1744-9081-1-16.fm

Behavioral and Brain Functions
Research
Efficacy of atomoxetine in adult attention-Deficit/Hyperactivity
Disorder: a drug-placebo response curve analysis
Stephen V Faraone*1, Joseph Biederman2, Thomas Spencer2,
David Michelson3, Lenard Adler4, Fred Reimherr5 and Stephen J Glatt6
Address: 1Department of Psychiatry, SUNY Upstate Medical University, Syracuse, NY 13210, USA, 2Department of Psychiatry, Harvard Medical School, Massachusetts General Hospital, Boston, MA 01880, USA, 3Lilly Research Laboratories, Indianapolis, IN 46285, USA, 4New York University School of Medicine, New York, NY 10016, USA, 5Mood Disorders Clinic, Department of Psychiatry, University of Utah Health Sciences Center, Salt Lake City, UT 84132, USA and 6Institute of Behavioral Genomics, Department of Psychiatry, University of California, San Diego, La Jolla, CA 92093-0603, USA Email: Stephen V Faraone* - [email protected]; Joseph Biederman - [email protected]; Thomas Spencer - [email protected]; David Michelson - [email protected]; Lenard Adler - [email protected]; Fred Reimherr - [email protected]; Stephen J Glatt - [email protected] * Corresponding author Published: 03 October 2005 Received: 22 August 2005Accepted: 03 October 2005 Behavioral and Brain Functions 2005, 1:16
2005 Faraone et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licenswhich permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background: The objective of this study was to evaluate the efficacy of atomoxetine, a new and
highly selective inhibitor of the norepinephrine transporter, in reducing symptoms of attention-
deficit/hyperactivity disorder (ADHD) among adults by using drug-placebo response curve
methods.
Methods: We analyzed data from two double-blind, placebo-controlled, parallel design studies of
adult patients (Study I, N = 280; Study II, N = 256) with DSM-IV-defined ADHD who were recruited
by referral and advertising. Subjects were randomized to 10 weeks of treatment with atomoxetine
or placebo, and were assessed with the Conners Adult ADHD Rating Scales and the Clinical Global
Impression of ADHD Severity scale before and after treatment.
Results: Those treated with atomoxetine were more likely to show a reduction in ADHD
symptoms than those receiving placebo. Across all measures, the likelihood that an atomoxetine-
treated subject improved to a greater extent than a placebo-treated subject was approximately
0.60. Furthermore, atomoxetine prevented worsening of most symptom classes.
Conclusion: From these findings, we conclude that atomoxetine is an effective treatment for
ADHD among adults when evaluated using several criteria.
however, the use of other agents, such as bupropion and Several compounds are now recognized as effective treat- desipramine, has also received some support. In addition ments for the major symptoms of attention-deficit/hyper- to these, atomoxetine, a highly selective noradrenergic activity disorder (ADHD) in adulthood. The most reuptake inhibitor with little affinity for other neurotrans- effective of these include methylphenidate and dextroam- mitter systemhas been shown to be well tolerated phetamine (or mixed dextro- and levoamphetamine); and effective in reducing the symptoms of ADHD in (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
Table 1: Summary of Effects of Atomoxetine and Placebo on ADHD Symptoms in Adults*
Placebo (n = 134) Atomoxetine (n = 133) Placebo (n = 124) Atomoxetine (n = 124) Total ADHD Symptom Scores Inattentive Score Self-Rated CAARS Total ADHD Symptom Score Inattentive Score *Data adapted from Michelson et al., 2003Data are presented as mean ± standard deviation adulthood. In fact, the benefits of atomoxetine for adults with ADHD have now been demonstrated in three studies of adult patients [t establishing the To provide a foundation for interpreting the results of superiority of atomoxetine over placebo in reducing inat- drug-placebo response curve analysis of the effects of ato- tentive, hyperactive, and impulsive symptoms of the ill- moxetine, we have displayed in Tableary of nea result of its demonstrated efficacy and low results of the standard analyses of these data, which were occurrence of clinically meaningful side effects originally presented by Michelson et al. []. Although data oxetine recently became the first non-stimulant medica- on clinical global impression (CGI) endpoints or Conners tion approved for use in the United States for the Adult ADHD Rating Scale (CAARS) ADHD index scores treatment of ADHD in adults.
were not presented in the initial report, it is clear that ato-moxetine had significant efficacy relative to placebo on all Thus, the ability of atomoxetine to reduce symptoms of other measures derived from the CAARS or CGI assess- ADHD among adults has been sufficiently established; ments. The most reliable reductions in ADHD symptoms however, several key questions about its clinical utility elicited by atomoxetine were seen for clinician-rated glo- remain unresolved. For example, although the initial bal impressions of ADHD severity, and investigator- and studies of the efficacy of atomoxetine provided useful self-rated total ADHD symptoms. Atomoxetine more information for clinicians treating adults with ADHD, robustly reduced inattentiveness than hyperactivity and such as the average magnitude of the decrease in ADHD impulsivity, as assessed by both investigators and sub- symptoms associated with drug treatment and the reliabil- jects. Overall, these data provided strong and conclusive ity of this effect, the standard methods of data presenta- evidence that atomoxetine was superior to placebo in tion in these reports do not provide information about reducing the symptoms of ADHD in adulthood, warrant- the full range of effects of this compound. To further char- ing further analysis by drug-placebo response curve acterize the clinical performance of atomoxetine, we com- pleted a drug-placebo response curve analysis of the datainitially reported by Michelson et al. [] This method, CAARS Investigator Ratings
described by Faraone et al. [], is a generalization of Figures, w the results for investigator ratings receiver operating characteristic (ROC on the CAARS. Figureares the effects of atomoxet- has been widely applied to assessing the accuracy of diag- ine on total ADHD score with those of placebo. For differ- nostic tests [he goal of this method is to identify ent definitions of responsiveness (i.e., different CAARS additional characteristics of drug-placebo differences that cutting scores), the points on the curve illustrate the have already been shown to be statistically significant, results of two calculations: the rate of response to drug including: 1) the size of the effect using different response and the rate of response to placebo. For example, the criteria; 2) the nature of individual responses; and 3) the point in Fi is located at coordinates [0.4, portion of the drug's effect that is due to symptom 0.54], which indicates that 40% of those treated with pla- improvement, the prevention of symptom worsening, or cebo achieved a change of -7 in total investigator-rated ADHD symptoms on the CAARS, whereas 54% of those (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CAARS Investigator- Rated Total ADHD Score CAARS Investigator-Rated Total ADHD Score.
treated with atomoxetine attained the same level of each figure) also shows the diagonal line of responsiveness. Sequential points further down the curve no effect, which allows us to visualize the size of the drug (i.e., toward the origin) specify increasingly stringent effect as the degree to which the drug-placebo response thresholds for defining improvement (i.e., larger curve rises above it. If outcome on drug were worse than decreases in investigator-rated total ADHD score), while outcome on placebo, then the drug-placebo response points further up the curve denote the proportions of each curve would fall below the line of no effect; however, as treatment group that responded to treatment as deter- Figure hows, atomoxetine produced a drug-placebo mined by increasingly lenient criteria for improvement.
response curve that was always above the diagonal line of Thus, these points and the curve that joins them illustrate no effect. This indicates that atomoxetine outperformed how the drug- and placebo-response rates change as the placebo throughout the full range of outcome scores. The cutting score used to define improvement is incrementally area under the curve (AUC) is 0.60, which means that ato- changed. From the curve, it is clear that if response criteria moxetine outperformed placebo 60 percent of the time, between total ADHD symptom change scores of -1 and - regardless of cutting score.
14 are used as the cutting score, a greater proportion ofatomoxetine-treated individuals than placebo-treated In addition to this information, and unlike a traditional patients will attain that level of symptom improvement.
statistical anine if In other words, over this range of cutting scores, the the effects of atomoxetine were due to its ability to majority of individuals judged as responsive to treatment improve the symptoms of ADHD, prevent their worsen- will have received atomoxetine rather than placebo. Near ing, or both. For example, for CAARS total ADHD symp- the most extreme cutting scores (in this case, the points tom change scores, a value of 0 indicates no change, and labeled 15 and -46), those treated with placebo were as this point on the drug-placebo response curve is labeled.
likely as those treated with atomoxetine to reach response At this point on the curve, we see that 82% of subjects receiving atomoxetine had a score of 0 or greater, i.e., only18% of atomoxetine-treated patients experienced a wors- (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CAARS Investigator-Rated Inattention CAARS Investigator-Rated Inattention Subscale.
ening of their symptoms. In contrast, 71% of placebo- was lower than that for inattention (0.61). These results treated subjects had a score of 0 or greater, which indicates indicate that atomoxetine was less effective in reducing that symptoms worsened in 29% of these patients. Thus, (and preventing the worsening of) hyperactivity than in the response curve clearly demonstrates that atomoxetine improving attention. However, it is still clear that, relative not only reduced symptoms, but prevented their worsen- to placebo, atomoxetine was an effective treatment for ing as well.
hyperactivity when any symptom change score criteria ofless than 0 was used as the response criterion.
Fiillustrates the effects of atomoxetine on investiga-tor-rated inattention on the CAARS. As with total ADHD In Figurects of atomoxetine and placebo on symptoms, inattentive symptoms seem to be more effec- investigator-rated ADHD index scores are plotted, and the tively treated by atomoxetine than placebo, as the drug- more restricted rise in this curve is immediately apparent placebo response curve was above the diagonal line of no relative to that seen in earlier figures. The AUC for this effect and the AUC was 0.61. Relative to total ADHD curve (0.59) was significant (p < 0.001), indicating that symptoms, there exists for inattention a narrower range of atomoxetine was more likely than placebo to reduce the cutting scores over which the greatest difference in respon- ADHD index score. However, there is a distinct peak in siveness was seen between atomoxetine- and placebo- this curve at a response criterion of approximately -8, treated subjects. At cutting scores of -4 through -8, approx- where atomoxetine-treated subjects were approximately imately 50% more atomoxetine-treated subjects reached twice as likely as placebo-treated subjects to attain this the response criterion than did placebo-treated subjects.
level of symptom improvement; at other cutting scores(e.g., -3), the benefits of atomoxetine were much more Figure s the effects of atomoxetine and placebo on investigator-rated hyperactivity on the CAARS. Relative toFition), the drug-placebo response curve in CAARS Self Ratings
Finot rise as far above the diagonal line of no Figures for self-ratings on effect and, consequently, the AUC for hyperactivity (0.58) the CAARS. In general, the effects of atomoxetine on self- (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CAARS Investigator-Rated H yperactive Subscale CAARS Investigator-Rated Hyperactive Subscale.
ratings on the CAARS mirrored its effects on investigator more effective than placebo in reducing inattentive symp- ratings. From Figu it is clear that, as with investigator- toms. Of note, the drug-response curve for self-ratings of ratings of this measure, self-rated total ADHD symptoms hyperactivity very closely matched that for investigator- were more likely to be reduced in atomoxetine-treated ratings of this measure (Figureg an apprecia- subjects than in those receiving placebo. In fact, the AUC ble amount of divergence in subjects' perceptions of the for this measure (0.61) was virtually identical to that effects of treatment on their constituent symptom clusters.
observed on investigator ratings of this measure, and ben- As expected, self-rated ADHD index change scores eficial effects of atomoxetine were observed over roughly mirrored investigator ratings in showing a sizeable benefit the same range of cutting scores. More than 60% of sub- of atomoxetine over placebo, especially when symptom jects treated with atomoxetine attained the median change scores in the range of -4 to -9 were used as response score (-6), while only approximately 40% of response criteria (Figure those treated with placebo saw this level of improvement;thus, when the median response score was used as the cut- CGI Clinician Ratings
ting score for defining responsiveness, atomoxetine had Figure depicts the effects of atomoxetine and placebo on greater efficacy on self reported total ADHD symptoma- ADHD symptom severity change scores for the CGI. As tology than on investigator ratings of this measure (cf, Fig- with the CAARS measures, the drug-placebo response curve occupied the space above the diagonal line of noeffect and the AUC approximated 0.60, indicating an The shape and position of the drug-placebo response advantage of atomoxetine over placebo in reducing clini- curve for self-rated CAARS inattention change scores (Fig- cian-rated ADHD severity. Also in accord with the CAARS ure quite similar to the curve for investigator rat- measures, a small protection from symptom worsening ings of this measure (Figifically, the curve was was afforded by atomoxetine, as approximately 10% of above the diagonal line of no effect and had a significant subjects who received the drug deteriorated clinically, AUC, indicating that atomoxetine administration was while approximately twice as many placebo-treated (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CAARS Investigator- CAARS Investigator-Rated ADHD Index.
subjects did so. The median change in CGI in the com- Drug-placebo response curves provide an easily interpret- bined group of atomoxetine- and placebo-treated subjects able format for further evaluating clinically informative was -1, a response criterion attained by almost 55% of characteristics of a compound with proven efficacy.
drug-treated subjects but by only approximately 40% of Because atomoxetine has demonstrable efficacy, drug-pla- those receiving placebo. As expected, these greater rates of cebo response curve analysis of its performance against improvement (and protection from deterioration) led to placebo was warranted. Collectively, the drug-placebo the attainment of lower CGI endpoint scores in the atom- response curves presented here for each of the different oxetine-treated gro reporters and the various dependent measures paint aconsistent picture of the benefits of atomoxetine. First, it is clear that atomoxetine is superior to placebo in reduc- The results of two large randomized, double-blind, pla- ing total ADHD symptoms as well as individual symptom cebo-controlled trials of atomoxetine for the treatment of clusters, such as inattention and hyperactivity. For each of ADHD in adults were initially reported by Michelson et al.
these measures, the drug-placebo response curve was umented the superiority of this compound always situated above the line of no effect, indicating that relative to placebo in reducing total ADHD symptoms, as subjects were more likely to respond to atomoxetine than well as inattentive and hyperactive symptoms of the ill- to placebo over the entire range of possible criteria of ness. Due to the ample size and rigorous design of these responsiveness. In addition, it is clear that atomoxetine studies, as well as the strong statistical significance of their targeted the core features of ADHD rather than only one results, the efficacy of atomoxetine has been firmly estab- of its most conspicuous features of inattention and hyper- lished. However, the simple knowledge that, on average, activity, as AUCs across total, inattention, and atomoxetine is efficacious does not tell clinicians much hyperactivity change scores were quite similar. Second, about its full range of effect.
responsiveness to atomoxetine was reliably assessed by (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
Rated Total ADHD Score CAARS Self-Rated Total ADHD Score.
clinicians, investigators, and patients, as the AUCs for the details that often are lost in a standard analysis, such as various dependent measures varied little (0.58–0.61) the ability of atomoxetine to improve outcome and across reporters. Third, atomoxetine not only reduced the prevent worsening throughout the full range of outcome symptoms of ADHD, but prevented the worsening of scores. The present drug-placebo response analysis these symptoms as well, a finding that has been seen for provided strong support for the efficacy of atomoxetine drug-placebo response curve analyses of other medica- relative to placebo for reducing inattention, hyperactivity, ti]. In contrast however, these prior drug-pla- and total ADHD symptoms assessed by a variety of cebo response curve analyses have also revealed stronger reporters, and for preventing the worsening of these effects of other medications on clinician-rated ADHD symptoms. The finding that atomoxetine is efficacious symptomatology, as evidenced by AUCs of 0.86 for through the full range of outcome further emphasizes the A], 0.89 for methylphenid clinical value of treating ADHD adults with this and 0.93 for desipramine [,compared to the AUC of approximately 0.60 presently observed foratomoxetine.
In conclusion, we have extended the statistical results of Two identical randomized, double-blind, placebo-con- Michelson et al. by using drug-placebo response curves trolled studies were conducted concurrently at 17 (Study to describe the clinical significance of the efficacy of ato- I) and 14 (Study II) outpatient sites in North America.
moxetine in the treatment of ADHD among adults. Our Each site's institutional review board evaluated and method of data presentation provides readers and clini- approved the study protocol, and written informed con- cians with a means of understanding the nature of the sent was obtained from each patient. Adults who met effects of this drug, and the degree to which they are clin- DSM-IV criteria for ADHD as assessed by clinical inter- ically relevant. Rather than collapsing individual view and confirmed by the Conners' Adult ADHD Diag- responses into means or single rates of response, the drug- nostic Interview for DSM-IV were recruited from clinics placebo response curve illustrates clinically meaningful and by advertisement. Patients were required to have at (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
Rated Inattention Subscale CAARS Self-Rated Inattention Subscale.
least moderate symptom severity, and the diagnosis had oxetine, while 211 placebo-treated subjects completed the to be corroborated by a second reporter for either current trial, a difference that was not significant.
symptoms (by a significant other) or childhood symp-toms (by a parent or older sibling). Patients who met diagnostic criteria for any other Axis-I disorder using the The outcome measures examined in this study were Structured Clinical Interview for DSM-IV were excluded, derived from the CAARS and the CGI. A clinician com- as were patients with serious medical illness or habitual pleted the CGI before and after the treatment regimen, substance abuse.
while both the subject and an investigator completed theCAARS before and after treatment. The three groups of pri- Atomoxetine and Placebo Administration
mary dependent measures of this study included: 1.) cli- Following an initial one-week medication washout and nician-rated CGI ADHD Severity change scores and evaluation period, patients entered a two-week placebo endpoint scores; 2.) investigator-rated inattention, hyper- lead-in phase. Patients who maintained the initial severity activity/impulsivity, total symptoms, and ADHD index criteria required for study entry were randomized to scores on the CAARS; and 3.) self-rated inattention, receive atomoxetine or placebo for a 10-week period. Ato- hyperactivity/impulsivity, total symptoms, and ADHD moxetine was administered in evenly divided doses in the index scores on the CAARS.
morning and late afternoon/early evening beginning at atotal daily dose of 60 mg. Patients with residual symp- Drug-Placebo Response Curve Analysis
toms received higher doses of up to 90 mg/day after two The rationale and methodology for drug-placebo weeks and 120 mg/day after four weeks. If patients devel- response curve analysis methods are described in detail by oped problems tolerating this regimen, the dose could be Faraone et al. [] The goal of response curve analysis is not decreased to the last tolerated dose or an increase in dos- to demonstrate statistically significant group differences; age could be omitted. Across both studies, 270 subjects rather, this method provides an alternative means of dis- received atomoxetine, while 263 subjects received pla- playing differences that have already been demonstrated cebo. Of these, 197 completed acute treatment with atom- to be statistically significant. Thus, it does not replace a (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CAARS Self-Rated Hyperactive Subscale Figure 7
CAARS Self-Rated Hyperactive Subscale.
standard statistical analysis, but augments that analysis by variable is a change score, also label the point correspond- showing the clinical significance of drug effects. For the ing to no change; 6.) Plot the line of no effect, which is the present study, the use of drug-placebo response curve diagonal line from the [0, 0] point to the point. Each analysis is warranted, as the statistically significant effects point along a curve represents an observed outcome score of atomoxetine on reducing symptoms of ADHD in adults on that measure, and the points on each plot are then con- have been documented previously.
nected by line segments. The line of no effect comprises allpoints for which the proportion of subjects who respond The drug-placebo response curve is constructed in the fol- to drug is the same as the proportion who respond to lowing six steps: 1.) Choose an outcome variable, for example the change in CAARS Inattention score frombaseline to the end of the study; 2.) At each observed The drug-placebo response curve is a graphical method of score, calculate separately for the drug and placebo groups describing results from a clinical trial, not a statistical test.
the proportion of subjects having that score or a better It is most sensibly used to describe an effect that has been score. For CAARS change scores, therapeutic change is demonstrated with appropriate statistical tools. Neverthe- indicated by negative numbers, i.e., a decrease in the less, the drug-placebo response curve's roots in (ROC) symptom score; 3.) For each observed score, plot these analysis motivate the computation of one statistic, the proportions for the drug group on the vertical axis against AUC, which is computed through integration. The area the proportions computed for the placebo group on the under the drug-placebo response curve ranges from 0.5 horizontal axis; 4.) Connect the plotted points and label (when the drug effect equals the placebo effect) to 1.0 those that correspond to the best response, the 25th per- (when the drug is completely effective and the placebo centile of response, the median response, the 75th percen- has no effect). The AUC is a useful index of clinical signif- tile of response and the worst response; 5.) If the outcome icance because it equals the probability that a randomly (page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CAARS Self-Rated ADHD Index.
selected member of the drug group will have a better Joseph Biederman, MD.- Joseph Biederman receives
result than a randomly selected member of the placebo research support from the following sources: Shire Laborato- group [,i.e., the probability that drug will outper- ries, Inc and Eli Lilly & Company, Pfizer Pharmaceutical, form placebo.
Cephalon Pharmaceutical,, Janssen Pharaceutical, Neuro-search. Pharmaceuticals, Stanley Medical Institute, Lilly In summary, the placebo-response curve provides four Foundation, Prechter Foundation, NIMH, NICHD and pieces of clinically relevant data not typically available from traditional statistical analyses of outcomes data.
First, the effect size of a drug on an outcome measure can Dr. Joseph Biederman is a speaker for the following speaker's be determined as the distance between the curve and the bureaus: Eli Lilly & Company, Pfizer Pharmaceutical, line of no effect at any given cut-point. Second, the ratio Novartis Pharmaceutical, Wyeth Ayerst, Shire Laborato- of drug responders to placebo responders across the range ries Inc, McNeil Pharmaceutical, and Cephalon of outcomes can be determined as the area under the curve. Third, the likelihood of a drug to elicit a specificoutcome (e.g., a clinically meaningful cut-point) can be Dr. Joseph Biederman is on the advisory board for the following determined as the proportion of drug-responders to pla- pharmaceutical companies: Eli Lilly & Company, CellTech, cebo-responders at any given cut-point. Fourth, the ability Shire Laboratories Inc, Novartis Pharmaceutical, Noven of a drug to improve functioning vs. prevent worsening of Pharmaceutical, McNeil Pharmaceuticals, Janssen, John- functioning can be determined as the proportion of drug- son & Johnson, Pfizer, and Cephalon Pharmaceuticals responders to placebo-responders at the outcome scorerepresenting no change.
Thomas Spencer, MD- Dr. Thomas Spencer receives research
support from the following sources: Shire Laboratories, Inc
and Eli Lilly & Company, Glaxo-Smith Kline, Pfizer Phar- Stephen V. Faraone, PhD.- Stephen Faraone receives
maceutical, McNeil Pharmaceutical, Novartis Pharmaceu- research funding from Lilly, McNeil and Shire.
(page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CGI ADHD Severity (Change Scores) Figure 9
CGI ADHD Severity (Change Scores).
Dr. Thomas Spencer is a speaker for the following speaker's Stephen J Glatt, PhD- Stephen Glatt has no conflicts of
bureaus: Glaxo-Smith Kline, Eli Lilly & Company, Novartis interest to declare.
Pharmaceutical, Wyeth Ayerst, Shire Laboratories Inc,McNeil Pharmaceutical Authors' contributions
Stephen V. Faraone, PhD- Steve Faraone contributed to
Dr. Thomas Spencer is on the advisory board for the follow- the analysis and interpretation of the data, the drafting ing pharmaceutical companies: Shire Laboratories, Inc and revision of the manuscript.
and Eli Lilly & Company, Glaxo-Smith Kline, Pfizer Phar-maceutical, McNeil Pharmaceutical, and Novartis Joseph Biederman, MD- Joseph Biederman contributed
to the analysis and interpretation of the data, the draftingand revision of the manuscript.
David Michelson, MD- David Michelson is a Lilly
employee
Thomas Spencer, MD- Thomas Spencer contributed to
the conception and design of the study, the acquisition of
Lenard Adler, MD- Lenard Adler receives grant and
data, interpretation of data and drafting/reviewing the Research Support, is a Consultant or on Advisory Boards: Abbott Laboratories, Bristol-Myers Squibb, Eli Lilly andCo., McNeil/Johnson & Johnson, Merck & Co., Inc., Neu- Lenard Adler, MD- Lenard Adler contributed to the con-
rosearch, Novartis Pharmaceuticals Corp., Pfizer Labs, ception and design of the study, the acquisition of data, Cortex Pharmaceuticals, Cephalon and Shire interpretation of data and drafting/reviewing the Fred Reimherr, MD-Fred Reimherr has been part of Lilly
David Michelson, MD- David Michelson contributed to
advisory board.
the study conception design, the data acquisition as wellas critical reviewing of the manuscript.
(page number not for citation purposes) Behavioral and Brain Functions 2005, 1:16
CGI ADHD Severity (En Figure 10
CGI ADHD Severity (Endpoint Scores).
Fred Reimherr, MD- Fred Reimherr contributed to the
Journal of Clinical Psychopharmacology 2000, study conception design, the data acquisition as well as McNeil BJ, Hanley JA: Statistical approaches to the analysis of
critical reviewing of the manuscript.
receiver operating characteristic (ROC) curves. Med Dec
Making 1984, 4:137-150.
Stephen J Glatt, PhD- Stephen Glatt contributed to the
JAMA 1982, 248:548-549.
analysis and interpretation of the data, and the drafting Siegel B, Vukicevic J, Spitzer RL: and revision of the manuscript.
Journal of Psychiatric
Research 1990, 24:293-311.
Siegel B, Vukicevic J, Elliott GR, Kraemer HC: This work was supported in part by a grant from Eli Lilly (Dr. Faraone, PI) J Am Acad Child Adolesc Psychiatry 1989, 28:542-548.
Faraone SV, Pliszka SR, Olvera RL, Skolnik R, Biederman J: Efficacy
of adderall and methylphenidate in attention-deficit/hyper-
Fuller R, Wong D: Effects of antidepressants on uptake and
activity disorder: a drug-placebo and drug-drug response
receptor systems in the brain. Prog Neuro-Psychopharmacol & Biol
curve analysis. Journal of Child and Adolescent Psychopharmacology
Psychiat 1985, 9:485-490.
Spencer T, Biederman J, Wilens T, Prince J, Hatch M, Jones J, Harding Faraone SV, Short E, Biederman J, Findling RL, Roe CM, Manos M M, Faraone SV, Seidman L: American Journal of Psychiatry 1998, 155:693-695.
International Michelson D, Adler L, Spencer T, Reimherr FW, West SA, Allen AJ, Journal of Neuropsychopharmacology 2002, 5:121-129.
Kelsey D, Wernicke J, Dietrich A, Milton D: Colditz GA, Miller JN, Mosteller F: Biol Psychiatry 2003, 53:112-120.
Int J Technol Assess Health Care 1988, 4:637-642.
Wernicke JF, Kratochivil C: Hanley JA, McNeil BJ Journal of Radiology Clinical Psychiatry 2002, 63:50-55.
Faraone SV, Biederman J, Spencer TJ, Wilens TE: (page number not for citation purposes)

Source: http://repository.countway.harvard.edu/xmlui/bitstream/handle/10473/2370/1262689.pdf?sequence=1