Show Summary Details
Page of

(p. 25) The Evidence Base for Cognitive-Behavioral Therapy for Pediatric Chronic Pain 

(p. 25) The Evidence Base for Cognitive-Behavioral Therapy for Pediatric Chronic Pain
(p. 25) The Evidence Base for Cognitive-Behavioral Therapy for Pediatric Chronic Pain

Tonya M. Palermo

Page of

date: 26 May 2019

In this chapter, I review the evidence for the effectiveness of cognitive and behavioral therapies in treating children and adolescents with chronic pain conditions, including relaxation-based therapies, biofeedback treatment, and multi-component CBT. This chapter is intended to serve as a resource; therefore, extensive detail is provided in order to not only identify the major published reviews, but to also evaluate the relative strength and conclusiveness of the evidence for CBT.

Considerable effort and attention has been devoted to summarizing the evidence base concerning psychological therapies for the treatment of children and adolescents with chronic and recurrent pain. Given that headache and abdominal pain are the most common types of recurrent pain in children, most of the treatment literature has focused on these two populations. Review articles and meta-analyses are described that focus on the efficacy of psychological therapies for management of pediatric chronic pain. Next, individual trials that demonstrate specific treatment components and innovations in treatment content and treatment delivery models will be highlighted.

It is important to consider the context of CBT for pediatric chronic pain. CBT may be the sole intervention strategy provided to children, or it may be used in conjunction with other therapies, including medications, physical therapy, and complementary and alternative medicine (CAM) to reduce pain sensations, increase comfort, or reduce associated disability and dysfunction. There is a very limited evidence base for medications, physical therapy interventions, or CAM (p. 26) interventions for the treatment of children with chronic pain. In contrast, as reviewed below, there are several decades of research available on cognitive and behavioral therapies for pediatric chronic pain. Narrative reviews and meta-analytic reviews of psychological treatments for pediatric chronic pain are described.

Narrative Review Articles

Several narrative reviews have summarized findings from controlled and uncontrolled trials of psychological therapies for children and adolescents with headache, abdominal, musculoskeletal, and disease-related pain. These reviews are summarized below in Table 3.1. Three of these reviews were part of a series of papers published in 1999 on empirically supported treatments in pediatric psychology (Holden, Deichmann, & Levy, 1999; Janicke & Finney, 1999; Walco, Sterling, Conte, & Engel, 1999). (p. 27) In this series, the status of the science of psychological treatments was depicted using Chambless’s criteria of “well-established,” “probably efficacious treatments,” and “promising interventions,” following guidelines formulated by the American Psychological Association’s Task Force on Promotion and Dissemination of Psychological Procedures (1995).

Table 3.1 Narrative Reviews of Psychological Therapies for Pediatric Chronic Pain




Quality of Intervention

(Holden et al., 1999)

Recurrent headache

Relaxation, Thermal biofeedback, Multicomponent CBT

  • Well-established

  • Probably efficacious

  • Promising

(Janicke & Finney, 1999)

Recurrent abdominal pain


Probably efficacious

(Brent et al., 2009)

Functional gastrointestinal disorders



(Walco et al., 1999)

Disease-related pain (cancer, rheumatology, hematology)



Note: “Well-established” refers to at least two well-controlled between-group design experiments, conducted by at least two different investigators or investigating teams, finding that the treatment is superior to a pill or psychological placebo, or is equivalent to an already established treatment. “Probably efficacious” refers to at least two waiting-list control studies that demonstrated superior effectiveness; or if “well-controlled” criteria have been met, studies have only been conducted by a single investigating team. “Promising” treatments refers to one well-controlled experiment, two or more well-controlled studies with small numbers, or two or more studies by the same investigator (Chambless & Hollon, 1998; Task Force on Promotion and Dissemination of Psychological Procedures, 1995).

As part of this series, Holden, Deichmann, and Levy (1999) reviewed interventions for recurrent pediatric migraine and tension headache. Based on their review, they described relaxation as a “well-established and efficacious” treatment, and thermal biofeedback as a “probably efficacious” treatment for recurrent pediatric headache. Other multi-component CBT treatment approaches were considered to be promising interventions.

Several reviews of psychosocial interventions for pediatric functional gastrointestinal disorders and abdominal pain have been conducted. Janicke and Finney (1999) identified four case studies, or series, using CBT in children with abdominal pain, finding that CBT met the criteria for a “probably efficacious” intervention. Since that time, more-recent reviews have been published. For example, Brent, Lobato, and LeLeiko (2009) reviewed 12 studies examining psychoeducation, relaxation-based therapies, or multi-component cognitive-behavioral therapy for managing functional gastrointestinal disorders. They reported positive findings overall, but highlighted the need for more well-designed randomized controlled trials.

Psychological interventions have been least studied for chronic disease-related pain in children. A few small uncontrolled trials of psychological interventions for the management of disease pain were reviewed by Walco and colleagues (1999). Specifically, they reported on three case studies of imagery and hypnosis in children with cancer; three case series of breathing, relaxation, and biofeedback in rheumatologic populations; and six case studies or series of hypnosis, imagery, relaxation, and breathing strategies in adolescents and young adults in hematology populations (hemophilia and sickle cell disease). Since that time, Barakat, Schwartz, Salamon, and Radcliffe (2010) conducted a randomized controlled trial of a brief CBT pain intervention using a family support person in adolescents with sickle cell disease. They did not find any changes in pain outcomes in youth receiving the CBT pain intervention in comparison to an education control group. The authors suggested that more comprehensive interventions may be needed in this population. Further empirical studies are needed to evaluate the efficacy of CBT for children with chronic disease–related pain and to understand relevance of particular intervention strategies.

Meta-Analytic Reviews

Table 3.2 below lists findings from five meta-analytic reviews conducted over the past 15 years that pooled data from multiple trials to summarize the evidence base for psychological therapies in the treatment of children and adolescents with chronic pain. The most commonly reported treatment outcome in meta-analytic (p. 28) (p. 29) reviews is clinically significant pain reduction. Treatment success is typically defined by a reduction in pain of 50 percent or greater at treatment completion or short-term follow-up compared to baseline (Eccleston, Yorke, Morley, Williams, & Mastroyannopoulou, 2003; Morley & Williams, 2006); for example, for a child reporting a pain intensity level of six out of 10 at baseline, a 50 percent reduction would be a score of three or lower at post-treatment. Although more conservative rates for determining intervention success have been used in adult populations (such as 30 percent pain reduction), the benchmark for several decades in pediatric trials has been a 50 percent reduction.

Table 3.2 Meta-Analytic Reviews of Psychological Therapies for Pediatric Chronic Pain Management




Effect Sizes on Primary Outcomes

(Hermann et al., 1995)

Headache, n = 12 trials

  • Biofeedback; Biofeedback + progressive muscle relaxation;

  • Multi-component CBT

Large effect sizes ranging from 1.41 for multi-component CBT, 2.57 for biofeedback alone, and 3.09 for biofeedback + relaxation

(Eccleston, Yorke et al., 2003)

Headache, n = 12 trials

Biofeedback, relaxation therapies, CBT

Large effect size for clinically significant pain reduction OR = 9.62

(Trautmann et al., 2006)

Headache, n = 10 trials

Biofeedback, relaxation therapies, CBT

  • Small effect size of g = 0.35 for pain intensity, duration, and frequency;

  • Large effect size for clinically significant pain reduction of g = 0.87

(Eccleston et al., 2009)

Headache, abdominal pain, fibromyalgia; n = 29 trials

Biofeedback, relaxation therapies, CBT

  • Large effect size (OR = 5.51 at post-tx and OR = 9.91 at 3 months) for clinically significant pain reduction;

  • Small effect sizes for mood (SMD = 0.07) and disability (SMD = 0.37)

(Palermo et al., 2010)

Headache, abdominal pain, fibromyalgia; n = 25 trials

Biofeedback, relaxation, multi-component CBT

  • Large effect size for clinically significant pain reduction (OR = 5.92 at post-tx and OR = 9.88 at 3 months follow-up);

  • Small effect sizes for disability (SMD = 0.24) and emotional functioning (SMD = 0.12)

Note: Standardized effect sizes are used in meta-analyses to convey the average difference between groups. Continuous data are typically presented using mean difference effect sizes (such as Hedge’s g, Cohen’s d, or standardized mean difference [SMD]) while binary outcomes are presented using odds ratios. Odds ratios (OR) can be interpreted as the chance of achieving the outcome; for example, an OR of 5.3 means that the treatment group has a five times greater chance of clinically significant pain reduction compared to the control group.

The first meta-analysis to summarize psychological treatments for pediatric migraine was conducted by Hermann, Kim, and Blanchard (1995). They reported on 17 behavioral treatment studies using controlled or multiple-baseline treatment designs and 24 pharmacological treatment studies. Overall, thermal biofeedback and biofeedback combined with progressive muscle relaxation had the largest effect sizes, and when directly compared to other behavioral and pharmacological treatment, these biofeedback treatments outperformed pharmacological interventions (Hermann et al., 1995), suggesting excellent efficacy of psychological interventions for pain reduction in children with migraine.

Two more recent reviews also used data-pooling techniques to summarize psychological treatments administered to children and adolescents with headache (Eccleston, Morley, Williams, Yorke, & Mastroyannopoulou, 2002; Trautmann, Lackschewitz, & Kroner-Herwig, 2006). In their first published review for the Cochrane Collaboration (available at, a library of systematic reviews in health care and policy, Eccleston and colleagues (2003) found that, in 12 randomized controlled trials (RCTs) of children with headache, psychological treatments were effective in reducing pain intensity, with a large and significant effect. Trautmann, Lackschewitz, & Kroner-Herwig (2006) conducted a meta-analysis including 23 RCTs of psychological treatment for recurrent headache in children; 10 of the 23 trials allowed for a between-group comparison on headache outcomes. The mean effect size across three headache outcomes, frequency, duration, and intensity, was small (g = 0.35). However when considering clinically significant improvement in pain intensity at post-treatment, they found a large and significant effect size. Success rates of 70 percent for treatment conditions and 30 percent for control conditions were found for clinically significant pain reduction.

Eccleston and colleagues (2009) performed an update to their 2003 Cochrane review to add new trials and to specifically examine mood and disability outcomes in addition to pain reduction. Thirty-four RCTs (n = 29 in the meta-analysis) were included. Findings revealed large treatment effects for pain reduction, both for children with headache pain as well as for children with non-headache pain (abdominal pain, fibromyalgia). However, for disability and mood outcomes, small and nonsignificant effect sizes were found in the few trials that reported on these outcomes.

More recently, Palermo and colleagues (2010) adapted the updated Cochrane review and published a comprehensive meta-analysis of trials of psychological (p. 30) therapies conducted over the previous 25 years. Only RCTs reporting on pain, disability, or emotional functioning outcomes were included in the meta-analysis, resulting in 25 trials with 1,247 participating children. Most treatments were brief, averaging just 6.4 hours of treatment contact, and were conducted either individually or in group sessions. Most studies tested multi-component CBT packages, with relatively fewer trials of relaxation and biofeedback therapies. Meta-analytic findings demonstrated a large positive effect of psychological intervention on pain reduction at immediate post-treatment (odds ratio = 5.92) and at three-month follow-up (odds ratio = 9.88) in children with headache, abdominal pain, and fibromyalgia. Small and nonsignificant effect sizes were found for improvements in disability and emotional functioning, although there were limited data on these outcomes available in the included studies. The type of intervention strategy was examined by coding trials into three categories: multi-component cognitive-behavioral therapy, relaxation therapy, and biofeedback. All three types of interventions produced significant and positive effects on pain reduction, suggesting similar effectiveness. Studies directly comparing the effects of self-administered (at-home) versus therapist-administered (in-clinic) interventions found similar positive effects on pain reduction, suggesting that CBT is effective using different modes of treatment delivery.

Taken together, results from these five meta-analytic reviews are consistent in demonstrating that psychological therapies, particularly CBT interventions, are effective in reducing pain in children with headache, musculoskeletal, and recurrent abdominal pain. Large and robust effects for pain reduction have been found across different types of psychological interventions, and the effects are maintained in the short term after treatment. There is little evidence available to judge the effectiveness of CBT on disability or mood outcomes, and these remain important areas to assess in future clinical trials.

Special Topics in the Evidence Base

Comparative Effectiveness

The relative or comparative effectiveness of different interventions has been examined in studies of headache and abdominal pain in children. Hermann and colleagues (1995) reported on the relative effectiveness of cognitive and behavioral treatments and pharmacological treatments for headache in children. As mentioned above, they found that thermal biofeedback and biofeedback combined with progressive muscle relaxation had effects superior to those of pharmacological treatments in reducing headache pain in children and adolescents. Similarly, in the systematic review conducted by Weydert, Ball, and Davis (2003), a range of treatments for recurrent abdominal pain was reviewed. Findings demonstrated positive effects for CBT relative to pharmaceutical, botanical, and dietary interventions, which had very weak evidence. Such work highlights the value of CBT relative to other forms of intervention for children and adolescents with chronic pain.

(p. 31) Multi-Component Treatments Involving Parents

Parents have been included in CBT in various ways, and the use of specific parent intervention strategies is an important treatment focus in CBT for pediatric chronic pain. In the meta-analysis performed by Palermo and colleagues (2010), seven of the 25 trials incorporated some instruction for parents, primarily around operant strategies to reinforce child adaptive behaviors and minimize negative pain behaviors. A few investigations have demonstrated that CBT can possibly be enhanced with the addition of parent management strategies (Allen & Matthews, 1998; Allen & McKeen, 1991; Sanders, Shepherd, Cleghorn, & Woolford, 1994) and that this combined treatment is effective for reducing pain related to headaches (Allen & Shriver, 1998; Arndorfer & Allen, 2001) and recurrent abdominal pain (Sanders et al., 1994).

However, only a few RCTs using operant parent strategies have been conducted. In the first study of children with recurrent abdominal pain, Sanders et al. (1994) compared a CBT intervention (contingency management training for parents and self-management training for children) to a standard care intervention (office-based pediatric care). Compared to standard care only, in families who received the CBT intervention, children had a higher rate of complete elimination of pain and lower levels of interference in daily activities as a result of pain. A replication and extension of this work was completed by Robins and colleagues (2005), who also found significant treatment effects for CBT using parent operant strategies in reducing the frequency of pain and school absences in a sample of children with recurrent abdominal pain. Similarly, social contingency interventions have been evaluated in pilot studies of children with headache (Allen & Shriver, 1998), finding that children who received biofeedback training plus parent training in pain behavior management experienced greater reductions in headache frequency and greater improvements in adaptive behavior at three-month follow-up compared to children who received biofeedback alone.

In a large trial conducted by Levy and colleagues (2010), 200 children (ages seven to 17 years, with functional abdominal pain) and their parents participated in a brief CBT intervention. These families were randomized to either three sessions of CBT, or three 75-minute sessions of an educational control condition where they participated in the sessions over three weeks. The CBT program included a primary focus on training parents in social learning strategies and training children in relaxation training and cognitive restructuring. These investigators found that children in the CBT condition had greater decreases in pain and symptom severity than children in the education attention–control condition. However, there were no significant differences in disability, depression, or anxiety between groups. Importantly, parents in the CBT condition had greater decreases in solicitous responses to their child’s symptoms, and children reported increases in their own coping skills, compared with parents and children in the control condition. This trial was innovative in that parents were conceptualized as the primary agents of behavioral change, and the study highlights potentially important therapeutic change processes in CBT.

(p. 32) Chapter 7 in this book provides information about interventions with parents. Additional work is needed to better understand which parental interventions are effective in children across the developmental range. To date, operant strategies have been the most consistently applied and tested. However, several other strategies, such as parental communication training, have been incorporated into multi-component interventions (e.g., Palermo et al., 2009).

Inpatient and Residential Treatment

CBT may be used in the residential treatment setting; that is, treatment delivered in an inpatient or day hospital setting. To date, there have been several published case reports and uncontrolled studies of multi-component interventions (involving CBT) in inpatient and residential treatment settings. For example, in a case report (Palermo & Scher, 2001) of inpatient treatment of an 11-year-old girl with severe debilitating pain, a 22-day inpatient hospitalization incorporating rehabilitation modeling and operant behavioral techniques was effective in improving her physical and psychosocial functioning and decreasing the longer-term costs of medical diagnostic testing, medications, and other healthcare use.

The first well-described interdisciplinary day treatment program for adolescents with chronic pain was at the Bath Pain Management Unit, by Eccleston and colleagues (2003). A formal evaluation of the program was conducted, with 57 adolescents with severe chronic pain admitted to the unit for interdisciplinary treatment (rehabilitative and cognitive-behavioral therapy) in a three-week day hospital treatment program. Their findings demonstrated significant improvements at post-treatment in adolescent disability, physical functioning, and emotional functioning. Because the Bath Pain Management Unit has a particularly strong training and research base, it has become a model for programs around the world.

Treatment may also be organized to occur at the inpatient treatment service where children stay in hospital. For example, a relatively large sample who received inpatient treatment in Germany at Vodafone Foundation Institute for Children’s Pain Therapy and Pediatric Palliative Care was described by Hechler and colleagues (2009). Pre-treatment and three-month post-treatment outcomes on 167 adolescents admitted to the program demonstrated significant changes in pain, disability, school absence, and emotional distress from pre- to post-treatment follow-up. In addition, 55 percent of adolescents demonstrated “overall amelioration,” defined by this team as improvement in pain-related disability or number of school absences.

As more residential treatment programs continue to develop, further outcome data will emerge, particularly about which children and adolescents seem to benefit from this form of treatment, the effectiveness of day hospital versus inpatient programs, and whether treatment gains can be maintained over the long term (12 months and longer).

(p. 33) Outpatient Group-Based Delivery of CBT for Chronic Pain

CBT has been delivered in the outpatient setting in group formats to children and adolescents with different types of chronic pain (e.g., headache, abdominal pain). Group-based treatment can be an efficient way to use professional time to reach a larger number of patients, especially for practitioners working in rural areas or who have a small clientele with chronic pain. Although there are no empirical data concerning the social support that children receive from pain management groups, a potential benefit of group-based treatment is to reduce isolation and provide increased feelings of support to children and adolescents and their parents.

Several examples of group treatment for children and adolescents with headache, abdominal pain, and inflammatory bowel disease have been published (e.g., McCormick, Reed-Knight, Lewis, Gold, & Blount, 2010). A wide range of group treatment sessions has been used, from several sessions that are one to two hours in length, to a single full-day session that is six to eight hours long. Often, separate content is delivered to parents and children in small groups. In some programs, time in each session or in select sessions is spent with children and parents to review the skills jointly. Group facilitators are most typically psychologists or psychology trainees.

In one pilot trial of group CBT for children with inflammatory bowel disease (McCormick et al., 2010), 24 adolescents were randomized to treatment or waiting-list control groups. The treatment group received coping skills for reducing pain and somatic symptoms in a one-day workshop. Adolescents were taught disease-related coping skills, pain management, relaxation techniques, and communication, and their parents were taught limit-setting strategies. The treatment day was followed by six weeks of Web-based skill review, including homework assignments and weekly group chat sessions. Following treatment, improvements were found in adolescents’ reports of somatic symptoms and adaptive coping strategies in the group receiving the coping skills intervention.

Similarly, in an uncontrolled pilot trial, Logan and Simons (2010) evaluated a group intervention with adolescents with chronic pain and depressive symptoms that focused specifically on teaching coping skills to improve their functioning in school. Group treatment involved either four two-hour sessions or a one-day workshop, depending upon the family’s preference. After treatment, pain and school attendance improved for the adolescents. Overall participation in the trial was low, however, and these authors speculated that it may have been difficult for families to come to the hospital to access these services. Unfortunately, one disadvantage of group-based treatment is the need to travel to the treatment center.

Group-Based CBT in the School Setting

The pioneer of group-based treatment of chronic pain in the school setting is Dr. Bo Larsson, a psychologist in Sweden (Larsson & Melin, 1986; Larsson, Melin, Lamminen, & Ullstedt, 1987). (p. 34) Treatment at school reduces multiple barriers that families face in accessing treatment in their local communities. In a series of studies focused on relaxation training for children and adolescents with headache, Larsson and colleagues found that adolescents could successfully acquire relaxation skills in brief group sessions of 20 to 30 minutes over several weeks. These studies demonstrated that children and adolescents receiving group-based relaxation treatment had significant reductions in their headache pain. Importantly, interventions were successfully delivered not only by psychologists, but also by school staff (e.g., school nurses).

Despite the success shown in delivering behavioral pain-management interventions in the school setting, there has been limited research in the context of schools. The school is a setting with a great potential for a cost-efficient mode of reaching a broad population of young people. School psychologists are also in a position to intervene directly with children in the school. It is surprising that psychologists have not yet developed approaches to pain prevention that are delivered in the school; this represents an important future opportunity. (Other school interventions are discussed in Chapter 9.)

Distance Treatment and Computer-Based Applications

Another way of addressing the barrier of geographical distance to treatment centers has been the application of distance treatment methods using written manuals, telephone counseling, and more recently, computer-based applications of CBT. Minimal therapist contact, and self-administered treatments (e.g., manual and relaxation tapes used at home) focused on relaxation and CBT for children with headaches were found to produce results equal to or better than equivalent therapist-led treatments (Burke & Andrasik, 1989; Kroner-Herwig & Denecke, 2002; McGrath et al., 1992). In one study of relaxation therapy, the self-administered treatment was less than one-third as costly compared to the in-office treatment (Larsson, Daleflod, Hakansson, & Melin, 1987).

In the meta-analysis conducted by Palermo and colleagues (2010), a subgroup analysis was reported for the efficacy of distance and computer-based applications on clinically significant pain reduction. This analysis found large and significant effects for pain reduction when treatment was delivered in the home; the odds ratios were similar (suggesting same-sized effects) to those found for clinic-delivered treatments. Similarly, two small pilot studies using innovative technologies to deliver minimal-therapist-contact psychological treatments for children with chronic or recurrent pain were examined separately. The first trial was conducted by Hicks, von Baeyer, and McGrath (2006) to evaluate the efficacy of a CBT distance treatment among children ages nine to 16 years with recurrent headache or abdominal pain. The intervention was delivered via the Internet (reading material online) and through telephone counseling with a therapist. These investigators found significant reductions in pain levels for children receiving the distance treatment, versus a standard-medical-care control group. The second trial was conducted by Connelly and colleagues (2006), who evaluated (p. 35) a CD-ROM pain management program (focused on relaxation training) in children ages seven to 12 years with recurrent headache. They also found significant improvements in headache activity (less intense and less frequent pain) in children who received the four-week CD-ROM intervention, versus children who received standard care.

Recently, several trials have been conducted specifically using interactive, self-guided Internet interventions for children with chronic pain. These are behavioral interventions that are transformed for delivery on the Internet using a variety of website components such as interactivity, personalization, and feedback and monitoring. Palermo developed and evaluated an interactive, self-guided family CBT Internet intervention in adolescents with chronic pain (Long & Palermo, 2009; Palermo et al., 2009). This program, “Web-based Management of Adolescent Pain” (Web-MAP), uses a travel theme, where children and parents journey through eight destinations covering education about chronic pain, training in behavioral skills (e.g., deep breathing and relaxation), cognitive skills (e.g., positive thoughts), and parental interventions (e.g., communication, operant strategies). In a randomized controlled trial with 48 children and adolescents (ages 11–17 years) and their parents, a significant reduction in activity limitations and in pain intensity in children receiving the online family CBT program was found in comparison to a waiting-list control condition.

In Germany, Trautmann and Kroner-Herwig (2010) randomized 65 children and adolescents with recurrent headache to one of three treatment conditions delivered on the Internet: multimodal CBT, applied relaxation, or an educational control intervention. Findings demonstrated that all three groups showed significant reductions in headache frequency, duration, and pain catastrophizing immediately after treatment. More children and adolescents in the CBT condition achieved clinically significant pain reduction (63 percent) than did the children in the other two treatment conditions at post-treatment (but not at the six-month follow-up). However, the study was limited by a high dropout rate in all three groups, resulting in small numbers available for analyses at the six-month follow-up.

Clearly, innovations in treatment delivery are important to consider in this patient population. Children and adolescents with chronic pain face considerable barriers to receiving specialized pain care in their home communities. Distance treatment methods offer feasible and effective ways to deliver CBT to children and adolescents with chronic pain, and their parents. The Internet offers opportunities to extend the reach of face-to-face treatment and to provide access to care to those who cannot receive treatment in person, and it is clearly an emerging treatment focus. Over the coming years, there will be greater attention focused on innovative treatment delivery platforms (e.g., Internet, smartphone applications) for children with all forms of chronic pain.

Opportunities for Future Treatment Development

The knowledge base for the efficacy of CBT for pediatric chronic pain management continues to accumulate. The following are important areas for future (p. 36) research. Clinicians may also take stock of these gaps and limitations in interpreting the evidence base of CBT for pediatric chronic pain.

  1. 1. Most of the clinical trials performed with children with chronic pain have had small sample sizes (an average of only 20 children per treatment condition) and have been conducted at one treatment center, limiting their power to detect treatment effects and generalize the findings. Larger, multi-center trials of CBT will extend the questions that can be answered in the field.

  2. 2. Very few clinical trials have measured children’s functional outcomes (e.g., physical function, school attendance, etc). Recent trials have focused more comprehensively on a range of outcome domains, and published recommendations are now available for domains and measures to consider for clinical trials in children with chronic pain (McGrath et al., 2008). Therefore, knowledge concerning functional outcomes is expected to increase substantially and is an important research priority.

  3. 3. There has been limited specific therapeutic content focused on reductions in disability. Many CBT interventions (e.g., biofeedback, relaxation) were developed for the purpose of altering pain perception. Fewer specific CBT interventions have been described for the purpose of enhancing children’s daily function, which may be achieved either directly (through specific strategies such as behavioral exposure) or indirectly (through parent operant strategies). Better linking of theory to specific CBT interventions may address this important gap in our knowledge.

  4. 4. Most of what is known about CBT outcomes applies to the period of time immediately following six or eight weeks of treatment. Extended follow-up over longer time periods (six to 12 months) will help us to better understand whether children are able to sustain these improvements.

  5. 5. Few trials of psychological therapies for chronic pain have identified predictors of children’s treatment response; that is, which patients will benefit from treatment. Large, multicenter trials with sufficient sample sizes will enable progress to be made toward understanding sociodemographic or individual predictors of children’s treatment response.

  6. 6. There is not yet information available on what strategies are necessary for behavior change or symptom improvement, or what processes influence change. Building measurement of change processes into the design of new CBT trials will help close this important gap in our knowledge.

  7. 7. The role of CBT in the interdisciplinary management of children and adolescents with chronic pain has received limited research attention. There are opportunities for designing trials to understand the interplay of CBT with other forms of treatment, such as medications and physical therapy.

  8. 8. Brief CBT (e.g., a one-day workshop) has been gaining support for delivering coping-skills training. Future studies are needed to understand what intensity of treatment is necessary to achieve positive changes.

  9. (p. 37) 9. It is not clear at this time which parental interventions are most effective, and for children at what ages. Future studies are needed to understand the role of parental thoughts, beliefs, and behaviors in influencing the treatment process.

Summary and Conclusions

Psychological treatments have demonstrated excellent efficacy in treating children and adolescents with different recurrent and chronic pain conditions. Relative to all other forms of intervention for pediatric chronic pain, CBT has produced the most positive treatment outcomes. Clinicians who work with children with chronic pain can convey optimistic news to children, families, referring providers, and insurers that CBT has a robust evidence base. Many children experience clinically significant pain reduction with CBT, and variations in treatment content and delivery methods have thus far resulted in effective treatments. (p. 38)