In the United States, identification of attention-deficit/hyperactivity disorder (ADHD) among children and adolescents showed the greatest increase of all categories of psychosocial problems 1, and the percentage of children with a diagnosis of ADHD receiving medications increased from 32% in 1979 to 78% in 1996 123. For the mid-1990s, overall psychostimulant treatment prevalence rates in children and adolescents were found at about 2.5% 4, ranging from 1.9% in a California health plan 5 to 3.0% in national database analyses 67. In the late 1990s, growth rates of stimulant prescriptions accelerated in the United States 89. For 1997, treatment prevalence was reported at 4.1% (or 78% of children with ADHD) 10. In a survey of school nurses conducted in Baltimore, Maryland, in 1998, 2.92% of all public school students (N = 816,465) were administered a medication for ADHD in school; 84% of those (2.46%) received methylphenidate 4. In 1999, in a nationally representative, commercially insured sample population 5 to 14 years old, the one-year prevalence of stimulant treatment was 4.2 percent in the United States 11. According to another survey conducted between 1997 and 1999 among parents of elementary school children in public elementary schools in North Carolina, medication prevalence was even 7% in the population studied, with stimulants accounting for 93% of the prescriptions 12. More recent data have been somewhat contradictory, with one analysis finding a treatment prevalence of 4.8% in children (age 6–12 years) and 3.2% in adolescents (age 13–19 years), suggesting that the steep increase of the late 1990s may have attenuated in these age groups 13, whereas another study published by Medco Health Solutions 14, a large U.S. pharmacy benefit management organization, still found the number of individuals age 19 years or younger using ADHD medications increasing by 49% (males) and 82% (females) between 2000 and 2004.

Meanwhile, the growth of pediatric psychotropic prescriptions has become an international phenomenon 15. For stimulant treatment, similar trends as in the US – albeit at lower absolute levels – have been observed in Canada 16 and Europe. For instance, in Spain the consumption of methylphenidate increased by 8% annually from 1992 to 2001 17. A study in six general practices in the Netherlands showed a threefold rise of methylphenidate users between 1998 and 2003 18. In England, the number of methylphenidate prescriptions dispensed in the community increased from 126,000 in 1998 to 389,200 in 2005 (or +207%; cf. Figure 1a), compared to a total growth of prescriptions of +40% during the same period 19. In Germany, the number of defined daily doses (DDDs) of methylphenidate prescribed for outpatients insured by statutory sick funds ("Gesetzliche Krankenversicherung", GKV, covering approximately 90% of the German population) grew 47-fold between 1992 and 2005 20, while total prescriptions written in Germany decreased by 41% in the same period (cf. Figure 2).

Much of the debate triggered by this trend has focused on medical aspects, such as concerns about potential overuse of psychostimulants 2122 and adverse treatment effects 23. It has, however, been argued that underuse may be more of a problem, at least in Europe 24, and low rates of recognition of ADHD or hyperkinetic disorder in the United Kingdom were associated with insensitive diagnostic assessments 25. Recent German methylphenidate prescription analyses based on a regional sample of 11,235 children and adolescents with a diagnosis of ADHD in Germany (year 2003) did not reveal overuse 26. Even for the United States, no evidence was found to substantiate claims of abuse and inappropriate use of methylphenidate 27, although there have been occasional reports on diversion to peers of stimulant medications, especially short-acting formulations 2829.

Beyond clinical implications, in an era of limited available resources, the economic dimension associated with increased health care utilization can no longer be ignored. In economic terms, the opportunity cost of medical interventions will be approximated by their budgetary impact – in particular, if a payers' perspective is adopted. Indeed, budgetary impact analyses are requested by a growing number of health care policy makers as an input into the decision-making process about health care resource allocation 30. In practice, results of such analyses are frequently interpreted with a notion of "affordability" in mind. Unfortunately budget impact may exceed the pure effects of increased treatment rates if, for instance, rates of diagnosis (administrative prevalence), unit costs (for instance owing to introduction of novel, usually more expensive treatment options, which may be used in addition to or substituting cheaper alternatives), or treatment intensity (in terms of dose or duration) change simultaneously. No doubt, all these factors come into play in ADHD among children and adolescents. More than this, they may interact with each other, reinforcing their combined effects: for instance, availability of more expensive novel medications will result in enhanced communication and promotional efforts by their manufacturers, which in turn will – in addition to their direct influence on physician treatment choices (i.e., increase market share) – influence awareness and identification of the disorder (i.e., increase market size and aggregate number of prescriptions 253132). Furthermore, recent results of long-term clinical trials such as the landmark NIMH Multimodal Treatment Study (MTA) have demonstrated the effectiveness, in particular in terms of ADHD core symptom relief, of an intense medication management strategy 333435363738), thus contributing to the increasing acceptance and intensity of pharmacotherapy.

With the introduction of new products commanding higher units costs (cf. Table 1), growth of pharmaceutical spending has outpaced the rising number of prescriptions. In the United States, prescription drug expenditures for ADHD in children increased by 183% between 2000 and 2003. By 2003, spending on behavioral medications to treat children had overtaken both the antibiotic and asthma segments, which are traditionally high-use categories in pediatric medicine 39. According to IMS data, total U.S. sales of ADHD drugs reportedly rose from US-$759 million in 2000 to US-$3.1 billion in 2004 40. Analysts from Merrill Lynch expect the U.S. ADHD market (including adult patients) to exceed US-$4 billion by 2010 on the back of new products (cf. below, Table 241).

In Germany, total GKV outpatient spending for psychostimulants rose from €1.25 m in 1995 to €23.7 m in 2002 (before the first modified-release product had been launched) and €51.6 million in 2004 20, 56% of which were accounted for by MPH-MR12. Spending for ADHD medications reached €82 million in 2005, exceeding our earlier forecasts, which had predicted annual expenditures of approximately €108 million (range from €62 to €155 million) by 2009 4243. Apart from the launch of atomoxetine in March 2005, this substantial increase was primarily driven by spending for two modified-release preparations of methylphenidate (cf. Table 1), both of which ranked among the top-100 products in terms of sales 44.

In England spending for ADHD-related pharmacotherapy, including modafinil and atomoxetine, increased from 1998 (₤3.1 million) to 2005 (₤24.4 million) by +695% (Figure 1b), again exceeding the growth of prescriptions (for ADHD treatments between 1998 and 2005, +132%)19. One new product (MPH-MR12, cf. Table 1) alone accounted for 42% of prescriptions, and 45% of pharmaceutical expenditures, for ADHD in 2005 19. Average costs per prescription rose steadily from ₤13.68 in 1998 to ₤46.94 in 2005, driven by price increases for dexamphetamine and a shift to more expensive new products: Combined, all new products with once-daily administration (in principle, MPH-MR08, MPH-MR12, Ritalin SR, and ATX) accounted for 54% of total ADHD-related prescriptions and 65% of sales in 2005 19. These Prescription Cost Analysis data for England do not include items dispensed in hospitals.

The high rate of adoption of new products with once-a-day administration schedules reflects the well-known problems associated with administration of a mid-day dose (during school) in children and adolescents with ADHD 4546. Obviously, here a disorder-specific clinical need contributes to the fact that current trends of ADHD prescribing behavior mirror a more general pattern encountered in European pharmaceutical markets, namely, that the diffusion of new products consistently represents the single most important growth driver 47. These observations raise the urgent question whether, and to what extent, pharmaceutical spending for children and adolescents with ADHD may continue to escalate in the foreseeable future. It is the purpose of this paper to shed some light on this issue, using the examples of England and Germany.

A forecasting model was developed to project the likely pharmaceutical expenditures for children and adolescents (age 6–18 years) with a diagnosis of ADHD in England and Germany through 2012, specifying assumptions and assumed relationships between variables in a transparent manner. Model outcomes were used to estimate the impact on drug spending by individual German physicians involved in care for ADHD patients, after validation of the model by replicating historic data (from 1998 to 2005). Key uncertainties were addressed using scenario analysis, combining objective information with specific assumptions about future events 48.

The model was restricted to pharmaceutical spending. It combined, in a hierarchical structure, (1) epidemiological information (demographic and prevalence data) with assumptions on (2) recognition rates (diagnosis prevalence), (3) rate of patients receiving drug therapy (treatment prevalence), (4) availability and adoption of new products, incorporating information on therapeutic profiles, (5) diffusion and market shares for alternative preparations, by category and by product, including generic substitution, (6) treatment intensity (expressed as average number of days times defined daily doses), (7) acquisition cost per defined daily dose for each product, from the perspectives of the NHS (England) or the GKV (Germany), respectively. For validation of the model, available data on the variables above were used to compare model outcomes with the historic evolution of drug spending in England and Germany. Extensions of the model were used to estimate the impact of the projections on drug budgets of individual pediatricians and specialists in child and adolescent psychiatry participating in care for patients with ADHD in Germany. Details on data sources and assumptions are provided in Table 2.

Without exception, all plausible scenarios for England and Germany point to a further increase of expenditures for ADHD medication. The striking extent of the projected increase may appear counterintuitive, but it is easily explained by the simultaneous and multiplicative effects exerted by four variables, i.e., increases in diagnosis prevalence, treatment prevalence, treatment intensity, and the shift to more expensive new products. Over a wide range of assumptions, all projections concur indicating a strong impact of this increase on the individual drug budgets of physicians treating patients with ADHD.

Some important limitations of the present analysis warrant discussion. First, the present analysis does not purport to convey value judgments. It has little to nothing to say about the clinical appropriateness of the prescriptions analyzed; its mere focus is on their budgetary impact.

Second, there is substantial uncertainty around future events. Compounds in development may be discontinued, marketed drugs may be withdrawn because of serious adverse events, safety concerns may slow down diffusion of new products, and so on. The dynamic health policy environment is yet another factor. However, it would hardly be a prudent response to uncertainty to abandon planning and taking a proactive role in the pursuit of efficient service provision.

In strategic management, scenario analysis is a well-established tool to deal with environmental uncertainty 65. The practice of scenario planning implicitly acknowledges that "best guesses" of future events may be wrong. Therefore, any scenario should not be confused with a forecast of the future. Multiple scenarios are pen-pictures of a range of plausible futures (Figures 3 and 4). Though each individual scenario has an infinitesimal probability of actual occurrence, combined a range of scenarios can shed light on possible future outcomes. This in turn should enable to plan for the range of futures that could, plausibly, unfold.

A third important limitation of this study concerns its restricted scope. Although pharmaceutical spending is likely to rise faster than other ADHD-related expenditures, it is certainly not the only cost component to be considered in a complete budgetary impact analysis 3066. Even more importantly, cost analyses illuminate just one half of the health economic equation; they do not provide information on "value for money", frequently analyzed in terms of cost-effectiveness 67. Cost-effectiveness analyses of ADHD treatment strategies, however, have been rare 68, are only now beginning to appear 697071, and have generally been limited to a one-year time horizon 72.

In combination, the scenarios developed here strongly suggest that the trend of rising drug expenditures for ADHD may not abate in the near future (Figure 3).

They further indicate that physicians involved in providing care for children and adolescents with ADHD should anticipate escalating expenditures and challenging questions from payers (Figure 5). In particular for German child and adolescent psychiatrists, as well as for pediatricians in private practice, there will be a growing need to demonstrate appropriate prescribing practices and the achievement of clinical benefits commensurate with levels of spending.

Likewise, pharmaceutical manufacturers developing and selling products for the treatment of patients with ADHD will need to produce timely evidence supporting the economic value of their products.

Finally, it should be noted that the present study was limited to ADHD in children and adolescents age 6 to 18 years. There is reason to assume that adult ADHD may be associated with substantial spending dynamics, which has only just begun to show 73. This fact, however, lends further support to the key conclusion of the present study, namely, that there will be a rapidly growing need for health economic evidence on the value of clinical interventions for ADHD.

ARM – armodafinil

ATX – atomoxetine

DEX – dexamphetamine

dMPH – dexmethylphenidate

FDA – Food and Drug Administration, the United States regulatory agency deciding on marketing authorizations for pharmaceutical products

GKV – Gesetzliche Krankenversicherung, statutory health insurance in Germany, covering approximately 90% of the population

IR – immediate release (formulation)

LisDEX – lisdexamphetamine, a prodrug of amphetamine in phase III clinical development

MAS – mixed amphetamine salts

MOD – modafinil

MPH – methylphenidate

MR – modified release (formulation)

MR08 – modified release formulation with duration of action of approximately 8 hours

MR12 – modified release formulation with duration of action of approximately 12 hours

NHS – National Health Service (England)

SHI – statutory health insurance in Germany (see GKV)

TDS – transdermal system