|Home | About | Journals | Submit | Contact Us | Français|
Because of several recent clinical and regulatory changes regarding Attention Deficit Hyperactivity Disorder (ADHD) in the United States, we quantified changes in ADHD diagnosis and medication management from 2000 through 2010.
We used the IMS Health National Disease and Therapeutic Index™, a nationally representative audit of office-based providers, to examine aggregate trends among children and adolescents under 18. We also quantified how diagnosis and treatment patterns have evolved based on patient and physician characteristics and the therapeutic classes used.
From 2000 to 2010, the number of physician outpatient visits where ADHD was diagnosed increased 66% from 6.2 million [M] (95% confidence intervals [CI] 5.5- 6.9M) to 10.4M visits (CI 9.3-11.6M). Of these visits, psychostimulants have remained the dominant treatment, used in 96% of treatment visits in 2000 and 87% of treatment visits in 2010. Atomoxetine use declined from 15% of treatment visits upon product launch in 2003 to 6% of treatment visits by 2010. The use of potential substitute therapies – clonidine, guanfacine, and bupropion – remained relatively constant (between 5-9% of treatment visits) during most of the period examined. Over this period, the ADHD management shifted towards psychiatrists (from 24% to 36% of all visits) without large changes in illness severity or the proportion of ADHD treatment visits accounted for by males (73%-77%).
In ten years, the ambulatory diagnosis of ADHD increased by two-thirds and is increasingly managed by psychiatrists. The effects of these changing treatment patterns on children's health outcomes and their families are unknown.
Attention deficit hyperactivity disorder (ADHD) is common among children and adolescents in the United States. The prevalence of parent-reported ADHD among children aged 4–17 years increased 21.8% during 2003–2007, from 7.8% to 9.5%, representing 5.4 million children (1). The disease also poses a considerable economic burden on children and families, both due to the direct costs of diagnosis and treatment as well as indirect and downstream costs such as lower educational attainment, occupational instability, and lost income (2, 3, 4).
During the past decade several important regulatory and clinical changes regarding ADHD have occurred without a clear analysis of the effect on diagnoses, treatments, and practice patterns. In 2000 the American Academy of Pediatrics (AAP) issued its first clinical practice guideline for the diagnosis and evaluation of ADHD (5). One year later guidelines for the treatment of school-aged children were released (6) with the next update occurring nearly a decade later (7). In addition, new pharmacotherapies have been brought to market, including atomoxetine (Strattera), a norepinephrine reuptake inhibitor initially developed to treat depression but later found to have efficacy in treating ADHD and approved for its use in 2003 (8). Thus, clinicians currently have a variety of pharmacotherapeutic options available, including psychostimulants - methylphenidate and amphetamine and their derivatives – atomoxetine and substitute therapies such as clonidine, guanfacine, and bupropion. Finally, during the past decade safety concerns have been identified regarding many of these therapies, leading the Food and Drug Administration (FDA) to issue a series of communications regarding cardiovascular events associated with amphetamine and dextroamphetamine (Adderall) (February 2005), sudden death and suicidal ideation with atomoxetine (September 2005), potential cardiovascular risks or adverse psychiatric symptoms with all approved ADHD medications (February 2007), and exacerbation of behavior and thought disturbances in patients with pre-existing psychosis (June 2008). Also in 2008, concern regarding ADHD medication use in children with some underlying cardiovascular abnormalities led to statements from the American Heart Association and the American Academy of Pediatrics suggesting careful assessment of children for heart conditions who need pharmacotherapy (9,10), though the necessity of such assessments remains debated (11,12).
The effect of the recent ADHD regulatory and professional society advisories as well as fluctuations in market factors on ADHD management and treatment is largely unknown. We used a nationally representative audit of office-based providers to quantify changes in ADHD diagnoses, medication treatments and other practice patterns over the past decade. In addition to describing aggregate changes in diagnosis and medical treatment, we examined whether there were changes in ADHD management based on patient characteristics (e.g., illness severity), physician specialty, and therapeutic classes used (13,14).
We quantified diagnosis and treatment patterns among children and adolescents less than 18 years of age using the IMS Health National Disease and Therapeutic Index™ (NDTI). The NDTI was established in 1958 by IMS Health, a provider of information services for the health care industry, and provides nationally representative diagnostic and prescribing information on patients treated by non-federally employed U.S. office-based physicians. As of 2010, the collection of NDTI data uses a two-stage sampling procedure and includes 4,300 physicians randomly selected from the American Medical Association and American Osteopathic Association master files within strata defined by specialty and geographic area. Information regarding specialty affiliations is derived from self-report as well as secondary rosters from professional societies. To ensure that all workdays in a report period are covered, participating physicians are randomly assigned to record audit information on all patient contacts during two consecutive workdays in each calendar quarter. The majority of contacts (approximately 85%) occur in office-based settings; the audit also captures information based on patient contacts that may occur by phone or during physician visits to patients in hospitals and nursing homes. For this study, we focused on office-based care alone.
The NDTI generates approximately 350,000 annual contact records. Though the number of annual contact records as well as specialty composition varies year to year, the application of sample weights allows for yearly comparison of nationally representative estimates. For each record, physicians record all applicable diagnoses and then for each diagnosis record the specific medications used to treat that condition. This allows for a direct correspondence between a drug's use and a specific clinical application. Drug reporting reflects the physician's best knowledge of new or continuing medications. Illness severity is coded as mild, moderate, or severe, according to providers’ global assessment. Each patient encounter record contains information specific to the encounter organized by diagnosis, as well as additional data specific to the physician and his/her practice. Our prior work and other investigations comparing the NDTI against the most similar publicly available data source, the National Ambulatory Medical Care Survey, a nationally representative survey of office-based providers conducted by the National Center for Health Statistics, suggest that the two audits are similar in breadth and scope (15,16,17).
We used descriptive statistics to examine trends in the diagnosis and treatment of ADHD from 2000 through 2010 among children less than 18 years of age. We defined ADHD using diagnostic codes for “Attention Deficit Disorder” and “Overactivity not otherwise specified” which are similar to ICD-9 code 314-Attention Deficit Disorder and ICD-9 code 312.01-Attenion Deficit with Hyperactivity, respectively. We focused on four groups of conventional therapies: (1) stimulants, (2) atomoxetine, (3) three therapy alternatives or substitutes, clonidine, guanfacine and bupropion, and (4) antipsychotics. Stimulants are the mainstay of therapy. We included atomoxetine because it is both a FDA-approved treatment for ADHD and its market debut occurred during the study timeframe. The alternative therapies and antipsychotics were chosen based on clinical expertise and confirmed by review of the top medications prescribed for ADHD that were neither stimulants nor atomoxetine. We calculated 95% confidence intervals for estimates using tables of relative standard errors that account for the NDTI's two-stage stratified cluster sampling design.
We report both total visits where ADHD was coded as a diagnosis and, of these, total visits where one or more treatments was mentioned, which we refer to as treatment visits. Because individuals may have received more than one pharmacotherapy during a clinical encounter, a single office visit may generate more than one treatment visit, and thus the total treatment visits such as those depicted in Table 2 may exceed 100%. Data on non-pharmacological treatments, including behavioral therapies, counseling, or alternative therapies, were not available. All data used is de-identified and the study was exempted from Institutional Review Board review.
Figures 1 and and22 depict trends in diagnosis and overall medical treatment rates. Figure 1 shows the overall trend over time in monthly ADHD treatment visits, and includes key events such as atomoxetine's market entry and a series of FDA advisories regarding risks associated with amphetamine and dextroamphetamine (Adderall), atomoxetine, and all ADHD medications. As shown in Figure 2, visits where ADHD was coded increased steadily from 6.2 million [M] visits in 2000 (95%; CI 5.5M – 6.9M) to 10.4M (95% CI; 9.3M-11.5M) in 2003, and then reached a plateau between 9.5M to 10.6M visits through 2010 (95% CI; 9.4M-11.56M). Of these, the fraction of visits where pharmacotherapy was used as treatment ranged from 93-96%. Thus, only a small proportion of visits where ADHD was diagnosed did not result in pharmacological treatment in this audit.
Visit characteristics for individuals receiving treatment did not vary markedly over the study period (Table 1). For example, the fraction of treatment visits accounted for by males ranged from 73% to 77% and mild severity fluctuated between 19% and 23% throughout the period of observation. There was a modest trend towards visits among older subjects. There was also a modest trend toward increasing treatment visits paid for by Medicaid and decreasing visits paid for by private insurers. The percentage of total treatment visits that were first visits ranged from 12% to 14% across the years examined among all subjects as well as analyses limited to specialists.
Of ADHD-coded visits resulting in treatment, the fraction receiving stimulants decreased from 98% (2001) to 87% (2010) and the share receiving atomoxetine decreased from 15% when the drug entered the market in 2003 to 6% in 2010 (Table 2). Between 2000 and 2010, 5%-13% of coded visits wherein patients were prescribed medical therapy were treated with clonidine, guanfacine, or bupropion. The share receiving these substitute medications increased to a high of 13% in 2010 following the introduction of extended release guanfacine (Inutniv). Between 2000 and 2010, the share receiving antipsychotics ranged from 1%-4%. “Other” therapies, such as mood stabilizers, antidepressants, vitamins, hormones, and antihistamines, represented between 4% and 7% of treatment visits, and were often used in conjunction with our therapies of interest.In the majority of visits, therapies to treat ADHD were used alone rather than in conjunction with another psychotropic medication. For example, in 2010, of visits coded with ADHD treatment, 87% prescribed stimulants, and of these, 83% prescribed stimulants alone rather than jointly with another medicine. Rates of joint psychotropic use with stimulants increased from 7% of all stimulant users (2000) to 18% (2004) before declining to 13% (2010). Rates of long-acting stimulant medication use increased from 14% (2000) to 87% (2010) while short acting stimulant medication use simultaneously decreased from 86% (2000) to 13% (2010).
Between 2000 and 2010, the proportion of visits coded with an ADHD diagnosis increased for psychiatrists and decreased for pediatricians. For example, in 2000, approximately one-fourth (24%) of all visits where ADHD was coded occurred during care by a psychiatrist. By 2010, more than one-third (36%) of diagnosis visits were by psychiatrists, while the share of ADHD diagnosis visits by pediatricians decreased from 54% to 47% (Table 3). There were similar shifts in ADHD treatment towards specialists.
In this analysis of a nationally representative audit of office-based providers between 2000 and 2010, we found large increases in the number of visits coded with an ADHD diagnosis, modest decreases in the rates of stimulant use, and a gradual shift towards psychiatrists in place of pediatricians in diagnosing and treating ADHD. These findings complement and update other investigations documenting the growth of stimulants among children and adolescents in the late 1990s (18,19) and a more recent report limited to those with commercial insurance indicating increasing ADHD diagnosis and treatment prevalence between 2000 and 2005 (20).
Several factors may account for the observed increase in the number of visits coded with an ADHD diagnosis that occurred primarily during the first half of the decade. Advocacy efforts and publicity campaigns (21) may have led to greater public awareness. Coupling this with increasing provider knowledge of ADHD from the dissemination of clinical guidelines and continuing medical education may have contributed to an increased likelihood of parents, children, and providers identifying behavioral and conduct disorders as ADHD. In addition, the availability of an increasing number of treatments (including atomoxetine) (22), as well as marketing and promotion on the part of pharmaceutical firms (23), may also have fostered the increasing number of diagnoses. Given the short time frame examined and the substantial fluctuation over short periods (e.g., a single year), a change in the true prevalence of this disorder is unlikely to account for the findings.
Our analysis was an office-based, rather than a population-based sample, and as such, should not be used to derive population-level estimates of the frequency of the disorder or its treatment. Similarly, given the nature of this ambulatory audit, these data likely underestimate the proportion of individuals receiving non-pharmacologic therapies. Nevertheless, the increases in the number of visits coded with an ADHD diagnosis that we describe are supported by a recent CDC report based on the National Survey of Children's Health, a telephone survey that also suggested increasing rates of ADHD diagnosis among children 4 to 17 years of age, from 7.8% in 2003 to 9.5% in 2007 (1).
We also found substantial decreases in use of atomoxetine from the year of its market debut, 2003, through 2010. Initially atomoxetine was celebrated by some as an important alternative for ADHD treatment because it allowed 24-hour dosing and, as a selective norepinephrine reuptake inhibitor, was not a controlled substance and reportedly would have less abuse potential. However, clinical trials of atomoxetine suggested smaller effect sizes than those observed with stimulant treatments (24). The decline in atomoxetine prescriptions also occurred during a period where scientific reports, FDA communications, and media coverage all generated awareness among the public and providers of potential adverse events associated with stimulant and atomoxetine use. However, these declines in stimulant and atomoxetine use were not accompanied by large increases in the use of alternative medications such as clonidine, guanfacine or buproprion. The reasons for the lack of increase in the use of potential substitutes may have included the absence of any industry promotion for their use as well as clinicians’ lack of faith in their effectiveness.
While the majority of children and adolescents with ADHD are still managed by primary care physicians, the results also show a substantial shift away from pharmacological management of ADHD by primary care physicians (e.g., pediatricians) to specialists (e.g., psychiatrists). We found no evidence that primary care physicians were electing to forgo pharmacologic treatment at higher rates. Instead, we found a shift in the proportion of all ADHD visits accounted for by specialists. Growing parent and provider awareness of the potential risks of prescription drugs used to treat ADHD may have contributed to the increasing fraction of all diagnosis and treatment accounted for by psychiatrists. Such a shift to specialists is plausible and has been previously demonstrated with antidepressants as regulatory advisories and safety concerns among children increased (13). Given the short supply of pediatric psychiatrists or psychiatrists specializing in ADHD relative to general pediatricians, increasing responsibility of ADHD management by pediatric specialists may pose access barriers, especially in the context of higher disease prevalence among the poorest children (25 26). The shift towards greater specialist care was not accompanied by evidence of an increase in illness severity in treated cases, as might be hypothesized with increased patient and provider awareness of the potential cardiovascular risks of pharmacotherapy.
Our analysis has several limitations and raises questions for future research. First, our focus was on the office-based diagnosis and pharmacologic treatment of children and adolescents, and thus we do not have data regarding the variety of non-pharmacologic treatments that may play an important role in the care of patients with ADHD. Future analyses are needed to examine the use of other interventions that some providers and parents may pursue for children with ADHD, such as diet, complementary/alternative medications or cognitive and behavioral therapies. Similarly, our analyses preclude an investigation of how the changes we describe were associated with changes in individuals’ educational environments such as individualized learning plans. Second, we used one audit of office-based encounters; future work should examine the trends that we describe with other office-based audits such as the National Ambulatory Medical Care Survey. Third, our study was not designed to specifically estimate the impact of regulatory and professional society advisories on ADHD treatment. Finally, we did not examine the prevalence and treatment of other disorders that may overlap with ADHD, such as bipolar affective disorder, conduct disorder, or oppositional defiant disorder, and our analysis does not allow for an assessment of the health impacts of the changes identified.
In the United States, visits related to the diagnosis and treatment of ADHD among children and adolescents increased substantially over the last decade. The magnitude and speed of these shifts is more likely to be due to changes in practice patterns than a change in the population prevalence of ADHD. Further work is needed to better characterize the precise role that various influences may have played and most importantly, to identify the effects of these changes on the health of children and their families.
We describe recent trends in the diagnosis and pharmacologic treatment of ADHD, including large increases in ADHD visits over the decade, changes in medication use, and shifts in care from pediatricians to psychiatrists.
The statements, findings, conclusions, views, and opinions contained and expressed in this article are based in part on data obtained under license from the following IMS Health Incorporated information service(s): National Disease and Therapeutic Index™ (2000-2010), IMS Health Incorporated. All Rights Reserved. The statements, findings, conclusions, views, and opinions contained and expressed herein are not necessarily those of IMS Health Incorporated or any of its affiliated or subsidiary entities.
This work was supported by the Agency for Healthcare Research and Quality (RO1 HS0189960). Dr. Dorsey was supported by the Robert Wood Johnson Physician Faculty Scholars Program and Dr. Huskamp was supported by a Robert Wood Johnson Foundation Investigator Award in Health Policy Research. Dr. Dusetzina was supported by a Ruth L. Kirschstein-National Service Research Award (T32 MH019733-17). The funding sources had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; and preparation, review, or approval of the manuscript for publication.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
All authors contributed to the research questions and analytic approach. SZ and ASH and RK derived the analytic extracts of data. CFG, ERD, JMP and GCA analyzed and interpreted the data. All authors reviewed these interpretations and provided substantive feedback. CFG drafted the manuscript. All authors reviewed the manuscript and edited it for intellectual content. All authors approved of the final manuscript. GCA supervised the study.