although effect sizes were borderline for clinical importance. The evidence suggests that neither treatment was substantially better tolerated than the other.

There is limited evidence suggesting a difference favouring individual psychotherapy over group psychotherapy on reducing re-experiencing and avoidance symptoms at 12 months using the K–SADS-based Orvaschel scale (Orvaschel, 1989) (k=1; n=56; SMD=−0.49, 95% CI −1.02 to 0.05). [I]

9.5.2.5. Comparing EMDR with other treatment (CBT)

One unpublished study (JABERGHADERI) compared EMDR with CBT. Each treatment arm entailed up to 12 sessions (duration not specified) and CBT incorporated a degree of exposure work. The evidence was inconclusive for both child-reported and parent-reported PTSD severity, using the Child Report and the Parent Report of PTSD Symptoms, respectively (Greenwald & Rubin, 1999), and there was no evidence that one treatment was better tolerated.

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between EMDR and CBT on reducing child self-report PTSD severity as measured by CROPS (k=1; n=14; SMD=−0.49, 95% CI −1.55 to 0.58). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between EMDR and CBT on reducing parent-rated PTSD severity as measured by PROPS (k=1; n=14; SMD=−0.18, 95% CI −1.23 to 0.87). [I]

9.5.3. Other trauma

9.5.3.1. Debriefing

One unpublished study (STALLARD) compared single-session debriefing with a generally supportive talk to children aged 7–18 years who had been involved in road traffic accidents, within approximately 2 weeks of the accident occurring. The interventions were of approximately equal duration (68 min). The evidence suggested that there is unlikely to be a clinically important difference at 8 months’ follow-up between single-session debriefing and supportive talk for self-rated measures of PTSD severity, depression and anxiety. The evidence was inconclusive for PTSD diagnosis and tolerability.

There is evidence suggesting there is unlikely to be a clinically important difference between structured debriefing and attention control on reducing child self-rated PTSD severity at 8 months’ follow-up (k=1; n=132; SMD=−0.05, 95% CI −0.39 to 0.3). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between structured debriefing and attention control on reducing PTSD diagnosis at 8 months’ follow-up (k=1; n=158; RR=0.97, 95% CI 0.58 to 1.62). [I]

9.5.3.2. CBT interventions

One study (STEIN 2003) compared group CBT against waiting list (delayed intervention) for children who had been exposed to violence. The intervention consisted of ten group sessions and was delivered in a school mental health clinic. The children were approximately 10–12 years old.

Group CBT versus waiting list

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between group CBT and waiting list on reducing the severity of PTSD symptoms post-treatment as measured by CPSS (self-report) (k=1; n=117; SMD=−0.71, 95% CI −1.08 to −0.33). [I]

The evidence is inconclusive and so it is not possible to determine if there is a clinically important difference between group CBT and waiting list on depressive symptoms as measured by CDI (k=1; n=117; SMD=−0.38, 95% CI −0.74 to −0.01). [I]

9.6. Clinical summary of psychological interventions

The above evidence suggests that psychological interventions, specifically trauma-focused cognitive–behavioural psychotherapy, can be effective for the treatment of post-traumatic stress symptoms in children and young people who have been sexually abused.

In contrast, there is very little evidence from RCTs for the efficacy of any psychological interventions for children or young people who suffer from PTSD arising from other forms of trauma. This reflects not the inconclusive nature of the evidence but rather the lack of RCTs. This means that conclusions about the effectiveness of psychological interventions for this group are reliant on extrapolation from other areas, principally work on PTSD with sexually abused children and psychological interventions for adults. Considerable caution is therefore required in drawing conclusions, particularly when drawing on downward extension of results from work with adults. Nevertheless, the limited psychological trials available suggest that trauma-focused CBT, whether delivered to children and young people with PTSD or to children who have developed PTSD in the context of childhood sexual abuse, may be of value.

Some of the trials involving children who have suffered sexual abuse included in this review specifically considered to whom and in what combination treatment should be given. For children over 7 years old who have suffered sexual abuse, treatment of the mother alone seems to be ineffective when compared with treatment of the child alone. Indeed, delivering CBT to the mother as well as the child does not of itself seem to confer much advantage over treatment of the child alone on PTSD symptoms. The lack of clinically important effects for trauma-focused CBT treatments delivered to the child and non-abusing parent or caretaker are particularly striking, given that for DEBLINGER 1999 and KING 2000 the ‘joint’ treatment condition essentially entails additional therapeutic time for the parent/caretaker.

No other psychological intervention has yet established a comparable evidence base, but other interventions such as EMDR show some promise, for example in the study by Chemtob et al (2002), which we were unable to include in the review owing to insufficient data being available for the control group, as well as in a number of non-randomised studies. The single study providing a comparison of EMDR against trauma-focused CBT (JABERGHADERI) suffered from the use of a non-standard PTSD measure. The single psychotherapy trial (TROWELL 2002) also used non-standard measures, which, combined with the lack of a waiting list or attention control, made drawing any significant clinical conclusions very difficult.

The evidence base from which to draw conclusions about the treatment of children under 7 years old suffering from PTSD is sparse. The lack of agreement on and use of a common set of measures is particularly of concern for studies of PTSD in very young children, and adds to the difficulties of interpreting an extremely limited data-set. All treatments need to be adapted to accommodate young children’s less mature ways of thinking about their world, and often clinicians will use play materials and drawings to help children focus on what happened to them and how they feel. However, there is a lack of high-quality (randomised controlled trial) evidence that specific types of play therapy or art therapy have therapeutic value in treating PTSD in young children.

The evidence does not support the use of single-session debriefing for children of any age.

9.7. Pharmacological interventions

Although the use of some psychotropic drugs has increased since the 1990s (Riddle et al, 2001; Bramble, 2003; Wolraich, 2003), and there is belief in their efficacy, much of the increase is accounted for by prescribing by doctors who are not child mental health experts. Although drugs are prescribed less often for childhood disorders in the UK than in the USA, there is none the less a considerable rate of prescribing psychotropic drugs for children by general practitioners in the UK (Montoliu & Crawford, 2002).

Few psychotropic medicines are licensed for use with children. Thus, many prescriptions have to be made ‘off licence’ on a named patient basis. The Royal College of Paediatrics and Child Health (2000) states that:

‘The use of unlicensed medicines or licensed medicines for unlicensed applications is necessary in paediatric practice where there is no suitable alternative. Such uses are informed and guided by a respectable and responsible body of professional opinion’.

This advice was given prior to the Medicines Control Agency advising against the use of all but one SSRI for the treatment of major depression in young people, following the discovery that reports of adverse reactions indicating an increased risk of self-harm had been suppressed.

There are major difficulties in conducting adequate drug trials with children and young people, but these need to be undertaken responsibly if potentially useful help is to be made available, if only while awaiting the application of more powerful treatments that are less available.

From the main search for randomised controlled trials (see Appendix 6), no study of drug treatments was identified by the Guideline Development Group as meeting the inclusion criteria. Only one RCT (Robert et al, 1999) was identified, which compared imipramine with chloral hydrate for 25 child burns victims aged 2–19 years for 1 week of treatment. However, this study did not meet the inclusion criteria as outcomes were recorded in the form of remission rates across a range of symptoms rather than a specific measure of PTSD. Open-label trials suggest that propranolol (Famularo et al, 1988), clonidine (Perry, 1994; Harmon & Riggs, 1996; Horrigan, 1996) and carbamazepine (Loof et al, 1995) resulted in symptomatic improvement, but no comparison group was studied.

9.8. Clinical summary of pharmacological interventions

At present there is too little evidence from RCTs, controlled trials, open-label studies or case–control studies to recommend the use of any psychotropic medication to treat PTSD in children or young people.

9.9. Clinical practice recommendations

9.9.1. Assessment

9.9.1.1.

When assessing a child or young person for PTSD, healthcare professionals should ensure that they separately and directly question the child or young person about the presence of PTSD symptoms. They should not rely solely on information from the parent or guardian in any assessment. [GPP]

9.9.1.2.

When a child who has been involved in a traumatic event is treated in an emergency department, emergency staff should inform the parents or guardians of the possibility of the development of PTSD, briefly describe the possible symptoms (for example, sleep disturbance, nightmares, difficulty concentrating and irritability) and suggest that they contact their general practitioner if the symptoms persist beyond 1 month. [GPP]

9.9.2. Early intervention

9.9.2.1.

Trauma-focused cognitive–behavioural therapy should be offered to older children with severe post-traumatic symptoms or with severe PTSD in the first month after the traumatic event. [C]

9.9.3. Chronic PTSD

9.9.3.1.

Children and young people with PTSD, including those who have been sexually abused, should be offered a course of trauma-focused cognitive–behavioural therapy adapted appropriately to suit their age, circumstances and level of development. [B]

9.9.3.2.

Where appropriate, families should be involved in the treatment of PTSD in children and young people. However, treatment programmes for PTSD in children and young people that consist of parental involvement alone are unlikely to be of any benefit for PTSD symptoms. [C]

9.9.3.3.

The duration of trauma-focused psychological treatment for children and young people with chronic PTSD should normally be 8–12 sessions when the PTSD results from a single event. When the trauma is discussed in the treatment session, longer sessions than usual are usually necessary (for example, 90 min). Treatment should be regular and continuous (usually at least once a week) and should be delivered by the same person. [C]

9.9.3.4.

Drug treatments should not be routinely prescribed for children and young people with PTSD. [C]

9.9.3.5.

When considering treatments for PTSD, parents and, where appropriate, children and young people should be informed that, apart from trauma-focused psychological interventions, there is at present no good evidence for the efficacy of widely used forms of treatment of PTSD such as play therapy, art therapy or family therapy. [C]

9.10. Research recommendations

9.10.1. Trauma-focused psychological intervention for children

9.10.1.1.

Randomised controlled trials for children of all ages should be conducted to assess the efficacy and cost-effectiveness of trauma-focused psychological treatments (specifically CBT and EMDR). These trials should identify the relative efficacy of different trauma-focused psychological interventions and provide information on the differential effects, if any, arising from the age of the children or the nature of the trauma experienced.

Rationale

Post-traumatic stress disorder is a common and potentially disabling condition in children as well as adults (Giaconia et al, 1995). Although up to 50% of children may develop PTSD following a traumatic event (Yule et al, 2000), many individuals recover without specific intervention; however, a significant proportion of individuals, perhaps more than 30% of victims of major disasters, go on to develop a chronic disorder with associated psychological and social handicaps (Morgan et al, 2003). Trauma-focused psychological interventions are generally effective for the treatment of PTSD in adults but only a limited evidence base exists for children and young people (Cohen et al, 2000). In addition, much of the evidence is drawn from work with children who have experienced childhood sexual abuse as well as developing PTSD (Ramchandani & Jones, 2003) and therefore the evidence base for interventions for PTSD arising from other traumas is weaker. For children aged under 7 years who develop PTSD there are virtually no formal RCTs of appropriate psychological interventions. A number of non-controlled trials suggest that treatments (specifically CBT and EMDR) are efficacious, but these have not been formally tested (Cohen et al, 2000).

Footnotes

1

Data from this study have now been published as Jaberghaderi et al (2004).

Copyright © 2005, The Royal College of Psychiatrists & The British Psychological Society.

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 Royal College of Psychiatrists.

Bookshelf ID: NBK56490

Views

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...