National Trends in ADHD Diagnosis and Treatment: Comparison of Youth and Adult Office-Based Visits

In this cross-sectional study, data from medical office–based visits by youths (ages two to 17 years) and adults (18–64 years) were analyzed from the National Ambulatory Medical Care Survey (NAMCS) and the National Hospital Ambulatory Medical Care Survey (NHAMCS) for calendar years 1999 through 2010. The NAMCS and NHAMCS are annual surveys of patient visits to physicians in office-based practices and hospital outpatient departments (13). A multistage probability design is used for the sampling of patient visits, which are selected from primary sampling units (PSUs), physician practices within the selected PSUs, and visits from the selected practices. Because the data set consisted of deidentified data, the study was exempt from review by the institutional review board.

There were two main study outcomes: first, the primary measure was visits noting an ADHD diagnosis, expressed as a percentage of total office-based visits to physicians. The diagnosis of ADHD was ascertained with an ICD-9-CM code of 314.xx. The second outcome was the prescribing of stimulants—including methylphenidate and amphetamine salts—during ADHD-related medical visits.

The trend in outpatient visits with a diagnosis of ADHD was assessed across three four-year periods (1999–2002, 2003–2006, and 2007–2010). Other study covariates included gender, race-ethnicity, U.S. census region, payment type, and physician specialty. Adopting the ICD-9-CM codes from a previous study (14), we categorized other psychiatric diagnoses into the following groups: schizophrenia or other psychoses, pervasive developmental disorders and intellectual disability, bipolar disorder, disruptive behavior disorders, depressive disorders, anxiety disorders, and “other disorders,” a group that included psychiatric diagnoses not reported in the other noted diagnostic groups.

All statistical analyses were performed at the visit level with SAS, version 9.2. The analyses accounted for the multistage probability design (unequal weighting, clustering, and stratification) so that nationally representative estimates were produced.

The proportion of visits with a diagnosis of ADHD was assessed with bivariate analyses across 12 years for total visits as well as for visits stratified according to physician specialty, payment type, gender, and race-ethnicity. These analyses were conducted separately for visits by youths and adults. In addition, multivariable logistic regression models were conducted to assess the odds of receiving an ADHD diagnosis in visits made during 2007–2010 compared with those made during 1999–2002, with adjustment for study covariates. The models included an interaction term between time period and age group to assess whether the trend in ADHD diagnosis in physician office visits differed between youths and adults.

Sociodemographic, administrative, and clinical characteristics of ADHD visits were compared by age group with the use of chi-square analyses. Finally, in visits with a diagnosis of ADHD, we assessed stimulant prescribing patterns with bivariate analyses and multivariable logistic regression.

As a proportion of total office-based visits with physicians, visits from 1999 through 2010 involving an ADHD diagnosis increased from 3.9% to 5.2% for youths and from .3% to .7% for adults (Table 1). The increase in the diagnosis of ADHD was significantly greater in visits by adults (adjusted odds ratio [AOR]=2.7) than by youths (AOR=1.3) (interaction p<.001), and this finding did not differ whether payment type was public or private or for visits by males or females.

Among visits to nonpsychiatrist physicians, the proportion of visits resulting in an ADHD diagnosis increased significantly more for adults (.1% to .4%) than for youths (2.5% to 3.7%) (interaction p=.001). Among visits to psychiatrists, the proportion of visits for adults resulting in an ADHD diagnosis doubled (5.0% to 11.1%). However, this increase was not statistically different compared with the increase in diagnosis among youths (from 43.2% to 53.2%).

In the period of 2003–2010 (Table 2), ADHD visits by adults compared with those by youths included a significantly greater proportion of visits by females (50.7% versus 26.3%), Caucasians (87.8% versus 72.6%), and the privately insured (67.7% versus 50.0%) and those seeing psychiatrists (59.3% versus 32.8%). Among visits with a diagnosis of ADHD, comorbid psychiatric disorders were reported more often among adults than youths (55.6% versus 31.9%). Depressive disorders (29.9% of adults versus 9.2% of youths), anxiety disorders (17.2% versus 7.0%), and bipolar disorders (8.8% versus 4.1%) were recorded more often in visits by adults than by youths. By contrast, pervasive developmental disorders and intellectual disability (3.0% versus 1.1%) and disruptive behavior disorders (8.9% versus 1.1%) were recorded more often in visits by youths than by adults.

In the 2003–2010 period (Table 3), stimulant medication was prescribed similarly in ADHD visits for youths (71.7%) and adults (74.7%), regardless of gender and physician specialty. Racial-ethnic differences in stimulant prescribing were less pronounced in visits by adults than by youths. In visits resulting in an ADHD diagnosis, stimulant prescribing was significantly greater in visits by privately insured adults than by publicly insured adults (78.1% versus 62.7%; AOR=1.8).

In an analysis of youth and adult outpatient visits, Olfson and colleagues (5) found that from 1995 through 2010, visits noting a psychiatric diagnosis increased significantly more among youths than adults, and more than half of these mental health visits were with providers who were not psychiatrists, regardless of age group. In our study, which specifically focused on ADHD and spanned 12 years (from 1999 through 2010), visits with a diagnosis of ADHD grew significantly more among adults than youths, particularly for visits provided by nonpsychiatrists. Of note, most (59.3%) visits by adults with a diagnosis of ADHD were with psychiatrists, whereas only one-third of ADHD visits by youths were with psychiatrists. This contrast highlights varying characteristics of treatment patterns by physician specialty.

The findings underscore the need for a better understanding of treatment-seeking behavior by adults as well as the referral process for the treatment of ADHD. A survey of primary care physicians and psychiatrists revealed that most adult patients with previously undiagnosed ADHD were self-referred for their clinical visits, regardless of provider type (15).

Although the growth in ADHD diagnosis occurred at a greater rate in visits by adults than by youths regardless of gender and payment type, this study found significant differences in characteristics between visits by youths and adults. Compared with ADHD visits by youths, visits by adults represented significantly greater proportions of females (50.7% versus 26.3%), Caucasians (87.8% versus 72.6%), and privately insured visits (67.7% versus 50.0%). Also, consistent with previous reports (7,16), comorbid diagnoses of bipolar disorder, anxiety, and depressive disorders were more frequently reported in visits by adults with a diagnosis of ADHD. The study data do not shed light on whether ADHD was clearly the clinical problem leading to these visits or whether patients were experiencing inattentiveness and poor organization due to other problems, including treatment-emergent effects (17). The distinct etiology of adult versus childhood ADHD may contribute to the varying characteristics observed in our study (18).

Asherson and colleagues (19) posited that underdiagnosis of adult ADHD may be related to challenges with diagnostic criteria, reliance on self-reporting, and cultural or media effects on the representation of ADHD. Also, the differences may be the result of underdiagnosis of ADHD among female youths, which may be recognized in adulthood through self-referral (16).

Stimulant prescribing was frequent in ADHD visits both for youths (71.7%) and for adults (74.7%). Notably, racial-ethnic differences in stimulant prescribing were less pronounced in adult ADHD visits than in those for youths. This finding might be due to the higher socioeconomic status that characterizes adult ADHD visits. Similarly, in visits by adults, stimulant prescribing was substantially greater for those with private insurance than for their publicly insured counterparts (78.1% versus 62.7%). Lower stimulant use by publicly insured adults may reflect greater reluctance by providers to prescribe abusable substances to these adults, but further research is warranted to elucidate the issue. Overall, the increasing trend in the diagnosis of ADHD along with the widespread prescribing of stimulants in outpatient treatment of adults raises concerns given that most stimulant misuse by adults involves prescription drugs (20). Future studies with adults may expand the focus to psychiatric conditions comorbid with ADHD in that complex drug regimens may result from multiple comorbidities (21).

In this study, diagnoses were clinician reported, which may not be as accurate as diagnoses assessed under clinical research conditions. In addition, the severity of ADHD could not be ascertained. Data on prescribing of stimulants may not have been commensurate with actual stimulant consumption. In our study, which focused on stimulant use alone, we did not characterize the prescribing patterns for nonstimulant drugs. However, prescribing of nonstimulant drugs, such as atomoxetine, in the outpatient treatment of ADHD was modest (<10% in youth and adult ADHD visits—data not shown). The findings may not generalize to practice patterns in specific regions of the United States, given the widespread geographic variation in stimulant treatment of adults and youths (22). Our data were limited to four large census regions and did not permit small-area variations to be assessed. Nevertheless, our data provide recent national estimates of the practice patterns on the diagnosis and stimulant treatment of ADHD in youth and adult outpatient visits, reflecting increasingly changed medical practice.