Treatment of Attention-Deficit/Hyperactivity Disorder: Overview of the Evidence

INTRODUCTION

The American Academy of Pediatrics' (AAP) Committee on Quality Improvement, Subcommittee on Attention-Deficit/Hyperactivity Disorder, developed a clinical practice guideline pertaining to the treatment of school-aged children with attention-deficit/hyperactivity disorder (ADHD).^1,2 The review here covers the additional evidence gathered for and by the subcommittee regarding specific treatments for ADHD. It does not include more recent studies published since the publication of the guideline in 2001. Other evidence supports the notion that ADHD is a chronic health condition meriting the application of general principles of management of childhood chronic conditions by primary care clinicians. Longitudinal studies of Barkley et al³ and Biederman et al⁴ document the persistence of the condition over time. Previous policy statements by the AAP describe the elements of chronic-condition care, including educating parents and children about illness, developing individual treatment plans, helping to coordinate multiple services, and encouraging parents to have contact with other parents of children with chronic conditions. Because ADHD has pervasive effects on the child's daily life, including school performance, a partnership of clinicians and school personnel will help ensure the child's best progress and proper assessment of progress.

This report summarizes the empirical literature on which the practice guideline's recommendation for pharmacologic and/or behavioral intervention was based. Certain other aspects of clinical care have not been the focus of careful clinical trials or randomized, controlled studies to determine accurately optimal practices. Thus, many of these recommendations reflect consensus of best practices. The specific areas included in these consensus recommendations include (1) the best specific ways of titrating a child's medications, (2) the frequency of monitoring visits during titration phases and after stabilization, or (3) the specific content of monitoring. Again, these recommendations reflect assessment of best practices in long-term care.

Four main sources of data were examined in developing these treatment recommendations. The McMaster University Evidence-Based Practice Center (under contract with the Agency for Healthcare Research and Quality and in partnership with the AAP and other organizations) reviewed the short- and long-term efficacy and safety of pharmacologic and nonpharmacologic interventions for ADHD and the comparative efficacy of single versus combined treatments. For a full account of the evidence, see the technical report compiled by the former Agency for Health Care Policy and Research (now the Agency for Healthcare Research and Quality).⁵ The second source was a review of interventions for ADHD conducted by the Center for Community Health Research, British Columbia Research Institute for Children's and Women's Health, and the University of British Columbia⁶ for the Canadian Coordinating Office of Health Technology Assessment. The third source included the findings from the multisite treatment study conducted by the Multimodal Treatment Study for Children With ADHD (the MTA Cooperative Group),^7,8 which is supported by the National Institute of Mental Health.

After reviewing the findings and conclusions of the 3 sources listed above, the subcommittee conducted an additional search of the available literature to assess further the effectiveness of behavioral interventions both as stand-alone regimens and in combination with pharmacologic treatments. This search selected published reports of trials of behavior therapies in groups of school-aged children with ADHD. Individual case reports were excluded. The evidence from this search was compiled and evaluated by the subcommittee.

McMASTER UNIVERSITY EVIDENCE-BASED PRACTICE CENTER REVIEW

The goals of the evidence-based review conducted by the McMaster University Evidence-Based Practice Center Group⁵ were to examine the efficacy of nonstimulant medications and nonpharmacologic interventions for ADHD in children and adults and to examine the comparative efficacy of combined versus individual interventions. The technical review examined (1) drug-to-drug comparisons of specific stimulant medications, (2) stimulants versus antidepressant medications, and (3) comparisons of different forms of the same medication. The stimulant drugs examined were methylphenidate (MPH), dexedrine (DEX), and pemoline (PEM). The review also compared tricyclic antidepressants versus placebo and pharmacologic versus nonpharmacologic interventions. The report also examined long-term studies with a duration of 12 or more weeks. Final categories reviewed were studies examining the treatment of ADHD in adults, treatment combinations, and the adverse effects of pharmacologic interventions. This article reviews only findings pertaining to treatment of ADHD among school-aged children.

The McMaster review selected a total of 92 empirical articles reflecting 78 investigations from a pool of 2405 citations compiled from traditional databases (Medline, Cinahl, HealthStar, PsychINFO, Embase), The Cochrane Library (1997, issue 4), reference lists of articles identified in the previous sources, and additional citations suggested by members of the McMaster research team and partnering organizations. Two reviewers independently rated each article to determine the quality of the methodology used in the study. Studies were included in the evidence-based review if they were randomized, controlled trial (RCTs), involved human subjects, and were published as a full report in a peer-reviewed journal. Studies that included participants with diagnoses other than ADHD (eg, oppositional defiant disorder, conduct disorder) were included in the review only if the study provided a separate analysis for the study participants with ADHD.

A problem identified by the review team and associated organizations was the diversity of outcomes used in these many studies. Some studies used indicators or core symptoms of ADHD; others examined aspects of school or social behavior or behaviors at home. This diversity makes clear comparisons among treatment regimens difficult. The McMaster review noted important methodologic limitations in the numerous studies examining interventions for ADHD spanning a period of more than 25 years. Major limitations included small sample sizes and the use of heterogeneous outcome measures. The review found few studies in most of the study areas.

MULTIMODAL TREATMENT STUDY OF CHILDREN WITH ADHD

The National Institute of Mental Health Collaborative Multisite Multimodal Treatment Study of Children With Attention-Deficit/Hyperactivity Disorder^7,8,66 included 579 children with ADHD (combined subtype) who were assigned randomly to 1 of 4 treatment groups: state-of-the-art medication management, intensive behavioral intervention, a combination of the 2 interventions, and a community treatment control group who received “usual care” (most commonly medication). Outcomes were assessed in multiple domains and included measures reflecting the core symptoms of ADHD as well as measures of co-occurring problems in social skills, parent-child relations, oppositional defiant behavior, internalizing behavior problems (eg, anxiety), and academic achievement. Outcome data reflected assessments conducted while children in groups involving pharmacotherapy were still receiving medication, although the behavioral interventions in the behavioral and combined treatment groups had been completed.

In terms of treatment outcome, all 4 treatment groups showed improvements over time, with medication management and the combined intervention associated with greater improvement than the intensive behavioral intervention alone and the community treatment control group, when considering the core symptoms of ADHD.^7,8,66 However, only families assigned to the combined treatment group showed consistently greater benefits than the families in the community treatment group across other outcomes domains (eg, disruptive behavior, parent-child relations, social skills).^7,8,66 Children with ADHD who had coexisting anxiety disorders responded well in both of the treatment groups that included the intensive behavioral interventions.^67,68

Treatment outcome was also examined against a broad composite outcome measure that aggregated treatment responses across a broad array of symptom and functional domains including internalizing (ie, anxiety, depression) and externalizing (opposition, aggression) symptomatology and social skills.⁶⁹ The investigators added this analysis to address the concern that measures that assess primarily core symptoms of ADHD may be less sensitive to the effects of behavioral intervention. Using this composite measure as an outcome variable, analyses revealed that children who received the combined intervention exhibited the best treatment response.⁶⁹

Wells et al⁷⁰ examined the effect of treatment on parent and family stress measures. Data revealed no differences among the 3 study treatment groups and the community treatment control group on measures of family distress (eg, parenting stress, depressive symptoms among parents, marital adjustment). However, compared with the community treatment control group, parents in the medication management, intensive behavioral intervention, and combined treatment groups reported greater reductions in their use of negative or ineffective discipline.⁷⁰ Data were also examined to determine if such parent-reported changes in parenting behavior were associated with teachers' reports of child behavior at the end of treatment.⁷¹ Findings revealed that, at the end of treatment, teachers' ratings of disruptive child behavior fell within the normal range for families that participated in the combined treatment group, and this group reported the greatest reductions in negative and ineffective discipline. This effect was not found for families who participated in behavioral intervention alone.

One component of the intensive behavioral intervention arm of the MTA Cooperative Group study was a summer treatment program. Pelham et al⁷² evaluated 117 children participating in a summer program at 3 of the MTA Cooperative Group sites. Approximately half of these children were assigned to the behavior intervention alone group and half were assigned to the combined treatment group. Children in the behavioral summer program who were also medicated showed a better response to treatment on 5 measures (following rules, good sportsmanship, peer negative nominations, and summer program teacher ratings of ADHD symptoms). However, children responded similarly to treatment regardless of medication status on 30 other measures. For 6 of the 35 measures of child behavior, children in the combined treatment group were more likely than those in the behavioral treatment alone group to fall within the normative range.⁷²

ADDITIONAL SUBCOMMITTEE SEARCH AND REVIEW

The goal of the additional literature search was to identify additional investigational evidence to quantify the effectiveness of behavioral treatment in combination with drug treatment. This review originally provided ∼200 potentially relevant articles, most of which were already included in other sources of information. After excluding case reports and general review articles, the subcommittee formally reviewed 28 articles and included 12 articles of relevance that had not been cited previously in other reviews.^{45,47,51,72–79} Table 1 summarizes the evidence of these reviews. The subcommittee assessed the additional evidence obtained from this review and noted an imbalance in the quality of evidence available for drugs versus behavioral interventions. Although drug regimens were highly specific as to type of drug, dose, and duration of treatment, behavioral interventions lacked this specificity and were less quantified. Investigators did not use identical behavioral interventions and used varying degrees of detail to describe the interventions.† The subcommittee found diversity in the type, intensity, and duration of interventions; for example, some focused on the family setting, and others focused on the school setting. No data directly compared the benefits of these different modalities, such as comparing daily report cards and summer training programs. The costs of these programs varied widely. Subcommittee members noted that these factors made it unlikely that a strong evidence-based recommendation for behavioral intervention per se would be possible.

The subcommittee concluded, on the basis of data from this additional literature review, that drug treatment alone showed a consistent dose-sensitive effect in improving the core symptoms of ADHD. Behavioral interventions alone did not demonstrate statistically significant results. Medication treatment in combination with behavioral interventions was shown to be as beneficial as drug treatment alone. In addition, some studies found that combined modalities yielded 2 additional desirable outcomes: (1) they enhanced teacher and parent acceptance, and (2) they lowered the drug doses needed to achieve the same therapeutic benefits as with drug treatment alone.⁴⁷

CONCLUSIONS

ADHD is a chronic condition that requires ongoing management and monitoring. A robust evidence base attests to the efficacy of stimulant medications in helping to manage the symptoms of ADHD among school-aged children. The stimulant drugs tested seemed equally effective. Tricyclic antidepressants may be effective also but are recommended only when children have been refractory to 2 or more stimulant drugs or have intolerable adverse effects. When considering evidence from RCTs, the data in support of behavioral intervention are less compelling. None of the nonpharmacologic interventions tested were more effective than medication in treating the symptoms of ADHD. Long-standing clinical experience dictates that education and counseling of the patient, family, and school personnel are valuable and necessary adjuncts to drug therapy, as with most long-term treatments for chronic conditions.

ADDENDUM

Atomoxetine is a nonstimulant licensed by the Food and Drug Administration in November 2002 for the treatment of ADHD in children and adolescents.⁸⁰ It is a selective inhibitor of the presynaptic norepinephrine transporter in the central nervous system. Atomoxetine increases both norepinephrine and dopamine levels, especially in the prefrontal cortex. In a randomized, placebo-controlled trial in children and adolescents 8 to 18 years of age over an 8-week period, atomoxetine demonstrated a statistically significant reduction in core ADHD symptoms and improvement in social and family functioning compared with the placebo group.⁸¹ Atomoxetine was compared with MPH in a randomized, open-label trial in children with ADHD during a 10-week study period. Significant improvements in inattentive and hyperactive/impulsive symptom domains were similar with both medications when assessed by parents and clinicians. Adverse effects were also similar (appetite suppression, initial weight loss), with the exception that atomoxetine does not cause or worsen insomnia but in the early phase can cause drowsiness.⁸² Atomoxetine treatment was associated with small but statistically significant increases in mean systolic pressure in adults and diastolic pressure in children and adolescents.⁸³ Blood pressure and pulse tended to increase early in therapy, then stabilized, and returned toward baseline after drug discontinuation. There was no significant difference as revealed by electrocardiogram between atomoxetine and placebo groups in change in QT interval for all study populations. Discontinuations because of cardiovascular-related events did not occur in the child/adolescent group.

Atomoxetine has a slower onset to action than do stimulants; thus, effects may not be seen until the end of the first week of treatment, but atomoxetine seems to have a longer duration of action after a once-a-day dose with suggestions of symptom relief during the evening and early-morning hours. The treatment effect size (0.71) for core ADHD symptoms is similar when once-daily dosing is compared with twice-daily dosing, and parent ratings document a sustained effect late in the day.⁸⁴ Motor and verbal tics associated with atomoxetine have not been reported.

Committee on Quality Improvement, 1999–2000

Charles J. Homer, MD, MPH, Chairperson

Richard D. Baltz, MD

Gerald B. Hickson, MD

Paul V. Miles, MD

Thomas B. Newman, MD, MPH

Joan E. Shook, MD

William M. Zurhellen, MD

Liaisons

Betty A. Lowe, MD

National Association of Children's Hospitals and Related Institutions

Ellen Schwalenstocker, MBA

National Association of Children's Hospitals and Related Institutions

Michael J. Goldberg, MD

Council on Sections

Richard Shiffman, MD

Section on Computers and Other Technologies

Jan Ellen Berger, MD

Committee on Medical Liability

F. Lane France, MD

Committee on Practice and Ambulatory Medicine

Subcommittee on Attention-Deficit/Hyperactivity Disorder

James M. Perrin, MD, Co-chairperson

Martin T. Stein, MD, Co-chairperson

Robert W. Amler, MD

Thomas A. Blondis, MD

Heidi M. Feldman, MD, PhD

Bruce P. Meyer, MD

Bennett A. Shaywitz, MD

Mark L. Wolraich, MD

Consultants

Anthony DeSpirito, MD

Charles J. Homer, MD, MPH

Wendy S. Freeman, PhD

Liaisons

Karen Pierce, MD

American Academy of Child and Adolescent Psychiatry

Theodore G. Ganiats, MD

American Academy of Family Physicians

Brian Grabert, MD

Child Neurology Society

Ronald T. Brown, PhD

Society for Pediatric Psychology