Evidence of Overdiagnosis and Overuse of Medication
Gretchen B. LeFever and Andrea P. Arcona - Center for Pediatric Research, Eastern Virginia Medical School and Children's Hospital of the King's Daughters
David O. Antonuccio - University of Nevada School of Medicine, Veterans Affairs Sierra Nevada Health Care System.
Correspondence concerning this article should be addressed to Gretchen B. LeFever, Ph.D., Old Dominion University and Safety and Learning Solutions, 912 Queen Elizabeth Drive, Virginia Beach, VA 23452, firstname.lastname@example.org.
The 700% increase in psychostimulant use that occurred in the 1990s justifies concern about potential overdiagnosis and inappropriate treatment of child behavior problems. A critical review of epidemiologic research suggests that attention-deficit/hyperactivity disorder (ADHD) is not universally overdiagnosed; however, for some U.S. communities there is evidence of substantial ADHD overdiagnosis, adverse educational outcomes among children treated for the disorder, and suboptimal management of childhood behavior problems. Evidence of ADHD overdiagnosis is obscured when findings are reported without respect to geographic location, race, gender, and age. More sophisticated epidemiologic tracking of ADHD treatment trends and examination of associated outcomes is needed to appreciate the scope of the problem on a national level. Meanwhile, a public health approach to ADHD that includes the development and implementation of data-driven, community-based interventions is warranted and is underway in some communities. Guidelines for promoting judicious use of psychotropic drugs are suggested.
Until the latter half of the 20th century, treating childhood behavior problems with medication was an almost nonexistent practice. The current American proclivity toward psychiatric drug therapy for behavior-disordered children began in the 1960s, when the American medical profession deemed it acceptable to use psychostimulants (especially methylphenidate, commonly and hereafter referred to as Ritalin) to ameliorate symptoms associated with minimal brain dysfunction (MBD; Wender et al., 1971), or what is now described as attention-deficit/hyperactivity disorder (ADHD; American Psychiatric Association, 1994; Barkley, 1990). Over the last three decades the rate of drug treatment for behavior problems has increased exponentially, culminating in the prescription of ADHD drug treatment for at least 5 to 6 million American children annually (Diller, 1998; Sinha, 2001). The high rate of prescription for Ritalin and expensive brand-name drugs such as Adderall, Concerta, and Metadate reflect a more general reliance on psychotropic drugs in American healthcare practices. In 1998, doctors mentioned psychotropic drug treatment an estimated 85.8 million times during 36.7 million office visits (Health Care Financing Administration, 2001) averaging 2.3 documented references to psychotropic drug treatment per physician visit. The unprecedented levels of drug treatment for child behavior problems justify closer public and professional scrutiny. Available research has not supported the idea of a widespread overdiagnosis of ADHD across the country (Goldman, Genel, Bezman, & Slanetz, 1998); however, there are clear indications of overdiagnosis and overtreatment in a growing number of communities. The fact that these problems are not universal should not serve to dismiss concerns for communities in which children are being diagnosed with and treated for ADHD at remarkably high rates. Careful investigation into the extent of overdiagnosis of ADHD and overuse of drug therapy is necessary to develop appropriate methods for improving ADHD care. This paper provides background information regarding ADHD treatment and a summary of historical trends in ADHD treatment, which set the stage for a critical review of epidemiological research on ADHD and its treatment.
Defining ADHD Overtreatment
Defining mental health disorders based on the concept of statistical rarity is arguably problematic for many mental health conditions (e.g., Wakefield, 1992), but not for all. In fact, definitions of some disorders-including ADHD-are reliant on the concept of statistical rarity, or what is sometimes referred to as developmental deviance. Consider the case of mental retardation vis-à-vis intelligence. Mental retardation (the condition) is defined by intelligence (the construct) that is measured to be at least two standard deviations below the population mean. While some individuals may have low intelligence, only those whose intelligence is significantly developmentally deviant (i.e., statistically rare) are considered disordered. The diagnosis of ADHD is conceptually akin to that of mental retardation in that the definition of both disorders relies on the concept of developmental deviance. As with intelligence, the hallmark symptoms of ADHD (impulsivity, hyperactivity, and inattention) exist in all children to some degree, but ADHD is said to exist only when the behaviors are expressed to an extreme or statistically rare degree.
Given that the definition of ADHD is based on statistical rarity, only a limited number of children can qualify as having the disorder. As in the case of mental retardation, the ADHD prevalence estimate was set at 3% to 5%, which restricts the disorder to those children whose ADHD-related behavioral characteristics are approximately two standard deviations away from the mean. The 3% to 5% estimate may constitute a liberal estimate because, as with mental retardation, statistical rarity is only one of several criteria for the diagnosis. The problematic behavior must also be persistent, pervasive, impairing, and not attributable to other conditions or factors. Consistent with this logic, some pediatric and behavioral experts argue that ADHD may affect as few as 1% to 3% of children (Carey, 1999, 2000). This notion received considerable attention by scientific experts on the NIH-sponsored ADHD Consensus Development panel (Diagnosis and Treatment of ADHD, 1998), although the consensus was to maintain the decades-long prevalence estimate of 3% to 5%.
Some may argue that the current definition of ADHD is flawed and that the concept of developmental deviance or statistical rarity should be discarded. Because ADHD is presumed to be a biological disorder, there may be no natural limit to the number of children who could be affected by the disorder and the 3% to 5% prevalence would be not only arbitrary but also inappropriate. There is no pathognomonic biological marker for ADHD (Barkley, 1999; Todd, 2000) and no clearly defined and widely accepted ADHD assessment method (Kessler, 1980), making it impossible to know precisely how many children are actually affected by the disorder (Godow, 1997). Unless a biological marker is identified, an agreed-upon gold standard diagnostic procedure is established, or ADHD is redefined, a population-based ADHD rate exceeding 3% to 5% by definition represents a problem of ADHD overdiagnosis. Thus, the 3% to 5% prevalence estimate is presently the only benchmark that can be used to evaluate possible ADHD overdiagnosis and overtreatment. Throughout this paper, ADHD- related drug treatment trends and community-based assessments of ADHD diagnostic rates are evaluated against the 3% to 5% benchmark.
Review of National ADHD Drug Treatment Trends
For years, discussions about the overdiagnosis of ADHD and overprescription of Ritalin have been one in the same. Until the late 1990s the vast majority of children medicated for ADHD received a psychostimulant and in 90% of these cases Ritalin was prescribed (Hamilton, 2000; Wilens & Biederman, 1992). Most Ritalin prescriptions are associated with ADHD treatment among U.S. patients. Therefore, at least until the late 1990s, Ritalin consumption has been used to track general patterns of ADHD treatment in the United States (LeFever, Dawson, & Morrow, 1999). Prescribing trends prompted the United Nations International Narcotics Control Board to issue its second warning in recent years that American physicians may be overprescribing psychostimulants (United Nations Information Service, 1997). This may become an issue in other countries such as Australia (Carmichael, 1996) and Canada (Miller, Lalonde, McGrail, & Armstrong, 2001), where ADHD drug treatment is becoming increasingly popular.
The Early Years of Ritalin
Use As of 1960 negligible numbers of children were medicated for MBD/ADHD (Safer & Zito, 1999). A decade later, more than 150,000 school-age children were receiving psychostimulant treatment annually in the United States (Safer, 1971). Beginning with the conservative estimate that 50,000 American children were treated annually for ADHD in 1960, there was a six-fold increase in psychostimulant treatment between 1960 and 1975. The rising rate of stimulant treatment prompted one of the nation's foremost developmental psychologists to address the topic in the New England Journal of Medicine (Sroufe & Stewart, 1975). Psychologist Alan Sroufe and psychiatrist Mark Stewart cautioned against the dangers of continued escalation in psychostimulant treatment, including possible reduction in parent and teacher motivation to take other steps to help children, inadequate monitoring of drug treatment, and possible development of low self-esteem and drug abuse among individuals treated with stimulants. Although these issues have never been adequately resolved, stimulant treatment has continued to increase unabated. The trends Sroufe and Stewart observed during the 1970s pale in comparison with those of the last two decades.
The Post-1990 Era of Ritalin Use
During the 1990s there was a 700% increase in the use of psychostimulants, with the United States consuming nearly 90% of the world's supply of the drugs (Mackey & Kipras, 2001; Marshall, 2000). As of 1999, school nurses across the country delivered more medications for mental health conditions than for any other chronic health problem, and more than half of these were specifically for ADHD (McCarthy, Kelly, & Reed, 2000). As of today, up to an estimated 5 to 6 million American children receive ADHD-related drug treatment annually (Sinha, 2001). These figures indicate that from 1960 to the turn of the century there was a more than 100-fold increase in the annual rate of ADHD drug treatment among U.S. children. Moreover, the use of psychostimulants and other psychotropic drugs continues to rise.
Accumulating evidence indicates that Ritalin use is highly variable across the country, with widening variation in state-level and community-level ADHD drug treatment over time. Drug Enforcement Agency (DEA) data for the years 1990 to 1995 indicated that the rate of Ritalin use was 6 times higher in some states compared with others (Morrow, Morrow, & Haislip, 1998) and was 20 times higher in some communities compared with others (Eaton & Marchak, 2001; LeFever, Arcona, & Stewart, 2001). In the years 1997 to 1999, some states used 30 times more Ritalin than other states, and some communities used 100 times more than others. Although the validity of DEA data for capturing treatment patterns in not entirely clear, the data represent one of the few sources of information currently available for tracking treatment patterns nationwide. Monitoring such data is also important because if treatment varies substantially across geographic regions, then the "average" practice may not be a very helpful index of the legitimacy of concern (Angold, Erkanli, Egger, & Costell, 2000).
Other Psychotropic Drugs
National statistics on the use of psychotropic drugs suggest that the current ADHD debate is no longer just about such psychostimulants as Ritalin (Zito & Safer, 2001). Between 1995 and 1998, antidepressant use increased 74% among children under 18, 151% among children between 7 and 12, and 580% among children younger than 6 years of age. Mood stabilizers increased 400% among children under 18, while the use of new antipsychotic medications increased 300% among the same age group (Diller, 2000). Recent analyses of the Kansas Medicaid database indicate that antidepressants were prescribed to children twice as often as any other type of psychotropic drug from 1995 to 1996 (Fox, Foster, & Zito, 2000).
These changes are also relevant to ADHD treatment practices because general practitioners and psychiatrists increasingly prescribe a variety of psychotropic drugs and drug combinations to ADHD children (Boles, Lynch, & DeBar, 2001; Popper, 2000; Zarin, Suarez, Pincus, Kupersanin, & Zito, 1998). Perhaps the most common of these regimens is the combined use of psychostimulants and antidepressants (Findling & Dogin, 1998; Zito et al., 2002). This pattern was observed in over one quarter of the children treated for ADHD in southeastern Virginia (LeFever, 2000). These increases have occurred despite the lack of convincing efficacy data for antidepressants in children (Antonuccio, Danton, DeNelsky, Greenberg, & Gordon, 1999). Still fewer data are available regarding the use of other psychotropics and psychotropic drug combinations in children. In one study, almost one fifth of children receiving prescriptions for ADHD from psychiatrists received drugs other than psychostimulants (Zarin et al., 1998). Although ADHD drug treatment estimates are often based on psychostimulant data, failure to consider a broader set of drugs substantially underestimates the magnitude of ADHD drug treatment in the United States.
The Case against ADHD Overdiagnosis
Evidence Potentially Supportive of Current Practices
Pooling findings from years of data from Baltimore County schools and from state Medicaid claims databases, Safer and his colleagues sought to address recent concerns about Ritalin use. Their data sources indicated that the prevalence of ADHD drug treatment doubled about every four to seven years between 1971 and 1987 and more than doubled between 1990 and 1995; however, the rates of increase have been variable across age categories (Safer, Zito, & Fine, 1996). Despite these trends, Safer et al. considered ADHD treatment rates through the mid-1990s to be unremarkable. However, they cautioned that their data were insufficient to settle the matter of ADHD overtreatment and that if treatment rates continued to escalate, concerns about overtreatment would be justified.
In most of his work, Safer considered the rate of school-based administration of Ritalin as synonymous with the prevalence of ADHD. As such, Safer and colleagues established the following estimates of ADHD prevalence among American school children: 1% for the 1970s, 3% to 5% during the 1980s, approximately 5% among students during the early to mid-1990s, and 4% in the late 1990s (Safer, 2000; Safer & Krager, 1988, 1989, 1994; Safer & Malever, 2000; Safer et al., 1996). They suggested that the discrepancy between escalating national estimates of Ritalin use and relatively stable rates of ADHD could be reconciled by considering several changes in clinical practice. Specifically, they mentioned a growing use of the medication among girls and adolescents, a growing tendency to keep diagnosed patients on the medication for longer periods of time, and more lenient ADHD diagnostic criteria. Safer also attributed the rise in Ritalin use to a broader acceptance of psychotropic treatment by the American public (Safer et al., 1996).
Jensen and his colleagues conducted a study that has been widely accepted as evidence against ADHD overdiagnosis and overtreatment (Jensen et al., 1999). Using post hoc analyses of data from a larger household survey, Jensen et al. examined the rate of ADHD identification and treatment in three U.S. cities and San Juan, Puerto Rico, in the first half of 1992. They identified potential ADHD cases among children and adolescents (9 to 17 years of age) by having lay providers conduct structured interviews of parent-child dyads. If either the child or the parent reported significant ADHD symptoms, the child was classified as having probable ADHD. The rate of Ritalin use was then examined among children who were classified as either ADHD or non-ADHD. Accordingly, 5% of children were identified with probable ADHD and only 12% of these children were taking Ritalin at the time of the survey. Because the rate of Ritalin use was substantially lower than the rate of ADHD, they reported and commented publicly that ADHD underdiagnosis and undertreatment was a major issue for American children.
Inadequacy of the Evidence
Addressing concerns about ADHD overdiagnosis requires consideration of at least the following six factors: (1) source of data, (2) sample size, (3) reporting of data by known risk factors, (4) timeliness of data, (5) accuracy of data, and (6) interpretation of data. When these factors are considered, the studies previously described cannot be used to dismiss concerns about ADHD overdiagnosis and overtreatment among American children.
Much of Safer's work has relied exclusively on school records or Medicaid samples. It is increasingly difficult to capture the extent of ADHD drug treatment by relying on school records. Safer estimated that school records missed at least 20% of children who were medicated for ADHD, because they did not necessarily receive a dose of their medication in school (Safer & Krager, 1994). With the growing popularity of long-acting drug treatments, the number of children medicated exclusively at home has increased. It now appears that school records alone may underestimate ADHD drug treatment by as much as 50% by the late 1990s (LeFever, Villers, Morrow, & Vaughn, 2002) and by 75% by 2002 (LeFever 2002). When reviewing post-1995 studies of ADHD, accurate estimates of drug treatment may require a doubling of figures obtained from school records. Medicaid databases also underestimate ADHD-related drug treatment because African American children, who are overrepresented in these databases, are half as likely as nonminority populations to receive ADHD drug treatment (Fox et al., 2000; LeFever et al., 1999).
The Jensen et al. (1999) epidemiologic study was more representative in that it involved community-based samples drawn from three U.S. cities and San Juan, Puerto Rico. However, it included only 1,285 children and 66 ADHD cases-far too few children and cases to generate a representative national picture of ADHD diagnosis and treatment. It is also puzzling that only 5% of children were identified as probable ADHD cases, because similar community-based studies estimate probable ADHD cases to be as high as 16% to 26% (Bird et al., 1988; Costello, 1989; Esser, Schmidt, & Woerner, 1990; Offord et al., 1987; Wolraich, Hannah, Pinnock, Baumgaertel, & Brown, 1996). This unexplained discrepancy complicates the interpretation of the Jensen et al. findings.
Reporting Data by Known Risk Factors
Neither Safer nor Jensen reported ADHD rates by race and gender. Such reporting is important (in addition to reporting by age groups) because prevalence and treatment have been documented to vary by all three factors, with the highest rate of ADHD among 6- to 9-year-old white boys (LeFever et al., 1999; Safer & Zito, 1999; Safer et al., 1996). Safer (1999) has presented, although not published, data indicating that from the early to mid-1990s the rate of ADHD treatment (i.e., school-administered Ritalin) among white boys in Baltimore County elementary schools was over 15%. Failure to report more specific findings when they are available can distort patterns in ADHD care. The Jensen et al. (1999) sample involved individuals between 9 and 17 years of age. By excluding a large proportion of children in the age group (i.e., ages 6 to 9) most affected by and/or treated for the disorder (Cohen et al., 1993; Safer & Zito, 1999; Safer et al., 1996; Scahill & Schwab-Stone, 2000), the study may have yielded an artificially low rate of ADHD and its treatment.
Timeliness of Data
Data from Jensen et al. (1999) were collected during the early part of 1992. Given the continuous increase in ADHD drug treatment throughout the 1990s, those data may be outdated. As the authors noted, very different results might have been obtained if the study were repeated today. [c head] Accuracy of Data The decrement that Safer noted in Ritalin use from the mid- to late 1990s in his most recent study (Safer & Malever, 2000) is curious. Although broad in terms of the geographic study region, the most recent study was not as thorough or methodologically rigorous as his previous studies. The data were collected at the school level, rather than the student level, and pertained only to Ritalin rather than a broader array of stimulants and nonstimulant ADHD-related drug treatments. School nurses reported the summary data under a legislated mandate and no quality control checks were reported. This means that study results were based on a limited number of data points, captured only a subset of ADHD drug treatment, and may be of questionable accuracy. Thus, the estimated 4% ADHD prevalence rate (i.e., school-based stimulant treatment rate) may poorly represent the actual data and is most likely a significant underestimate of ADHD diagnosed and treated prevalence.
Interpretation of Data
The magnitude of geographic variation in Ritalin use and ADHD diagnosis is so substantial that no study has been conducted that by itself is sufficiently methodologically rigorous to dismiss concerns about ADHD overdiagnosis. The data from Jensen et al. (1999) and the series of studies by Safer and colleagues remain limited in their refutation of ADHD overdiagnosis trends. As outlined later, a better understanding of the issue is obtained by consideration of ADHD prevalence findings from regions of the country with low, moderate, and high rates of ADHD drug use.
Evidence of ADHD Overdiagnosis
In response to high rates of clinic referrals for ADHD, LeFever and colleagues (1999) examined the rate of medication administered (daily) during the 1995-96 school year to students in two school districts in southeastern Virginia (LeFever et al., 1999). ADHD was defined by medication administered in school during regular hours and physician diagnosis. To guard against inflation of estimated drug treatment rates, children enrolled in self-contained special education classes, a group known to have a very high rate of treatment, were excluded. Using this conservative method of assessing ADHD treatment among nearly 30,000 students in grades two through five, 8% to 10% of the students were treated with stimulants for ADHD. Such population-based studies of actual rates of diagnosis and treatment are essential for addressing the current controversy, yet this study has been excluded from major reviews on the topic (Brown et al., 2001; Phelps, Brown, & Power, 2001) because diagnostic criteria were not reported. National research has documented that primary-care physicians rarely adhere to standardized ADHD diagnostic criteria (Kellerher & Larson, 1998) and regardless of the diagnostic criteria, used by practicing physicians in southeastern Virginia, the study documented excessive rates of ADHD drug treatment across the large and highly populated geographic region.
Findings from this study were consistent between two racially, economically, and socially diverse school districts. In both school districts, the rate of ADHD medication use was highest among white males and lowest among black females; 17% of white males and 3% of black females received ADHD medication in school. As noted above, these findings are consistent with research in public schools in Maryland (Safer, 1999). They indicate that in southeastern Virginia and Baltimore County, Maryland, the rate of school- administered ADHD drug treatment among elementary students was at least 2 to 3 times higher than the national estimate of the disorder.
To address the underestimation of ADHD diagnosis and treatment that is inherent in exclusive reliance on school-based records, LeFever and colleagues examined the rate of ADHD diagnosis and treatment among children by means of parental report (LeFever et al., 2002). In the same Virginia region previously studied, parents of children enrolled in three elementary schools that comprise a representative sample of the entire school district's elementary population were asked to complete a school health survey that focused on ADHD and asthma. Parental report of school-based administration of ADHD medication (9%) was consistent with objective assessment of the same (8% to 10%). However, nearly twice as many students had been diagnosed with ADHD compared with the number medicated in school. Among elementary students, 17% of all students and 33% of white boys had been diagnosed with ADHD and the vast majority had been medicated for this condition at some time during the 1997-98 school year. At the time of the survey, which spanned the summer months (when drug therapy is sometimes temporarily discontinued), 12% of all elementary students were medicated for ADHD (LeFever et al., 2002). Ninety percent of all identified cases had been medicated for the disorder at some point and the majority had been treated for over two years. These findings suggest that as of 1998, school-based studies of ADHD prevalence captured approximately half of the cases treated in the community. Preliminary data from a follow-up study suggest that as of 2002, school records capture as few as 25% of ADHD cases (LeFever, 2002).
Angold and colleagues (2000) conducted an epidemiologic study of mental health status among children in 11 western North Carolina counties from 1992 to 1996 (Angold et al., 2000). The study included children and youth between 9 and 16 years of age and examined participants periodically for four years. Based on parent report, almost three quarters of children with an unequivocal ADHD diagnosis received medication for the disorder. The majority of individuals who received stimulants were never reported by their parents to have any impairing ADHD symptoms, although they had some degree of symptomatology per parent and teacher report. In other words, even when symptoms were reported, they fell well below the threshold for an ADHD diagnosis.
At some point during the Angold et al. (2000) study, 3.4% of children met unequivocal diagnostic criteria for ADHD and 7.3% of children were treated with psychostimulants. This rate of treatment is almost twice as high as would be expected based on the estimated base rate of 3% to 5% and the assumption that some lesser percentage requires drug treatment. The figure is even more striking considering that the sample excluded the majority of the age group most commonly diagnosed and medicated for ADHD (i.e., children between 5 and 10 years of age). Treatment for ADHD has been observed to decline by as much as 20% a year between the ages of 10 and 20 (Cohen et al., 1993). Furthermore, study participants were drawn from rural and presumably underserved populations. A more recently conducted study of a single rural North Carolina county revealed that 10% of elementary students had been diagnosed with ADHD (Rowland et al., 2002).
Goldstein and Turner (2000) attempted to replicate the LeFever et al. (1999) school-based study in an entire Utah school district. Examining rates of school-based administration of ADHD medications for the 1997-98 school year, Goldstein and Turner found that less than 2% of students in Grades 1 through 5 received medication for ADHD. They concluded that ADHD went unrecognized and untreated among a sizable portion of children. This figure is not surprising given that among the 50 states, Utah ranked 44th in Ritalin consumption for the years 1997 to 1999 (Eaton & Marchak, 2001; LeFever et al., 2001). Because research suggests that rates of ADHD drug treatment may be at least twice as high as school nurse records indicate (LeFever et al., 2002), a 3% to 4% prevalence figure for Utah is a realistic estimate of the number of children who were medicated for ADHD.
Strength of the Evidence
These studies by LeFever et al. (1999, 2001), Angold et al. (2000), and Goldstein and Turner (2000) suggest that ADHD treatment practices are highly variable, which is supported by per capita assessments of DEA Ritalin distribution data. Rates of Ritalin distribution can be divided into quartile ranges such that 1st quartile reflects relatively high-use regions, the 2nd and 3rd quartiles reflect moderate-use regions, and the 4th quartile reflects low-use regions. As such, Virginia is a high-use state, North Carolina a moderate-use state, and Utah a low-use state. This corresponds to the fact that epidemiologic studies suggest that regional ADHD drug treatment rates among elementary students are as high as 17% in Virginia (high-use state), 7% to 10% in North Carolina (moderate-use state), and 3% in Utah (low-use state).
Although some have argued that the relatively low rate of school-based ADHD drug treatment observed in Utah (i.e., Salt Lake City area) is indicative of ADHD underdiagnosis, this relatively low rate of treatment may actually reflect appropriate care. At most 70% of children have a positive response to ADHD drug therapy and up to 30% require alternative treatments. Given this fact, the range of treatments available, the range in severity of the disorder, and parental preferences, drug treatment may be appropriate for well below 100% of affected children and perhaps no more than 70% of affected children. Accordingly, the rate of ADHD drug treatment would not exceed 2.1% to 3.5% (i.e., 70% of the estimated 3% to 5% of affected children). The use of stimulant medication alone among American children has already exceeded this rate (American Academy of Pediatrics, 2001). If drug treatment is at a reasonable level in a state with one of the lowest rates of Ritalin use (Utah), ADHD overdiagnosis may be significant in the 36 states with moderate to high rates of Ritalin use, suggesting that ADHD overdiagnosis and corresponding drug treatment may be more widespread than previously acknowledged by the American Medical Association (Goldman et al., 1998).
A combination of drug, behavioral, and educational interventions is generally recommended for the treatment of ADHD. Available evidence suggests that the underuse of behavioral interventions may be as problematic as the overuse of drug therapy. A survey of parents in the Jensen et al. (1999) study revealed that close to 90% of children received medication when their parents felt they needed medication for their behavior problems. However, only about half of those who felt the need for school or behavioral services actually received them and about one third who felt their children needed counseling received such services.
The Value of ADHD Treatment- Overselling the Drugs?
Clinicians sometimes justify the high rate of drug treatment for ADHD because "it works." Clinical efficacy studies are designed to determine whether drugs reduce narrowly defined symptoms (e.g., hyperactivity) without adverse side effects, as delivered in controlled research or laboratory settings and over short periods of time. Hundreds of clinical efficacy studies have documented that psychostimulant treatment reduces core symptoms associated with ADHD. Clinical effectiveness studies are designed to ascertain the benefit of drugs on major outcome variables (e.g., educational success) as delivered in routine community settings and over long periods of time (Wolraich, 1999). Little information is available regarding treatment effectiveness. Moreover, the accumulating evidence suggests that, as currently delivered in routine community settings, ADHD drug treatment is not nearly as effective as generally assumed (Pelham, 1999).
The wide-scale use of Ritalin makes it possible to evaluate the impact of such treatment on the health and well being of large cohorts of children. On a national level, increased use of psychostimulants does not appear to have led to reduced associated risks of more serious problems such as substance abuse, depression, suicide, and school dropout. Since the 1970s, when ADHD treatment began, teenage depression has skyrocketed, the rate of adolescent suicide has doubled (Centers for Disease Control and Prevention, 2000), and rates of high school dropout have remained unchanged (U.S. Department of Commerce, 2000). These population indicators beg for an examination of long-term outcomes associated with ADHD drug treatment.
In southeastern Virginia, the region with the highest documented rate of ADHD drug treatment of any community, students identified with ADHD were 3 to 7 times more likely than their peers to experience adverse educational outcomes. Regardless of whether children diagnosed with ADHD were medicated, they were far more likely than their schoolmates to be expelled or suspended from school, require special education services, and repeat a grade (LeFever et al., 2002). Such findings underscore the need for research to move beyond a focus on clinical efficacy studies to clinical effectiveness studies. Forthcoming effectiveness studies also need to consider major outcome measures (e.g., grade retention) over long periods of time (e.g., several years).
Improving Care of Children with Behavior Problems
Establishing a Public Health Agenda
Despite pronouncements to the contrary by leading ADHD researchers (Tennant, 1996), some of the communities experiencing elevated rates of ADHD identification have been compelled to act. For example, Johnston County, North Carolina, and southeastern Virginia have taken a public health approach to dealing with this complex and controversial issue (Pills for preschoolers, 2000; Simpson, 2000). These communities responded to identification of high rates of ADHD treatment by forming regional coalitions, each with a primary focus on improving the health and education of children with ADHD and related disorders. The southeastern Virginia coalition  has increased the community's awareness and understanding of ADHD issues and has been instrumental in the development and funding of several ADHD intervention and prevention grants (LeFever et al., 1999). The coalition worked with state legislators to pass legislation requiring study of the prevalence and impact of ADHD treatment in schools throughout Virginia. Through its ADHD position statement and related activities, the coalition is encouraging other communities to consider adopting a public health approach to ADHD issues (LeFever, 2001).
The efforts of such communities as southeastern Virginia are laudable; however, a more comprehensive and national public health agenda is urgently needed with regard to child mental health generally (U.S. Department of Health and Human Services, 2000) as well as ADHD specifically. The agenda should include ongoing surveillance of the rate of ADHD identification and treatment as well as outcomes associated with varying levels of treatment. Although reviews of DEA data and Medicaid claims contribute to our understanding of ADHD treatment, each has its limitations and more sophisticated tracking systems are needed (Fox et al., 2000). In addition to improving surveillance methods, strategies to prevent the emergence of disruptive behavior disorders are urgently needed. We know a great deal about the combined influence of child and environmental characteristics on the expression of developmental problems as well as how to reduce the risk of such problems (Shonkoff & Phillips, 2000). It is important to apply this knowledge to reduce the number of children who warrant psychiatric diagnoses (e.g., ADHD) and chronic psychotropic drug treatment to function adequately in school.
Promoting Judicious Use of ADHD Diagnostic Label and Associated Drug Treatment
We agree with others (Sameroff, 2001) that the overuse of psychotropic medications is a stopgap measure that is tantamount to placing the problem exclusively "in the child" rather than addressing the more complex issue of working to adapt the child and environment to each other. When it appears to be cost effective and efficient to "fix" the child through use of medication, society is unwilling to expend resources to design more development-enhancing environments that are responsive to the needs of behaviorally demanding children. Much has been learned through the years of research devoted to establishing biological treatments for ADHD. However, one must hope that society's belief in the safe and palliative nature of these drugs is warranted and that the 200 years of research showing that the long-term side effects of psychotropic drug treatment are almost always greater than initially anticipated (Olivieri, Cantopher, & Edwards, 1986) does not apply to ADHD treatments. In fact, a recent meta-analysis (Schacter, King, Langford, & Moher, 2001) concluded that (1) the outcome literature does not offer enough safety data, (2) there are many more side effects than usually appreciated, (3) few studies evaluate long-term results, and (4) there is a significant publication bias against studies showing no advantage of stimulants.
Given the lack of: (1) an identified biological marker for ADHD, (2) a single valid test or assessment approach for diagnosing ADHD, and (3) evidence of long-term effectiveness of psychotropic treatment, together with evidence that behavioral treatment is arguably as effective (Pelham, 1999) as drug treatments, judicious use of medical interventions is warranted. New practice guidelines for treating ADHD have recently been published (American Academy of Pediatrics, 2001). The following guidelines are offered to complement these new guidelines and to promote prudent use of drug treatment among children with ADHD-related symptoms and diagnoses.
- Before any kind of treatment is offered, a suspected case of ADHD requires a thorough diagnostic evaluation applying the full DSM-IV criteria (including the need to establish that the symptoms in question cannot be better accounted for by another condition and are inconsistent with developmental level).
- If a child receives a diagnosis of ADHD during the preschool years, drug treatment should be avoided, as many cases resolve by the first or second grade. Moreover, the safety and efficacy of drug treatments have not been adequately established in this population (Ghuman et al., 2001; Spencer, Biederman, & Wilens, 2000). Training in normal development and behavioral management is preferred prior to age 6.
- Although some contend that behavioral treatments do not add efficacy to medication alone (MTA Cooperative Group, 1999), behavioral interventions ought to be tried first because of their arguably comparable efficacy and lower medical risks than drug treatment (Pelham, 1999). This approach is consistent with the Hippocratic dictum "first do no harm" and reflects the costs and benefits of empirically supported treatments.
- If the child has not responded adequately after 6 months of behavioral intervention, other treatments ought to be considered, including empirically supported drug interventions.
- Psychotropic medications should not be combined unless data from controlled studies support the safety and efficacy of the combination in children (Guevara, Lozano, Wickizer, Mell, & Gephart, 2002).
ADHD is diagnosed and treated differently in communities across the United States, as evidenced by the 30-fold variation in per capita rates of Ritalin use. The probability that ADHD is diagnosed appropriately in some communities should not serve to dismiss concerns about overdiagnosis in all communities. Rates of treatment are consistently highest among younger (i.e., under age 10), nonminority, and male school-age children. The evidence of ADHD overdiagnosis is obscured when findings are reported without respect to geographic location, race, gender, and age. The fact that ADHD is clearly overdiagnosed in some communities and among some groups of children (e.g., one in every three white elementary-aged boys in southeastern Virginia) is lost in nationwide estimates of ADHD drug treatment. It is essential that mechanisms be established to track rates of child mental health diagnoses and psychotropic drug treatment and its outcomes among American children. Until we have a better understanding of these issues, it is appropriate to be judicious in our use of psychotropic medications and cautious about dismissal of concern about ADHD overdiagnosis.
- The School Health Initiative for Education (SHINE) is funded through a grant from Children's Hospital of The King's Daughter's Health System, Norfolk, Virginia.
American Academy of Pediatrics. (2001). Clinical practice guideline: Treatment of the school-aged child with attention-deficit/hyperactivity disorder. Pediatrics, 108, 1033-1044.
Angold, A., Erklani, A., Egger, H., & Costell, E. (2000). Stimulant treatment for children: A community perspective. Journal of the American Academy of Child and Adolescent Psychiatry, 39, 975-984.
Antonuccio, D., Danton, W., DeNelsky, G. Y., Greenberg, R., & Gordon, J. (1999). Raising questions about antidepressants. Psychotherapy and Psychosomatics, 68, 3-14.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. (4th ed.). Washington, D.C.: Author.
Barkley, R. (1999). Reflections on the NIH/NIMH Consensus Conference on ADHD. ADHD Report, 7(1), 1-4.
Barkley, R. A. (1990). Attention deficit hyperactivity disorder: A handbook for diagnosis and treatment. New York: Guildford Press.
Bird, H. R., Canino, G., Rubio-Stipec, M., Gould, M. S., Ribera, J., Sesman, M., et al. (1988). Estimates of the prevalence of childhood maladjustment in a community survey in Puerto Rico: The use of combined measures. Archives of General Psychiatry, 45(12), 1120-1126.
Boles, M., Lynch, F., & DeBar, L. (2001). Variations in pharmacotherapy for attention deficit hyperactivity disorder in managed care. Journal of Child and Adolescent Psychopharmacology, 11, 43-52.
Brown, R., Freeman, W., Perrin, J., Stein, M., Amler, R., Feldman, H., et al. (2001). Prevalence and assessment of attention-deficit/hyperactivity disorder in primary care settings. Pediatrics, 107(3), E46--E53.
Carey, W. (1999). Problems in diagnosing attention and activity. Pediatrics, 103, 664-667.
Carey, W. (2000, November 1). Presentation to Texas State Board of Education. Paper presented at the Texas State Board of Education, Austin, Texas.
Carmichael, A. (1996). Improving diagnosis and management of attention deficit hyperactivity disorder in Australia. Medical Journal of Australia, 165, 464-465.
Centers for Disease Control and Prevention. (2000). Suicide among children, adolescents, ad young adults-United States. Retrieved from http://www.cdc.gov/nchs/data/ hus00cht.pdf.
Cohen, P., Cohen, J., Kasen, S., Velez, C. N., Hartmark, C., Johnson, J., et al. (1993). An epidemiological study of disorders in late childhood and adolescence-I. Age- and gender-specific prevalence. Journal of Child Psychology and Psychiatry and Allied Disciplines, 34(6), 851-867.
Costello, E. (1989). Developments in child psychiatric epidemiology. American Academy of Child and Adolescent Psychiatry, Special Issue, 836-841.
Diagnosis and treatment of attention deficit hyperactivity disorder. (1998). NIH Consensus Statement, 16(2), 1-37.
Diller L (1998). Running on Ritalin: A physician reflects on children, society, and performance in a pill. New York: Bantam Books.
Diller, L. (2000, March 9). Kids on drugs. Salon. Retrieved April 2000 from http://dir.salon.com/health/feature/2000/ 03/09/kid_drugs/index.html.
Eaton,, S., & Marchak, E. (2001, June 10). Ritalin prescription rates vary widely across the country. Staten Island Sunday Advance, A23.
Esser, G., Schmidt, M. H., & Woerner, W. (1990). Epidemiology and course of psychiatric disorders in school-age children-Results of a longitudinal study. Journal of Child Psychology and Psychiatry and Allied Psychiatry, 59, 42-49.
Findling, R., & Dogin, J. (1998). Psychopharmacology of ADHD: Children and adolescents. Journal of Clinical Psychiatry, 59, 42-49.
Fox, M., Foster, C., & Zito, J. (2000). Building pharmacoepidemiological capacity to monitor psychotropic drug use among children enrolled in Medicaid. American Journal of Medical Quality, 15, 126-136.
Gadow, K. (1997). An overview of three decades of research in pediatric psychopharmacoepidemiology. Journal of Child and Adolescent Psychopharmacology, 7(4), 219-236.
Ghuman, J. K., Ginsburg, G. S., Subramaniam, G., Ghuman, H. S., Kau, A. S., & MA, R. (2001). Psychostimulants in preschool children with attention-deficit/hyperactivity disorder: Clinical evidence from a developmental disorders institution. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 516-524.
Goldman, L. S., Genel, M., Bezman, R. J., & Slanetz, P. J. (1998). Diagnosis and treatment of attention-deficit/hyperactivity disorder in children and adolescents. Journal of the American Medical Association, 279(14), 1100-1107.
Goldstein, S., & Turner, D. (2000). Initial report: The extent of drug therapy for ADHD among children in a large public school district. The ADHD Report, 8(1), 10-15.
Guevara, J., Lozano, P., Wickizer, T., Mell, L., & Gephart, H. (2002). Psychotropic medication use in a population of children who have attention-deficit/hyperactivity disorder. Pediatrics, 109, 733-739.
Hamilton, D. (2000). Cambrex mounting a challenge to Shire's big drug. Retrieved fromhttp://www.thestreet. com/stocks/biotech/1191861.html.
Health Care Financing Administration. (2001). National health expenditures projections: 2000-2010. Retrieved from http://www.hcfa.gov/stats/NHSProj/proj2000/default. html.
Jensen, P. S., Kettle, L., Roper, M. T., Sloan, M., Dulcan, M., Hoven, C., et al. (1999). Are stimulants overprescribed? Treatment of ADHD in four U.S. communities. Journal of the American Academy of Child and Adolescent Psychiatry, 38, 797-804.
Kellerher, K., & Larson, D. (1998). Prescription of psychotropics to children in office-based practice. American Journal of Disabled Children, 143, 855-859.
Kessler, J. (1980). History of minimal brain dysfunction. In H. Rie & E. Rie (Eds.), Handbook of minimal brain dysfunction: A critical review (pp. 18-52). New York: Wiley.
LeFever, G. B. (2000, November 15). ADHD: A major public health problem. Paper presented at the American Public Health Association, Boston, MA.
LeFever, G. B. (2001, June 8). University-community partnerships and school outcomes: Improving outcomes of students with ADHD and related disorders. Paper presented at the Society for Community Research and Action, Atlanta, GA.
LeFever, G. B. (2002). Limitations of using school records to assess ADHD prevalence. Norfolk, VA: Center for Pediatric Research.
LeFever, G. B., Arcona, A., & Stewart, D. (2001). Analysis of U.S. Ritalin consumption: 1997-1999. Norfolk, VA: Center for Pediatric Research.
LeFever, G. B., Dawson, K. V., Morrow, A. L. (1999). The extent of drug therapy for attention-deficit/hyperactivity disorder among children in public schools. American Journal of Public Health, 89, 1359-1364.
LeFever, G. B., Villers, M., Morrow, A., & Vaughn, E. (2002). Parental perceptions of adverse educational outcomes among children diagnosed and treated for ADHD: A call for improved school/provider collaboration. Psychology in the Schools, 39, 63-72.
Mackey, P., & Kipras, A. (2001). Medication for attention deficit/hyperactivity disorder (ADHD): An analysis by federal electorate. Retrieved July 23, 2001, from http:// www.aph.gov.au/library/pubs/cib/2000-01/01cib11.html.
Marshall, E. (2000). Epidemiology: Duke study faults overuse of stimulants for children. Science, 289, 721.
McCarthy, A., Kelly, M., & Reed, D. (2000). Medication administration practices of school nurses. Journal of School Health, 70(9), 371-376.
Miller, A. R., Lalonde, C. E., McGrail, K. M., & Armstrong, R. W. (2001). Prescription of methylphenidate to children and youth, 1990-1996. Canadian Medical Association Journal, 165, 1489-1494.
Morrow, R. C., Morrow, A. L., & Haislip, G. (1998). Methylphenidate in the United States, 1990 through 1995. American Journal of Public Health, 88, 121.
MTA Cooperative Group. (1999). A 14-month randomized clinical trial of treatment strategies for attention-deficit/hyperactivity disorder. Archives of General Psychiatry, 56, 1073-1086.
Offord, D. R., Boyle, M. H., Szatmari, P., Rae-Grant, N. I., Links, P. S., Cadman, D. T., et al. (1987). Ontario child health study. II. Six-month prevalence of disorder and rates of service utilization. Archives of General Psychiatry, 44(9), 832-836.
Olivieri, S., Cantopher, T., & Edwards, J. (1986). Two hundred years of anxiolytic drug dependence. Neuropharmacology, 25, 669-670.
Pelham, W. (1999). The NIMH multimodal treatment study for attention-deficit hyperactivity disorder: Just say yes to drugs alone. Canadian Journal of Psychiatry, 44, 981-990.
Phelps, L., Brown, R., & Power, T. (2001). Pediatric psychopharmacology: Facilitating collaborative practices. Washington, D.C.: American Psychological Association.
Pills for preschoolers: It's time to slow the rush to behavioral medication. Virginian-Pilot, B10.
Popper, C. (2000). Pharmacologic alternatives to psychostimulants for the treatment of attention-deficit/hyperactivity disorder. Child and Adolescent Psychiatric Clinics of North America, 9(3), 605-646.
Rowland, A., Umbach, D., Stallone, L., Naftel, A., Bohlig, E., & Sandler, D. (2002). Prevalence of medication treatment for attention deficit-hyperactivity disorder among elementary school children in Johnson County, North Carolina. American Journal of Public Health, 92, 231-234.
Safer, D. J. (1971). Drugs for problem school children. Journal of School Health, 41, 174-179.
Safer, D. J. (1999, August). Stimulant treatment in Maryland public schools. Paper presented at the American Psychological Association Annual Meeting.
Safer, D. J. (2000). Are stimulants overprescribed for youths with ADHD? Annals of Clinical Psychiatry, 12(1), 55-62.
Safer, D. J., & Krager, J. M. (1988). A survey of medication treatment for hyperactive/inattentive students. Journal of the American Medical Association, 260(15), 2256-2258.
Safer, D. J., & Krager, J. M. (1989). Hyperactivity and inattentiveness. School assessment of stimulant treatment. Clinical Pediatrics, 28(5), 216-221.
Safer, D. J., & Krager, J. M. (1994). The increased rate of stimulant treatment for hyperactive/inattentive students in secondary schools. Pediatrics, 94(4, pt. 1), 462-464.
Safer, D. J., & Malever, M. (2000). Stimulant treatment in Maryland Public Schools. Pediatrics, 106, 553-559.
Safer, D. J., & Zito, J. M. (1999). Psychotropic medications for ADHD. Mental Retardation and Developmental Disabilities, 5, 237-242.
Safer, D. J., Zito, J. M., & Fine, E. M. (1996). Increased methylphenidate usage for attention deficit disorder in the 1990s. Pediatrics, 98(6, pt. 1), 1084-1088.
Sameroff, A. (2001, July 12). Risk and resilience from infancy to adolescence: Is it better to change the child or the context? Paper presented at the 2001 OSEP Research Project Directors' Conference, Washington, D.C.
Schachter, H. M., King, J., Langford, S., & Moher, D. (2001). How efficacious and safe is short-acting methylphenidate for the treatment of attention-deficit disorder in children and adolescents? A meta-analysis. Canadian Medical Association Journal, 165, 1475-1488.
Scahill, L., & Schwab-Stone, M. (2000). Epidemiology of ADHD in school-age children. Child and Adolescent Psychiatric Clinics of North America, 9, 541-555.
Shonkoff, J., & Phillips, D. (2000). From Neurons to Neighborhoods: The Science of Early Childhood Development. Washington, D.C.: National Academy Press.
Simpson, E. (2000, March 5). Group seeks reasons for medication rate. Virginian-Pilot, B1.
Sinha,, G. (2001, June). New evidence about Ritalin: What every parent should know. Popular Science, 48-52.
Spencer, T., Biederman, J., & Wilens, T. (2000). Pharmacotherapy of attention deficit hyperactivity disorder. Child and Adolescent Psychiatry Clinics of North America, 9, 77-97.
Sroufe, L., & Stewart, M. (1975). Treating problem children with psychostimulants. New England Journal of Medicine, 289, 407-413.
Tennant, D. (1996, April 9). Combating ADD: Expert says children are not being overdiagnosed or overtreated for the disorder, which is genetic. Virginian-Pilot, E1.
Todd, R. (2000). Genetics of attention deficit/hyperactivity disorder: Are we ready for molecular genetic studies? American Journal of Medical Genetics, 96(3), 241-243.
United Nations Information Service. (1997). INCB sees continuing risk in stimulant prescribed for children. Vienna, Austria: Author.
U.S. Census Bureau. (2000). Status of dropout rates of 16- through 24-year-olds, by race and ethnicity. Current Population Survey, October. Washington, D.C.: U.S. Department of Commerce.
U.S. Department of Health and Human Services. (2000). Report on the Surgeon General's Conference on Children's Mental Health: A national action agenda. Paper presented at the Surgeon General's Conference on Child Mental Health, Washington, D.C.
Wakefield, J.C. (1992). The concept of mental disorder on the boundary between biological facts and social values. American Psychologist, 47, 373-388.
Wender, P., Bierman, J., French, F., et al. (1971). Minimal brain dysfunction in children. New York: John Wiley and Sons.
Wilens, T., & Biederman, J. (1992). The stimulants. Psychiatric Clinics of North America, 15, 1991-222.
Wolraich, M. (1999). The difference between efficacy and effectiveness research in studying attention-deficit/hyperactivity disorder. Archives of Pediatric and Adolescent Medicine, 153, 1220-1221.
Wolraich, M. L., Hannah, J. N., Pinnock, T. Y., Baumgaertel, A., & Brown, J. (1996). Comparison of diagnostic criteria for attention-deficit hyperactivity disorder in a county-wide sample. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 319-324.
Zarin, D., Suarez, A., Pincus, H., Kupersanin, E., & Zito, J. (1998). Clinical and treatment characteristics of children with attention-deficit/hyperactivity disorder in psychiatric practice. Journal of the American Academy of Child and Adolescent Psychiatry, 37, 1262-1270.
Zito, J., & Safer, D. J. (2001). Services and prevention: Pharmacoepidemiology and antidepressant use. Biological Psychiatry, 49(12), 1121-1127.
Zito, J. M., Safer, D. J., dosReis, S., Gardner, J. F., Soeken, K., Boles, M., & Lynch, F. (2002). Rising prevalence of antidepressants among US youths. Pediatrics, 109, 721-727.