1.1. Background

Attention deficit hyperactivity disorder (ADHD) is the single most prevalent behavioral and mental health problem in youth. Approximately 10 percent of U.S. children have received a clinical diagnosis of ADHD.¹ Clinical diagnoses have increased steadily over time,² though the higher rates may be attributable to changing clinical practices (including changes in diagnostic criteria, awareness, clinical practice guidelines, and educational policies that motivated clinical assessment and diagnosis), rather than to an increase in true population rates. The prevalence of ADHD based on rigorous diagnostic procedures is approximately 5.3 percent, a rate that is similar across geographic regions worldwide and that has remained constant over more than 20 years when diagnostic criteria have remained constant.³ This rate, when compared with the much higher rates of clinical diagnoses, suggests that a large number of youth may be receiving a diagnosis when they should not be. Alternatively, the increasing rates of diagnosis could represent the clinical recognition of youth who have clinically significant and functionally impairing ADHD symptoms but who may not meet full, formal diagnostic criteria,⁴ since increasing evidence suggests that ADHD symptoms are continuously distributed quantitative traits and therefore lie on a continuum of severity in the general population.^5–7 Some youth, however, are misdiagnosed as having ADHD when they in fact have symptoms of other disorders that are similar to, or overlap with, the symptoms of ADHD – difficulty concentrating, for example, is a symptom that occurs in many other conditions.⁸ ADHD is more than twice as likely to be diagnosed in boys than in girls,¹ though this sex-specific difference in prevalence is thought to derive at least in part from diagnostic biases and cultural influences, in addition to true underlying biological determinants.^{9, 10} ADHD is a more prevalent diagnosis in youth from low-income families¹¹ and in Caucasian compared to Black, Hispanic, and Asian youth,¹² although diagnostic bias, ethnocentrism, and cultural influences may again contribute to these socioeconomic, ethnic, and racial disparities in diagnostic rates.^{13, 14}

The first question patients, parents, teachers, and clinicians ask when considering ADHD is, “Does this child truly have ADHD?” Unfortunately, clinician judgement, especially by non-specialist clinicians in primary care, is poor in diagnosing ADHD¹⁵ compared with expert, research-grade diagnoses by mental health clinicians.¹⁶ Accurately identifying youth who have ADHD has proved difficult at a population level, in part because diagnoses are often made using subjective clinical impressions and limited diagnostic tools. These tools include structured and semi-structured parent, youth, and teacher questionnaires. They represent an improvement over unsupported clinician judgement, but they are nevertheless highly subjective, prone to disagreement across eporters,¹⁷ and likely overestimate the prevalence of ADHD.^{18, 19} More objective diagnostic tools have been proposed, including activity monitors,²⁰ neuropsychological test measures,^21–24 biomarkers such as genotyping,²⁵ electrophysiological indices,^{26, 27} and magnetic resonance imaging (MRI) measures,^{28, 29} though they are not yet established diagnostic tools.

It is essential to know how the comparative accuracy of these diagnostic tools varies by clinical setting, including primary care or specialty clinic, and/or patient subgroup, including age, sex, socioeconomic status, racial or ethnic group, co-occurring mental, emotional, or developmental disorders, or other risk factors associated with ADHD. The accuracy of an ADHD diagnosis is thought to be especially poor in preschool-aged children, for whom hyperactivity, general rambunctiousness, and difficulties with impulse control are often relatively normative and difficult to distinguish from ADHD-related behaviors. Preschool youth also typically do not have the same classroom expectations for behavioral self-regulation that are expected of children in elementary school,³⁰ further obscuring the distinction between ADHD and neurotypical early childhood behaviors. Numerous population-based studies have found that the youngest children in a school year are much more likely to be diagnosed as having ADHD or to receive ADHD medication than their older classmates.³¹

ADHD diagnosis is normally based on an assessment to determine whether the patient meets the criteria described in the Diagnostic and Statistical Manual of Mental Disorders, 5^th edition, text revision (DSM-5-TR).³² Rating scales, which can be completed by parents, teachers, and/or patients, are used to evaluate the frequency and severity of each of the 18 symptoms in DSM-5-TR³² (9 symptoms related to inattention, and 9 symptoms related to hyperactivity/impulsivity), as well as the degree of symptom-related impairment across settings (e.g., home, school, work). Rating scale data are integrated with a clinical interview to determine the onset, course, duration, and impairment associated with symptoms. In addition, screening and clinical evaluation of potential co-occurring psychiatric conditions is a key part of the diagnostic process. Important questions remain about the accuracy of this approach in primary care settings. A particular challenge opularng ADHD from other conditions that may appear similar (e.g., anxiety, conduct disorders) and determining whether another condition may better explain ADHD symptoms or is present as a co-occurring diagnosis. Co-occurring problems are the rule, as approximately half of youth with ADHD are diagnosed with an oppositional defiant or conduct disorder diagnosis, one-third have an anxiety disorder, and 20 percent have depression.²

Inaccurate diagnoses of ADHD can lead either to the administration of treatments, usually stimulant medications, in children who do not need them, or to the withholding of treatment and services for those who would benefit from such treatments.^{30, 33} Prescription of stimulant medications across the U.S. population has doubled in the last decade,³⁴ with a prevalence in 2019 of approximately six percent, and as high as 14 percent regionally.³⁵ These rates are higher than the 5.3 percent population prevalence of rigorously diagnosed ADHD,³⁶ suggesting that many youth may be receiving stimulants when they do not have ADHD.^{36, 37} These trends have created alarm in the lay public, policy makers, and healthcare providers.^{37, 38} Adding to their concern is that diversion and abuse of stimulants is common, particularly in college students³⁹ and not infrequently by parents.⁴⁰ Little is known or understood about how the risk for diversion and abuse of stimulant medications approved for ADHD varies with patient characteristics (e.g., as a function of age, race/ethnicity, or socioeconomic status). Conversely, only about half of U.S. children who receive a clinical diagnosis of ADHD are treated with stimulants,⁴¹ suggesting a large number of children are not receiving medication when perhaps they should be.

Additional important consequences of an incorrect diagnosis can include stigmatizing youth unnecessarily with a diagnosis of ADHD^{30, 42} on the one hand (i.e., “labeling harms,” which can impair self-esteem or reduce future educational attainment or career opportunities^43–45 or failing to provide a correct diagnostic framework for appropriate, timely, and evidence-based interventions on the other. Misdiagnosis of ADHD not only leads to its overdiagnosis or underdiagnosis, but it can also can lead to incorrectly diagnosing as ADHD other conditions that share symptoms with ADHD (e.g., anxiety, conduct disorders, bipolar disorder, complex trauma, difficult home environments, attachment problems, sleep disturbances, other medical disorders/diseases, speech or language delay, or developmental disorders).^46–49 Thus, treating disorders misconstrued as ADHD may withhold appropriate psychosocial and psychological therapies for those conditions and instead inappropriately treat them with stimulants and other ADHD therapies that may have little or no effectiveness in treating those conditions.

Once a diagnosis of ADHD is made, patients and their parents ask, “What treatment should be undertaken?” The answer to this question is challenging for most clinicians and requires a detailed and accurate understanding of the comparative safety and effectiveness of pharmacologic and behavioral treatments for improving not only the immediate symptoms of ADHD, but also the long-term impact that ADHD has on academic and occupational success, mental health, substance abuse, and conduct or antisocial behaviors.⁵⁰ This answer, however, is always conditioned on characteristics of the individual child or the child’s environment that are known to modify response to treatment. These “tailoring variables” can include patient age, ADHD presentation (primarily inattentive, hyperactive/impulsive, or combined), socioeconomic status, race and ethnicity, prior trauma history, co-occurring conditions (e.g., depression or anxiety), family conflict, and biomarker status (e.g., genotype, cognitive testing profile).^{51, 52} Possible benefits of medication must be weighed against risks and side effects. Many parents and clinicians do not have ready access to information that can help them identify and assess these potential risks and whether their child is likely to respond better or worse to any specific possible treatment they might undertake.

Treatment strategies for ADHD are diverse and can be divided into pharmacologic and nonpharmacologic therapies. The main categories of pharmacologic therapies include stimulants (either methylphenidate or amphetamine derivatives) or non-stimulants (norepinephrine reuptake inhibitors, alpha-2 agonists, and antidepressants). The current frontline treatment for ADHD is stimulant medication, with or without combined psychological and behavioral therapies. Nonpharmacologic therapies include psychosocial interventions (e.g., homework, organizational, and social skills training, sleep-focused interventions, dialectical behavior therapy, cognitive behavior therapy, and mindfulness training), school-based interventions (e.g., psychoeducation and expert consultation for class-room based interventions by teachers), cognitive training (e.g., training of working memory, executive function, and motor skills using interactive games and tasks), parent support (e.g., behavioral training for parents, in-home nurse visits, group psychotherapy, telephone-assisted self-help, psychoeducation, and parental friendship coaching), provider interventions (e.g., psychoeducation and training of providers, support for monitoring therapeutic response, and expert consultation) neurofeedback (e.g., learning to modulate electroencephalogram [EEG] activity), nutritional or dietary supplements (e.g., Omega-3, vitamins, herbs), complementary, alternative, or integrative medicine (acupuncture, homeopathy, physical therapy, and chiropractic treatment).

In children over the age of 5, the American Academy of Pediatrics (AAP) recommends stimulants as the first line of therapy.¹⁸ Whether combining behavioral therapy with stimulant medication confers a significant benefit over stimulants alone, or whether nonpharmacologic therapy alone may be effective, is at present unclear. Adverse effects of pharmacologic treatment depend on the specific intervention and may include gastrointestinal symptoms, changes in appetite, slowed somatic growth, and sleep disturbance.⁵³ Treatment can also lead to personality changes or perceived loss of spontaneity. Individuals who are initially misdiagnosed or who have inadequate monitoring may be overtreated with stimulant medications. Overtreatment leads to the risk of treatment with little or no benefit or to unnecessary side effects. Long-term adherence to medication regimens is often poor in youth who have ADHD and can limit the long-term, real-world effectiveness of medication.⁵⁴

Long-term outcomes for both medication and non-medication therapies have been less well studied,⁵³ and little is known about which treatment to begin first and for whom, or how best to sequence treatments for ADHD when the first intervention proves ineffective or insufficient. Recent advances in the development and testing of novel therapies for ADHD warrant a systematic review of their efficacy and effectiveness that will provide information eagerly awaited by the field. These novel therapeutics include cognitive training, game-based digital devices such as EndeavorRx, approved by the Food and Drug Administration (FDA), and neuromodulation techniques such as repetitive Transcranial Magnetic Stimulation and the FDA-approved external Trigeminal Nerve Stimulator.^55–66

Once treatment is begun, the central question is, “Is the treatment working?” The answer to this question is not as straightforward as it may at first appear, as ADHD symptoms and the capacity to compensate for them may vary over time and with circumstance (e.g., school day or weekend, the presence of psychosocial stress), by symptom presentation (e.g., hyperactivity, inattention, impulsivity), and by functional domain (academics, risk-taking behaviors, socialization). Thus, valid and reliable methods are needed to monitor treatment response easily and accurately. If the current treatment is not producing the desired response, or if side effects are limiting the dose of medication prescribed, the final question is what to do next to improve short- and long-term outcomes. For example, is it better to optimize dosing of the current medication, switch to another first-line medication, switch to a second-line medication, add an additional medication, or add an adjunctive psychological or behavioral therapy? And how does a clinician or parent prevent the complete abandonment of treatment, which is exceedingly common, when the first line treatment is ineffective or produces troubling side effects?⁶⁷

After a child is diagnosed with ADHD and an initial treatment strategy is determined, a monitoring strategy is applied to ensure that outcomes are evaluated over time, and modification of treatments are made when needed.⁶⁸ Ideally, repeat monitoring should provide the opportunity to intervene (e.g., modify the treatment) before the undesirable or adverse outcomes associated with ADHD occur or determine whether and which treatment for remains clinically indicated. Several instruments are available to assess treatment response and adverse effects over time, including the Vanderbilt, Conners, ADHD Rating Scale-5, and Swanson, Nolan, and Pelham Rating Scale (SNAP)-IV rating scales. Monitoring may also include assessment of any adverse treatment effects. The frequency of monitoring may depend on the age of the child, the specific treatment, duration of treatment, previous symptoms, co-occurring conditions, and family and healthcare provider preferences. For example, monitoring into adulthood is often desirable or needed, as one-third to one-half of patients with ADHD will have clinically significant symptoms that persist into adulthood. Monitoring for long-term adverse outcomes in domains distinct from ADHD symptom severity is important, since youth with ADHD are at increased risk for future problems associated with risk-taking, such as substance abuse, motor vehicle accidents, unprotected sexual intercourse, and criminal behavior. They are also at considerable risk as adults for chronic health problems, including diabetes, heart disease, and poor oral health, in part because they engage in behaviors that increase risk for these conditions, and they often fail to adhere to health-protective behaviors. They are also at risk for future depression, anxiety, suicide attempts, and problematic peer and family relationships.^{2, 50} In addition, the long-term effectiveness of standard and novel interventions for ADHD, and their potential long-term adverse effects, are not well known^69–73 and are difficult to detect and document for these diverse outcomes,^74–76 even though they are critically important considerations for patients, parents, and clinicians as they make treatment decisions. Knowledge of the ways in which unique patient characteristics modify these short- and long-term treatment outcomes is essential to tailor and personalize care for individual patients.⁷⁷

1.2. Purpose and Scope of the Systematic Review

This review updates prior AHRQ reviews on ADHD.^{11, 53, 78} It builds on the previous reports and will address important gaps in knowledge related to the diagnosis of ADHD, concerns about overtreatment and undertreatment, and conflicting literature about the effectiveness of long-term treatment. The review is especially intended to be a resource for clinicians, researchers, and policymakers, although through them, we hope the review will benefit the many youth who have ADHD, as well as their families and teachers. We anticipate that the analyses and results will be difficult for most parents, educators, and lay persons to understand, although the executive summary, key points, and discussion are intentionally crafted to be accessible to a much wider audience. Finally, this systematic review aims to inform a planned update of the current American Academy of Pediatrics (AAP) clinical guidelines for the diagnosis, evaluation, and treatment of ADHD.

Since the last AHRQ report was published, further diagnostic and treatment strategies have been suggested, warranting an update of the literature. Identified references address predominantly diagnostic questions such as the diagnostic validity of specific tests and suggested diagnostic tools. Furthermore, key studies that provide important information on the diagnosis of ADHD predate the most recent ADHD report. Hence, the current systematic review will include older studies. Searches for studies of diagnostic tools will extend back to 1980, when the diagnosis of ADHD and its diagnostic criteria were first introduced in the DSM as Attention Deficit Disorder with or without hyperactivity (DSM-III).⁷⁹

In addition, since the last AHRQ review, several studies have been published that explore novel interventions, such as game-based cognitive therapy or computer training. Furthermore, key studies that predate the most recent ADHD report provide important information on the treatment of ADHD. Hence, the current systematic review also includes older treatment studies. Searches for studies of ADHD interventions will therefore extend back to 1980, when long-acting stimulants were introduced, heralding the modern era of ADHD pharmacotherapy.

Given that the 2018 AHRQ report on ADHD identified no monitoring study, we removed limits on the search date for this question and will aim for a comprehensive review that considers older studies (the 2018 report included only studies published to 2009). Based on discussions and preliminary literature searches, we still do not expect to identify many studies for monitoring strategies and long-term outcomes, although we anticipated that some data may be available from the educational and school psychology literature, such as Response to Intervention – Behavioral (RTI-B) strategies to monitor behavioral and psychosocial interventions in the classroom that aim to improve ADHD outcomes.

To our knowledge, no prior reviews of ADHD have been as comprehensive as the current review in the range of diagnostic tools, treatments, clinical outcomes, participant ages, and year of publication for the included studies. We hope that it will be a valuable resource for patients, families, clinicians, educators, policymakers, and researchers for years to come.