This Web site is a component of the SAMHSA Health Information Network

Enter Keywords Choose Website This Site Mental Health at HHS All Mental Health

Online Publications Order Publications National Library of Medicine National Academies Press Publications Homepage

printer friendly page e-mail this page bookmark this page shopping cart  current or new account

This Web site is a component of the SAMHSA Health Information Network.

Section 3: Status of Mental Health Services at the Millennium

Chapter 12. Pharmacoepidemiology of Methylphenidate and Other Medications for the Treatment of ADHD

Julie Magno Zito, Ph.D.*

University of Maryland School of Pharmacy

This paper aims to describe the clinical and social characteristics associated with the treatment of Attention Deficit Hyperactivity Disorder (ADHD) in the United States during the 1990's. The objectives are the following: to review the 10-year psychopharmacologic prevalence trends for the stimulants and related medications based on community treatment patterns; to discuss the epidemiology of these treatments in terms of host (person), agent (psychotropic medications), and environment (clinical, educational, and public issues) so that variations in prevalence can be best understood; to interpret the findings broadly in light of trends in the United States regarding clinical factors, nosology, educational policy, public attitudes, and media effects; and to suggest both future research to understand the appropriateness of the increase in psychotropic medication during the past decade, and changes in clinical practice guidelines.

Research Methods

The methods used to ascertain health service utilization in the United States are far from ideal because a national health insurance system is lacking and no comprehensive system of gathering national medical treatment data exists. As a result, only a few population-based databases have been used to estimate the medication usage patterns in the usual practice setting (Zito & Safer, 1997). Consequently, analysts and the lay press often rely on marketing data, such as IMS America, a proprietary prescription survey that tracks market share principally for the pharmaceutical industry. A second source, the National Ambulatory Medical Care Survey (NAMCS), involves a national probability sampling of physician office visits; but this resource is limited, in part, because child mental health services involving prescription medications represent a very small, unreliable sample of these data. A larger, more promising source is administrative claims data from various clinical practice settings such as the Medicaid health insurance system, which covers persons with low income, impairment (Supplemental Security Income), or special placement (e. g., foster care).

In studies by the author and her colleagues, two large population data sets were obtained at three sites. Two sets of Medicaid data were selected: the first was from a Mid-Western Medicaid (MWM) State and the second was from a Mid-Atlantic Medicaid (MAM) State. The third data set included employed, insured individuals and their families. Health records of these families were gathered from a nonprofit health maintenance organization (HMO) from the northwestern region of the United States. After organizing the enrollment data from these systems and the administrative claims for reimbursement of medical and prescription services (for Medicaid) or prescription records (for the HMO), we undertook a comprehensive analysis of psychotropic medication prevalence. The analysis focused on stimulants, the psychotropic drug class most commonly used among children, and methylphenidate, the most common medication within the stimulant class. Prevalence was defined as the number of individuals with one or more prescriptions for a specific medication or medication class during the study year per 1,000 individuals enrolled in the Medicaid or the HMO health service system.

The trends for population-based total stimulant prevalence are illustrated in Figure 1. Stimulant use among those less than 20 years old who were treated in the HMO setting showed a 606 percent increase in use (from 0.36 percent in 1987 to 2.54 percent in 1996) while the HMO enrollment rose only 17 percent. Figure 2 shows that stimulant use among 5-to 14-year-olds in the Medicaid setting (MAM and MWM) was nearly twice that in the HMO setting. One-year stimulant prevalence was eight percent among 5-to 14-year-olds in MAM but only four percent in the HMO. These differences could be accounted for both by geographic differences and by target population differences as well as by possible prescribing practice differences.

Host (Sociodemographic) Factors That Influence Stimulant Prevalence

Host factors that influence drug prevalence include age, gender, ethnicity, geographical locale, and socioeconomic status. An analysis of these factors produced the following results:

(3) As illustrated in Figure 3, age-specific prevalence differed substantially among the four age groups (0–4; 5–9; 10–14; and 15–19 years old). For example, in MWM, the lower (0–4) and upper (15–19) age groups had a very low stimulant prevalence relative to the treatment of ADHD at the typical ages of 5–14 years (7 and 12 per 1,000 for 0–4 and 15–19-year-olds, respectively). When the age rates are compared with the stimuant prevalence of 70 and 68 per 1,000 for the 5–9 and 10– 14-year-olds, the difference was sixfold to tenfold. Between 1987 and 1996, there was a threefold and fivefold rate increase, respectively, among 5–9 and 10–14-year-olds. However, the increase in use during this decade was most dramatic (six-fold) for the 15– 19-year-olds, suggesting a longer duration of treatment with stimulants than in the previous decade. When the preschoolers in MWM were examined by year of age, 3-and 4-year-olds had stimulant rates of 1 and 2 percent by 1996, up approximately threefold over the previous decade (Figure 4).

(4) Gender-specific prevalence data typically illustrate the predominance of ADHD treatment among boys. However, when male-to-female preschooler ratios were compared across the decade, there was a marked change in the gender disparity. This is well illustrated among HMO youths. By 1995, the male: female ratio was half of the 1991 ratio (4.6:1 versus 9.8:1). This change suggests that girls entered treatment in increasing numbers during the 1990's.

(5) A racial disparity in stimulant use is observable from the race-specific data and this effect is influenced by age. Stimulant prevalence for Caucasian youths was approximately twice that of non-Caucasians in the MAM data source. The disparity is greatest for the oldest age group— a fact that may be consistent with differential school dropout rates among high schoolers or variable time in treatment.

(6) Geographical locale-specific prevalence for stimulants in 1996 showed a 5.1-fold variation across eight regions of the MAM system (Figure 5). Further analysis of race-specific and geographical locale-specific prevalence is instructive. It is important to consider the interaction of race and region, since each factor may independently influence health service utilization. To accomplish this analysis, a logistic regression model was developed with race and region as predictors of the odds of receiving methylphenidate. Caucasians were 2.6-fold more likely to receive this treatment compared with non-Caucasians, a ratio that dropped to 2.2-fold when region was accounted for in the model. The interaction of race and region was significant (p < 0.001), which suggests that race-specific prevalence varies according to the geographical locale. These 1996 data corroborate our earlier findings regarding Caucasian and African-American youths ages 5–14 years old who participate in Medicaid (Zito, dosReis, Safer, & Riddle, 1998).

When a comparison of Caucasian to African-American prevalence ratios was made for the leading psychopharmacologic classes of medication (stimulants, antidepressants, antipsychotics, and lithium) in relation to the leading nonpsychopharmacologic medication classes (e.g., antibiotics, topical agents, antitussives, and eye/ear/nose and throat remedies), the racial disparity was 58–79 percent greater for psychopharmacologic agents used to treat mental or behavioral disorders than for medications used to treat medical disorders. This fact suggests that cultural differences explain the lower psychopharmacologic use relative to medical drug use.

(7) Socioeconomic factors explain several differences in the attitudes, satisfaction, and knowledge of the medication experience reported in a survey of parents with children receiving methylphenidate. Survey responses from parents in the low socioeconomic class category who were receiving service in a State-supported mental health clinic were compared with higher income parents who were participating in an advocacy and support group for ADHD. Pronounced differences were noted: school referrals were 2.5-fold more frequent for the low-income group; school-day-only treatment regimens were more likely in the low-income group; and counseling was less likely in low-or middle-income groups. Better knowledge scores and fewer fears about medication but less satisfaction with social functioning were reported by the high-socio-economic-class parent group (dosReis, Zito, Safer, & Soeken, 1998).

Medication and Medication-Related Factors That Influence Stimulant Prevalence

Medication and medication-related factors influence drug prevalence. Among these factors are (1) marketing and promotion; (2) physician prescribing patterns within the class of stimulants; (3) the growing use of stimulants along with ancillary medications, most of which are off-label (without indications in the Food and Drug Administration (FDA)-approved labeling information for the product package insert); and (4) Federal and local advocacy issues influencing stimulant treatment in the United States.

Results of the medication analyses suggest the following inferences. First, promotion of a combination of four amphetamine salts (Adderall®) was very successful during the late 1990's and the effect is evident in the increase in sales according to recent National Prescription Audit data. From January 1996 to March 1999, Adderall sales increased more than fortyfold and, in March 1999, exceeded prescription sales of brand-name methylphenidate (Ritalin®) by 1.5-fold. Second, changes in the proportional data within the stimulant class suggest that other amphetamines and Adderall are enjoying increased use while methylphenidate and pemoline have slightly reduced proportions (Figure 6). Recent clinical reports of serious liver toxicity associated with pemoline use (Rosh, Dellert, Narkewicz, Birnbaum, & Whittington, 1998) were largely ignored until 1999. This fact reminds us of the length of time it takes to change clinical practice when we rely on voluntary reporting of adverse medication events. Third, trends in MWM between 1987 and 1996 for selected psychopharmacologic agents show that alphaagonists (clonidine and guanfacine) increased 53-fold while antidepressants increased 3.6 times and stimulants increased 3.7 times (Figure 7). Thus, considerable increased psychotropic medication use is observed and is likely to be explained by more youths in treatment, longer times in treatment, and the concurrent use of stimulants and ancillary medications (e. g., an alpha-agonist for insomnia related to ADHD or to stimulant use, or an antidepressant for comorbid depression).

Clinical and Environmental Factors That Influence Stimulant Prevalence

Clinical and environmental factors that influence stimulant prevalence include (1) nosological changes; (2) comorbidities and multiple medication practices; and (3) health service system changes. First, we note changes in the clinical symptoms to meet diagnostic criteria according to the latest version of the Diagnostic and Statistical Manual (DSM-IV; APA, 1994) relative to earlier versions and to the International Classification of Diseases. The 1994 DSM criteria make it easier for youth to meet criteria based on inattention alone (Wolraich, Hannah, Pinnock, Baumgaertel, & Brown, 1996). A second factor involves the increasing identification of comorbidities among those with ADHD. This trend partially explains the use of multiple medications, particularly antidepressants for comorbid depression. A review of the diagnoses related to stimulant use among youths in the MAM and MWM systems suggests that only 67 percent and 74 percent, respectively, of the stimulant-treatedindividuals had a diagnosis of ADHD (Figure 8). Nearly 20 percent of the stimulant-treated youths had no diagnosis during the study year (which may be an artifact of the cross-sectional research design) and a substantial proportion had psychiatric diagnoses other than ADHD, a finding that suggests that symptomatic treatment with stimulants is expanding among those with related psychiatric disorders. This conclusion is also supported by the MWM data comparing 1987 or 1991 with 1995 for individuals with ADHD alone and those with additional (comorbid) diagnoses (Figure 9). The disparity between ADHD and ADHD with comorbidities was greater in 1987 than in 1995. Prominent among the comorbidities were disruptive disorders such as conduct disorder and oppositional defiant disorder, which grew 27 percent, and depression, which had a 270 percent increase in prevalence during that 9-year period (Figure 10). When those with an ADHD diagnosis alone were reviewed, the increased use of other medications from 1987 through 1995 was pronounced. Examples include a 35-fold increase in the use of alpha-agonists (clonidine or guanfacine) and a 2.4-fold increase in the use of antidepressants. Methylphenidate treatment alone proportionately decreased by 9. 2 percent over the same period. Finally, when youths in MWM receiving multiple medications in 1987 were compared with those in 1995, those with two or more increased from 16 percent to 27 percent, while those receiving only one medication class decreased proportionately.

A third clinical area that explains variations in the prevalence for ADHD medication treatments concerns the treatment setting as defined by the health service system. From our 8-year analysis of NAMCS data, we reported the following findings: primary care providers (pediatrics, general practice, family practice, and internal medicine specialists) differ from psychiatrists when ADHD visits are compared to non-ADHD visits. Primary care provided 61 percent while psychiatry provided 25 percent of ADHD visits. Second, HMO insurance coverage, publicly insured (Medicaid), and privately insured (e. g., preferred provider organization insured) were significantly different with respect to ADHD and total other visits. HMO had only 11.7 percent of ADHD visits, although its share of non-ADHD visits was 17. 9 percent. Private insurance had 51 percent and public insurance had 23. 7 percent of ADHD visits. ADHD visits increased across the 8-year span, doubling in the latter half of the interval. Stimulant treatment as a proportion of ADHD visits increased from a mean of 62. 6 percent in 1989 to a mean of 76.6 percent in 1996 (Zito et al., 1999).

Educational policy changes in 1990 expanded the identification of ADHD and led to an increased role for schools in assessing the emotional health needs of students. As a result, school staffs became more accommodating and responded more to parental demands for psychological and educational testing of restless and inattentive youths for special education services, which increased their role in assessing ADHD. The role of the media has moved from largely negative reporting in the 1960's through the late 1980's to a more balanced, if not more positive, viewpoint. U.S. Federal mental health programs promoting the "decade of the brain" in the 1990's is a related development that may explain the greater acceptance of somatic treatments for ADHD by both teachers and families. Baltimore County, Maryland, public school survey data illustrate the increased duration of treatment among school-age youths from 1971 through 1997. No children from middle or high school were medicated during the school day according to the 1971 data, but 5.6 percent of middle schoolers and 1.6 percent of high schoolers were reported to be receiving medication for ADHD during the school day in April 1997 (Safer & Zito, 2000). The negative media effect is shown in the dip in stimulant prevalence during 1988 and 1989, years when newspapers in Baltimore carried details of a lawsuit against the county school system. The result was a 39 percent drop in stimulant treatment for public school students from 1987 to 1991 (Safer & Krager, 1992).

Conclusions from our findings to date are as follows: (1) there is a substantial difference in stimulant prevalence in public versus private health service systems, and (2) medication utilization for ADHD has increased substantially over the past decade. The stimulant prevalence increase is attributable to (1) expanded diagnostic criteria; (2) longer time in treatment, resulting in more teenage youths in treatment; (3) more girls in treatment; (4) a threefold increase in stimulant prevalence among 2–4 year olds (Zito et al., 2000); (5) an increased role of schools; and (6) more favorable attitudes of families and professionals. Concerns are raised in regard to (1) the appropriateness of medicating pre-schoolers for ADHD; (2) the long-term effectiveness of stimulants for the treatment of inattention as the sole symptom of ADHD; (3) long-term safety issues; (4) the efficacy and safety of off-label medications particularly when used in combinations for the treatment of ADHD; (5) racial and socioeconomic disparities; and (6) the role of cultural differences in the acceptability of ADHD and its treatment with medication.

Implications for Clinical Research and Clinical Practice

The appropriateness and the outcome of treatment in the usual practice setting need to be more intensively researched. Measures should include symptom improvement and consumer satisfaction as well as functional assessments in the crucial areas of academic performance, behavior, and social relations.

The results of this investigation clearly call for considerable additional research to help us understand the nature and extent of ADHD and its appropriate treatment in children and adolescents. Some key questions include: What is the prevalence of ADHD? Is it increasing or decreasing? In which age groups? Has accurate case finding improved over time with better diagnostic criteria and improved knowledge? Are children with ADHD getting appropriate treatment according to current knowledge about quality care? What is the appropriate mix of psychopharmacology and psychotherapy in the treatment of ADHD? Which children with ADHD are not getting any care at all? What is the role of the family, the school, and the community in the delivery of quality care? How can we implement new knowledge about improved assessment and treatment of ADHD? Each of these questions is of very high priority. The importance of our children to our future as a society demands no less.

References

Coyle, J. T. (2000). Psychotropic drug use in very young children. JAMA, 283(8), 1059–1060.

dosReis, S., Zito, J., Safer, D., & Soeken, K. (1998). Parental knowledge, attitudes and satisfaction with medication for children with Attention Deficit Disorder [Abstract]. Association for Health Services Research (1998 Abstract Book), 315–316.

Education of the Handicapped Act Amendments of 1990, PL 101–476. (1990).

Rosh, J. R., Dellert, S. F., Narkewicz, M., Birnbaum, A., & Whittington, G. (1998). Four cases of severe hepatoxicity associated with pemoline: Possible autoimmune pathogenesis. Pediatrics, 101(5), 921–923.

Safer, D. J., & Krager, J. M. (1992). Effect of a media blitz and a threatened lawsuit on stimulant treatment. JAMA, 268, 1004–1007.

Safer, D. J., & Zito, J. M. (2000). Pharmacoepidemiology of methylphenidate and other stimulants for the treatment of ADHD. In L. L. Greenhill & B. B. Osman (Eds.), Ritalin: Theory and practice (pp. 7–26). Larchmont, NY: MA Liebert.

Wolraich, M. L., Hannah, J. N., Pinnock, T. Y., Baumgaertel, A., & Brown, J. S. (1996). Comparison of diagnostic criteria for Attention-Deficit Hyperactivity Disorder in a county-wide sample. Journal of the American Academy of Child & Adolescent Psychiatry, 35(3), 319–324.

Zito, J. M., dosReis, S., Safer, D. J., & Riddle, M. A. (1998). Racial disparity in psychotropic medications prescribed for youths with Medicaid insurance in Maryland. Journal of the American Academy of Child & Adolescent Psychiatry, 37(2), 179–184.

Zito, J. M., & Safer, D. J. (1997). Sources of data for pharmacoepidemiological studies of child and adolescent psychiatric disorders. Journal of Child and Adolescent Psychopharmacology, 7(4), 237–253.

Zito, J. M., Safer, D. J., dosReis, S., Gardner, J. F., Boles, M., & Lynch, F. (2000). Trends in the prescribing of psychotropic medications to preschoolers. JAMA, 283(8), 1–6.

Zito, J. M., Safer, D. J., dosReis, S., Magder, L. S., Gardner, J., & Zarin, D. A. (1999). Psychotherapeutic medication patterns for youths with attention deficit hyperactivity disorder. Archives of Pediatrics & Adolescent Medicine, 153, 1257–1263.

* Acknowledgments: Co-investigators on these studies include Drs. Daniel Safer and Susan dosReis, James Gardner and Indrani Basu. The work was supported by R01 MH 55259 from the National Institute of Mental Health and the G. Harold and Leila Y. Mathers Charitable Foundation.

Index | Previous | Next