Use of antipsychotic medications in pediatric and young adult populations: future research needs.

Use of Antipsychotic Medications in Pediatric and Young Adult Populations: Future Research Needs

The use of antipsychotics, particularly second generation antipsychotics, among children and adolescents has increased markedly during the past 20 years. Existing evidence gaps make this practice controversial and hinder treatment decision-making. This article describes and prioritizes future research needs regarding antipsychotic treatment in youth, focusing on within-class and between-class drug comparisons with regard to key population subgroups, efficacy and effectiveness outcomes, and adverse event outcomes. Using as a foundation a recent systematic review of antipsychotic treatment among youth, which was completed by a different Evidence-based Practice Center, we worked with a diverse group of 12 stakeholders representing researchers, funders, health care providers, patients, and families to identify and prioritize research needs. From an initial list of 16 evidence gaps, we enumerated 6 high-priority research needs: 1) long-term comparative effectiveness across all psychiatric disorders; 2) comparative long-term risks of adverse outcomes; 3) short-term risks of adverse events; 4) differentials of efficacy, effectiveness, and safety for population subgroups; 5) comparative effectiveness among those with attention-deficit/hyperactivity disorder and disruptive behavior disorders and common comorbidities; 6) comparative effectiveness among those with bipolar disorder and common comorbidities. In this article, we describe these future research needs in detail and discuss study designs that could be used to address them. (Journal of Psychiatric Practice 2015;21:26–36) KEY WORDS: antipsychotic medications, pediatric, future research needs, treatment, effectiveness, attention deficit/hyperactivity disorder, bipolar disorder, disruptive behavior disorders

26

January 2015

ROBERT B. CHRISTIAN, MD BRADLEY N. GAYNES, MD, MPH LISSETTE M. SAAVEDRA, PhD BRIAN SHEITMAN, MD ROBERTA WINES, MPH DANIEL E. JONAS, MD, MPH MEERA VISWANATHAN, PhD ALAN R. ELLIS, PhD, MSW CAROL WOODELL, BSPH TIMOTHY S. CAREY, MD, MPH

First-generation antipsychotic medications (FGAs), also known as typical antipsychotics, were developed in the 1950s. FGAs are used to treat psychotic symptoms such as auditory and visual hallucinations and delusions through several proposed mechanisms, including the blockade of dopamine neuroreceptors. FGAs are associated with various adverse effects, including extrapyramidal symptoms (EPS) and the rare neuroleptic malignant syndrome (NMS). Second-generation antipsychotic medications (SGAs), also known as atypical antipsychotics, emerged in the 1980s. SGAs are generally thought to have a lower risk of EPS than FGAs.1,2 However, based on data from shorter term clinical trials, SGAs CHRISTIAN: University of North Carolina School of Medicine and Carolina Institute for Developmental Disabilities, Chapel Hill, NC; GAYNES, JONAS, and CAREY: University of North Carolina School of Medicine and Cecil G. Sheps Center for Health Services Research, RTI-UNC Evidence-based Practice Center, Chapel Hill, NC; SAAVEDRA, VISWANATHAN, and WOODELL: Research Triangle Institute International, RTI-UNC Evidence-based Practice Center, Research Triangle Park, NC; SHEITMAN: University of North Carolina School of Medicine; WINES and ELLIS: Cecil G. Sheps Center for Health Services Research, RTIUNC Evidence-based Practice Center Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. Please send correspondence to: Robert B. Christian, MD, Carolina Institute for Developmental Disabilities, University of North Carolina School of Medicine, CB #7255, Chapel Hill, NC, 27599. [email protected] The authors declare no conflicts of interest. Alan Ellis has received grant funding from Merck for unrelated research and from the UNC Center for Pharmacoepidemiology, which receives industry funding, for research on the use of psychotropic medications in children. Funding for this project was provided by the Agency for Healthcare Research and Quality (AHRQ) through the American Recovery and Reinvestment Act (ARRA) (Contract No. 290 2007 10056 I). The views expressed in this manuscript do not represent and should not be construed to represent a determination or policy of AHRQ or the U.S. Department of Health and Human Services. DOI: 10.1097/01.pra.0000460619.10429.4c

Journal of Psychiatric Practice Vol. 21, No. 1

Copyright © Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

USE OF ANTIPSYCHOTIC MEDICATIONS IN PEDIATRIC AND YOUNG ADULT POPULATIONS

have been found to be associated with a higher risk of a range of metabolic side effects, including weight gain, dyslipidemia, insulin resistance, the development of type 2 diabetes, and, rarely, hyperglycemic coma.1 The use of antipsychotics, particularly SGAs, in children and adolescents has increased markedly during the past 20 years.3–7 Prescribing antipsychotics to the pediatric population is controversial because of a relative lack of high-quality and longitudinal studies on which to base clinical practice recommendations. For the majority of antipsychotic drugs, approved indications for youth in the United States are restricted to the treatment of schizophrenia and bipolar disorder. The U.S. Food and Drug Administration (FDA) also approved risperidone in 2006 and aripiprazole in 2009 for the treatment of irritability associated with autism. However, offlabel prescriptions are given to younger children for a range of indications, including behavioral symptoms (eg, aggression) that are related to diagnosable conditions (eg, attention-deficit/hyperactivity disorder [ADHD]). Prompted by the recent increase in antipsychotic use among children, the Agency for Healthcare Research and Quality (AHRQ) worked with the University of Alberta Evidence-based Practice Center (EPC) in 2010 to complete a review titled “Comparative Effectiveness of First and Second Generation Antipsychotics in Pediatric and Young Adult Populations.”8 Tables 1–3 describe the first 3 key questions the review addressed and summarize the conclusions of the review. Key question 4 was “Do the effectiveness and risks of FGAs and SGAs vary in differing subpopulations (eg, sex; age group; race; comorbidities, including substance abuse and ADHD; cotreatment versus monotherapy; first episode psychosis versus a history of previous psychotic episodes; duration of illness; treatment naïve versus history of previous antipsychotic use)?” For this question, the review concluded that no subpopulation existed for which there was sufficient evidence for direct or indirect between-class or within-class comparisons of antipsychotics across disorders. The AHRQ review also included the authors’ recommendations for future research for each key question.8 The purpose of our work was to build on the AHRQ review by defining and prioritizing future research needs (FRNs) in a comprehensive and sys-


tematic way. Using strategies developed by the Research Triangle Institute-University of North Carolina (RTI-UNC) EPC and other EPCs,9–11 we gathered and synthesized the opinions of an informed and involved stakeholder group, asking them to help identify evidence gaps in the systematic review and to prioritize specific future research needs. The intent of the FRN process is to create a document to help guide a range of stakeholders in developing research priorities. In addition, this FRN project focuses on effectiveness rather than efficacy and by its nature intends to influence research that will ultimately help to provide guidance in realworld clinical practice.

METHODS Because the AHRQ review was produced by another EPC and the final report had not yet been published when our study began, we used the draft review to compile an initial list of evidence gaps: topics for which the review authors rated the strength of evidence as “low” or “insufficient” to support conclusions.12 Once the final review was published, we verified that our list of evidence gaps was consistent with the limitations described in the final version. We then scanned clinicaltrials.gov, HSRProj, NIH RePORTER, and the International Clinical Trials Registry Project to find relevant ongoing or recently completed research that might address any of the gaps. A description of the search strategy and results is available in the full report which is available online.13 We convened a 12-member national stakeholder panel, drawn from advocacy groups, researchers, providers of care, funders, and professional organizations. Our stakeholder group (listed in the Acknowledgments, p. 35) included 8 non-federal participants and 4 representatives from federal agencies with vested interest in medication or mental health research and policy. Each potential stakeholder completed a disclosure form, which was reviewed by both the project team and AHRQ for potential conflicts of interest. Had a substantial conflict of interest been discovered, the stakeholder involved would have been eliminated from consideration and another stakeholder would have been approached as a replacement. A complete description of our methods is provided in the full report.13 We engaged stakeholders through two teleconferences, electronic mail,

January 2015


27


Table 1. Summary of Key Question 1, level of evidence and conclusions8 Key Question 1: What is the comparative efficacy or effectiveness of FGAs and SGAs for treating disorder-specific and nonspecific symptoms? Disorders/ symptoms

Strength of evidence

Conclusions

Pervasive developmenttal disorders

Insufficient to low

x Insufficient evidence for comparison among specific SGAs and insufficient evidence for comparison of SGAs versus FGAs x SGAs favored over placebo for behavioral symptoms associated with PDDs with low SOE

ADHD and disruptive behavior disorders

Insufficient to moderate

x Insufficient evidence for comparison among specific SGAs and insufficient evidence for comparison of SGAs versus FGAs x Insufficient evidence concerning clinical effectiveness outcomes of interest (eg, social/occupational outcomes, health-related quality of life, and medication adherence) x No difference between SGAs and placebo for aggression or anxiety (low SOE) x Among those with ADHD and DBDs, SGAs superior to placebo for a number of behavior symptoms and clinical global impressions (moderate SOE)

Bipolar disorder


x Insufficient evidence for comparison among specific SGAs and insufficient evidence for comparison of SGAs versus FGAs x Insufficient evidence concerning clinical effectiveness outcomes of interest (eg, social/occupational outcomes, health-related quality of life, and medication adherence) x SGAs favored over placebo for treatment of mania (low SOE) x SGAs did not differ from placebo for depression symptoms in this group (low SOE) x SGAs favored over placebo for clinical global impressions (moderate SOE)

Schizophrenia and other psychotic disorders


x Insufficient evidence for comparison among most specific SGAs and insufficient evidence for comparison of SGAs versus FGAs x SGAs favored over FGAs for clinical global impressions (low SOE) x No difference between SGAs and FGAs on the PANSS x Several within-class SGA comparisons did not differ on positive and negative symptoms or clinical global impressions (low SOE) x SGAs favored over placebo for tics (moderate SOE)

Tourette syndrome


x Insufficient evidence for comparison among specific SGAs and insufficient evidence for comparison of SGAs versus FGAs

Obsessivecompulsive disorder


x Insufficient evidence concerning efficacy or effectiveness of specific medications or SGAs versus FGAs for all outcomes of interest

Posttraumatic stress disorder


x Insufficient evidence concerning efficacy or effectiveness of specific medications or SGAs versus FGAs for all outcomes of interest continued

28

January 2015




Table 1. Summary of Key Question 1, continued Disorders/ symptoms


Conclusions

Anorexia nervosa


x Insufficient evidence concerning efficacy or effectiveness of specific medications or SGAs versus FGAs for all outcomes of interest

Behavioral symptoms*

Insufficient

x Insufficient evidence concerning efficacy or effectiveness of specific medications or SGAs versus FGAs for all outcomes of interest from studies evaluating non-disorder specific treatment of behavioral symptoms.

ADHD: attention-deficit/hyperactivity disorder; DBD: disruptive behavior disorder; FGA: first-generation antipsychotic; PANSS = Positive and Negative Syndrome Scale; PDD: pervasive developmental disorder; SGA: second-generation antipsychotic; SOE: strength of evidence. *Behavioral symptoms = aggression, agitation, anxiety, behavioral dyscontrol, irritability, mood lability, selfinjurious behaviors, and sleep disorders

and two web-based prioritization exercises. During the first teleconference, we asked the stakeholders to refine the initial list of evidence gaps based on their expertise. After that call, we revised the list of gaps and asked each stakeholder to complete an online prioritization exercise using the following elements from the AHRQ Effective Health Care Program topic selection criteria: importance, desirability of new research/duplication, and potential impact.14 The online exercise used a forced prioritization method in which each person could distribute a limited number of votes among the evidence gaps. We reviewed the results of this exercise with the stakeholders on the second conference call. Then, in a second round of the same online exercise, the stakeholders assigned priorities among the evidence gaps that had been ranked highest in the first round. We used the results of the second round to determine the final list of high-priority evidence gaps. The study team considered the following components for each of the high-priority evidence gaps— population, intervention, comparator, outcome, timeframe, and setting (often referred to as PICOTS)—and developed a list of FRNs.15 We considered potential study designs for each research need using specific criteria, including ability to produce a valid result, resource requirements, availability of data or ability to recruit subjects, and ethical and legal considerations,10 and performed sample power analyses to help identify pragmatic barriers involved in the potential designs. Study design options were generated and refined by the multidisciplinary study team, which included methodologists, analysts, and


mental health services researchers. Several future research needs could be addressed by more than one study design. We did not ask stakeholders to propose or rank study designs. We then conducted power analyses based on hypothetical study designs. We focused on comparisons of means in 2 x 2 factorial designs because this scenario was appropriate for several of the high-priority research needs. The objective was to determine the approximate numbers of patients and measurement occasions needed to have a power of 0.80 to detect a clinically meaningful difference in outcomes with a two-tailed test at a significance level of 0.05. The power analysis results are reflected in the study designs suggested below; details appear in the full report.13

RESULTS Prioritization Results From an initial list of 16 evidence gaps, we completed two rounds of prioritization13 and determined a final prioritized list of 6 FRNs (Table 4). Each research need is expressed as an actionable research question. All of these FRNs are considered “high-priority” since they are a subset of the initial list of gaps. Given the extensive overlap of PICOTS across these 6 research needs, we present in Table 5 an expanded, integrated PICOTS table. These outcomes are organized into disorder-specific efficacy/effectiveness outcomes, outcomes related to psychiatric comorbidity, and adverse event outcomes.

January 2015


29


Table 2. Summary of Key Question 2, level of evidence and conclusions8 Key Question 2: Do FGAs and SGAs differ in the following medication-associated adverse events: overall adverse events, specific adverse events, withdrawals and time to withdrawal due to adverse events, and persistence and reversibility of adverse events? Adverse event


Conclusions

Extrapyramidal symptoms

Insufficient to low

x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons except as otherwise stated below* x SGAs significantly favored over haloperidol for EPS (low SOE)

Weight gain/ body composition


x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons except as otherwise stated below* x Quetiapine and risperidone favored over olanzapine for body composition (moderate SOE) x Haloperidol favored over olanzapine for body composition (low SOE)

Hyperglycemia/ insulin resistance

Insufficient

x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons except as otherwise stated below*

Dyslipidemia


x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons except as otherwise stated below* x Risperidone favored over olanzapine for dyslipidemia (moderate SOE)

Hyperprolactinemia /sexual development

Insufficient

Sedation

Insufficient

x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons except as otherwise stated below* x Olanzapine favored over risperidone for prolactin-related events (moderate SOE) x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons *

FGA: first-generation antipsychotic; SGA: second-generation antipsychotic; SOE: strength of evidence. *A number of within-class (SGA vs. SGA) comparisons revealed no significant difference with low SOE. For comparisons of SGAs vs. placebo for nearly all outcomes, the placebo group experienced significantly fewer adverse events than the group receiving SGAs (moderate SOE). Insufficient evidence for questions concerning withdrawals, time to withdrawal, or persistence/reversibility of adverse events.

Study Design Considerations The project team considered various study designs to address this set of prioritized research needs. Because these research needs share many PICOTS elements, their study design considerations also overlap. Accordingly, we review general study design considerations that apply to all of the research needs. Our considerations regarding randomized controlled trials, nonrandomized comparative designs, observational designs, and meta-analysis of individual patient data follow. A complete discussion

30

January 2015

of study design considerations can be found in the full report online.13 Randomized controlled trials. While an RCT may have many ideal design features, in general RCTs will not be practical for evaluating these complex questions regarding drug and subgroup comparisons. For example, our sample size estimation analysis demonstrated that if one wanted to conduct a drug-drug comparison trial of response rates for mania in bipolar disorder, each group may require up to 388 subjects to detect clinically significant differ-




Table 3. Summary of Key Question 3, conclusions8 Key Question 3: Do FGAs and SGAs differ in the following other short (occurring within 6 months) and long-term (occurring after 6 months) outcomes? Outcome Response rate, time to discontinuation, medication adherence/ persistence


Conclusions

Insufficient to low

x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons across disorders except as otherwise stated x Medication adherence not statistically or clinically different between SGAs and placebo for PDDs and ADHD/DBDs (low SOE) x Medication adherence statistically significantly better for placebo than for SGAs in bipolar disorder (low SOE) x No significant difference in medication adherence between FGAs and SGAs, olanzapine and quetiapine, olanzapine and risperidone, or SGAs and placebo in schizophrenia (low SOE)

Growth and maturation

Insufficient

x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons across disorders

Cognitive and emotional development

Insufficient


Suicide-related behaviors


x Insufficient evidence for direct or indirect between-class or within-class antipsychotic comparisons across disorders except as otherwise stated x SGAs and placebo did not significantly differ for suicide-related behaviors in bipolar disorder (moderate SOE)

School performance

Insufficient


Work-related functioning

Insufficient


Patient insight into illness

Insufficient


Patient/parent reported outcomes

Insufficient


Health-related quality of life

Insufficient


ADHD: attention-deficit/hyperactivity disorder; DBD: disruptive behavior disorder; SOE: strength of evidence; FGA: first-generation antipsychotic; PDD: pervasive developmental disorder; SGA: second-generation antipsychotic

ences between drugs. This subgroup size is as high as the total population in the largest trials in the field to date. This estimate was generated based on data from existing drug-versus-placebo trials using conservative assumptions. For studies of rare outcomes (eg, suicide) or differences that are difficult to detect


(eg, in cognitive performance), the needed sample sizes are likely unachievable. The full report provides further details on sample size estimations.13 Additional factors may make randomized trials difficult in this research area. The parents of child research subjects may be unwilling to accept random

January 2015


31


Table 4. Final highest priority research needs based on stakeholder input 1. What is the long-term comparative effectiveness between and within classes of antipsychotics as measured in outcomes related to the disorder of interest, its comorbidities, associated behavioral features, social-occupational outcomes, and outcomes identified as important by patients and their families? 2. What are the comparative long-term risks of medication exposure between and within antipsychotic classes? 3. What is the efficacy and effectiveness of first- or second-generation antipsychotics for adolescents and young adults with schizophrenia in the following outcome domains: core features of the disorder, commonly associated comorbidities and behavioral features, social/occupational functioning, patient- and parent-reported outcomes, outcomes related to high-risk behaviors, and suicide-related behavior? 4. Are there subgroups of patients, based on baseline demographic/clinical characteristics or physical and/or mental health comorbidities, for which first- and second-generation antipsychotics differ in efficacy, effectiveness, or frequency of adverse events? Sub-groups include sex, age, race/ethnicity, comorbidities, co-treatment, history of psychosis, history of treatment failure, or duration of illness. 5. What is the efficacy and effectiveness of first- or second-generation antipsychotics for individuals with attentiondeficit/hyperactivity disorder (ADHD) and disruptive behavior disorders in the following outcome domains: core ADHD symptoms, commonly associated comorbidities and behavioral features, social/occupational functioning, patient- and parent-reported outcomes, outcomes related to high-risk behaviors, and suicide-related behavior? 6. What is the efficacy and effectiveness of first- or second-generation antipsychotics for adolescents and young adults with bipolar disorder in the following outcome domains: core features of the disorder, commonly associated comorbidities and behavioral features, social/occupational functioning, patient- and parent-reported outcomes, outcomes related to high-risk behaviors, and suicide-related behavior?

assignment to treatment. In addition, when there is a perceived or real lack of equipoise (eg, in comparing FGAs versus SGAs), random assignment may be neither ethical nor acceptable. Nonrandomized comparative designs. Non-randomized designs may allow researchers to address the challenges of examining long-term outcomes in the youth population when randomization at the level of the individual is not feasible or ethical. Instead of random assignment, treatment may be determined by the provider or practice. Selection bias and unmeasured confounding are issues of greater concern with these designs. If the sample size is large, this design may be the best method to assess subgroup effects or incidence of rare harms. However, available statistical techniques for adjusting for baseline differences may not completely control for potential bias and unmeasured confounders. Observational studies. Prospective and retrospective cohort designs are also useful for addressing the

32

January 2015

complex comparative effectiveness questions presented here. Although similar to nonrandomized comparative designs (trials in which treatment is not assigned at random), the intervention is not controlled. Registries of individuals exposed to antipsychotics, with linkages to clinical data from claims or medical records, will be important but are only beginning to be developed in the United States and Canada. Practice-based research networks in child mental health may also be vehicles for evaluating needs, but such platforms are nascent in this field. Pooled analysis of individual patient data. Another appropriate design would involve pooling data from existing trials for simultaneous data analysis. Advantages of these study designs for producing a valid result include increased statistical power, increased ability to account for heterogeneity in the sample, and increased frequency of low-baserate behaviors/harms. As with non-randomized comparative designs, careful consideration of selection bias and unmeasured confounders is key. Integration




Table 5. Populations, interventions, comparators, outcomes, time frames, and settings (PICOTS) for the upper-tier evidence gaps and related research questions prioritized in round 2

The following descriptions of PICOTS define the scope of the evidence gaps and research questions identified through the Future Research Needs for the Comparative Effectiveness of First- and SecondGeneration Antipsychotics stakeholder engagement process. Specific outcomes may apply to one or more research questions. For example, some questions focus on health and behavioral outcomes and others focus on adverse events. Population All Questions: Children, youth, and young adults (24 years of age or younger) with one or more of the following disorders: PDD, ADHD, DBD, bipolar disorder, schizophrenia/schizophrenia-related psychosis, OCD, PTSD, eating disorders (anorexia nervosa, bulimia nervosa, eating disorder NOS), tic disorders (Tourette syndrome), or non-disorder specific severe behavioral issues (eg, aggression). For Subgroups: Demographic and clinical subgroups of children, youth, and young adults 24 years of age or younger with one or more of the disorders listed above including: age, racial groups, gender, genetically defined subgroups, socioeconomic groups, level of education, or physical and mental health comorbidities (other psychiatric disorders, treatment history, substance abuse, learning disabilities, developmental disorders, language impairments, cognitive abilities). Interventions All Questions: Any FDA-approved FGA or SGA Comparators All Questions: Any other FDA-approved FGA or SGA, placebo, or another dose of the same antipsychotic Outcomes (efficacy or effectiveness) Disorder/illness-specific symptom-related outcomes: The following are examples of symptom domains of interest for each disorder or illness and do


not represent an exhaustive list of all potential outcomes that might be measured: • PDDs: repetitive behaviors, social functioning/communication, agitation/aggression • ADHD and DBDs: impulsivity, defiant behavior, aggressive behavior • Bipolar disorder: mania, depression • Schizophrenia and related psychoses: positive and negative symptoms, cognitive function • OCD: severity of obsessions, severity of compulsions • PTSD: re-experiencing, avoidance behaviors • Anorexia nervosa/bulimia nervosa/eating disorder NOS: metabolic/nutritional outcomes, purging frequency/ intensity • Tourette syndrome: tic severity Non-disorder specific behaviors of interest: The following are examples of key common non-disorderspecific behaviors of interest that could be considered when evaluating efficacy/effectiveness: aggression, agitation, anxiety, behavioral dyscontrol, irritability, mood lability, self-injurious behaviors, and sleep disorders. Outcomes (related to key psychiatric comorbidities) The following are examples of key comorbidities for the specific disorders of interest listed above that could also be considered when evaluating efficacy/ effectiveness: • PDDs: ADHD, anxiety disorders • ADHD and DBDs: substance abuse, depression, anxiety disorders • Bipolar disorder: ADHD, substance abuse or dependence • Schizophrenia and related psychoses: depression, substance abuse or dependence, • OCD: Tourette syndrome/tic severity, ADHD • PTSD: depression, substance abuse • Anorexia nervosa/bulimia nervosa/eating disorder NOS: depression • Tourette syndrome: ADHD, OCD

January 2015


33


Table 5. PICOTS continued Outcomes (harms or adverse events) Medication-associated adverse events, withdrawal due to adverse events, persistence and reversibility of adverse events. Examples of major adverse events include the following: • Mortality • Cerebrovascular disease-related events • Development of diabetes mellitus • Diabetic ketoacidosis • Neuroleptic malignant syndrome • Seizures • Tardive dyskinesia • Cardiomyopathies • Cardiac arrhythmias • Agranulocytosis • Extrapyramidal effects General adverse events. Examples include the following: • Weight gain (eg, using body mass index growth charts) • Agitation • Constipation • Sedation • Elevated cholesterol • Elevated transaminases • Adverse events related to prolactin elevation • Galactorrhea/bloody galactorrhea • Exercise intolerance • Precocious puberty Outcomes (other short-term and long-term outcomes) • Response rates with corresponding dose, duration of response, remission, relapse, speed of response, time to discontinuation of medication • Growth and maturation • Cognitive and emotional development • Suicide-related behaviors or death by suicide • Medication adherence and persistence

of trials with multiple arms would allow testing of several hypotheses regarding agents. However, use of this study design requires underlying databases with common data definitions and the willingness of investigators to pool data.

34

January 2015

• • • •

•

• • • • •

School performance/attendance Work-related functional capacity Patient insight into illness High-risk behavior outcomes (ie, unintended pregnancy, accidental injury/death, sexually transmitted disease, substance abuse) Outcomes reported by patient or parent/care provider, including level of physical activity/inactivity, diet (ie, caloric intake, food preferences) Health-related quality of life Relationship functioning Interactions with legal/justice system (ie, arrests, detention) Utilization of health care system (eg, hospitalization rates, medication costs, outpatient expenditures) “Outcomes that matter” to children, youth, young adults, and their families. Examples of such functional outcomes could include those related to social success, development of autonomy, or capacity for spirituality. These outcomes are often tied to developmental level.

Timing • Short term: < 6 months • Long term: > 6 months, generally 1 or more years Settings All settings, including inpatient hospitalization and outpatient treatment, and practice-based research networks ADHD: attention-deficit/hyperactivity disorder DBD: disruptive behavior disorder FDA: U.S. Food and Drug Administration FGA: first-generation antipsychotic NOS: not otherwise specified OCD: obsessive-compulsive disorder PDD: pervasive developmental disorder PTSD: posttraumatic stress disorder SGA: second-generation antipsychotic

CONCLUSIONS Working with a group of stakeholders, we ultimately identified 6 high-priority research needs in the area of antipsychotic usage in youth. The 6 high-priority




research needs (Table 4) covered a broad range of issues cutting across disorders, key clinical outcomes, safety outcomes, and methodological concerns. The prioritization revealed an emphasis on safety and effectiveness needs across all mental health conditions that occur in children and adolescents and also highlighted the desire for more comparative effectiveness research in ADHD, bipolar disorder, and schizophrenia. PICOTS development aided our consideration of study design issues, and our sample power analyses demonstrated the clear pragmatic barriers that many of the potential designs present. While large long-term multi-site clinical trials may be the gold standard for addressing many of the important questions, feasibility concerns have greatly limited the number of such large pragmatic trials in mental health to date. It may be viable to conduct large prospective cohort studies of youth exposed to antipsychotics and such studies offer considerable analytic flexibility, but they are also costly. Patient registries with linkages to clinical data sets may allow for more efficient evaluation of some questions with advanced analysis methods, but the infrastructure for registry studies requires considerable investment, and its development may face considerable hurdles related to information privacy.10 Meta-analysis of current trial data and meta-analysis of individual patient data may prove helpful, but will likely be limited to evaluation of specific shorter term outcomes.10 The nature of the stakeholder process comes with certain limitations. First, the process requires a delicate balance between the need to create a list of clear, concise gaps and the need to remain faithful to the language and intent of the findings of the original AHRQ review, which was completed by a different EPC.8 It is possible that other important research gaps were not identified in the original comparative effectiveness review, and hence were not evaluated in this project. However, this balance ultimately may represent a strength of the process in that neither the findings of the systematic review nor the views of the stakeholder panel fully dominated the final product. In addition, this future research needs process was challenging in that it focused on gaps that cut across disorders as well as gaps concentrated on specific groups of disorders. The end result was a top tier needs list that included 3 specific disorder groups. This may create the impression that the panel did not see other particu-


lar disorder groups (eg, pervasive developmental disorders) as a higher order need and this may indeed be a limitation of the process. Fortunately, the process emphasized a range of effectiveness and safety-related future research needs across all disorders in addition to highlighting specific disorder groups. Second, participation by stakeholders in each exercise was not complete, which might unfairly favor a research need preferred by those participating at a particular time. However, we had at least 75% participation in each call and prioritization exercise. Furthermore, the broad spectrum of the high-priority research needs suggests that no one view was over-represented. Despite its limitations, the structured process used in this project may prove to be an effective way of reaching agreement on research priorities in this broad and complex topic area.

Acknowledgements We wish to thank our Task Order Officer at the Agency for Healthcare Research and Quality (AHRQ) Effective Health Care Program, Sonia Tyutyulkova, MD, and the members of our stakeholder panel for their contributions of time and expertise. The stakeholder panel members were: Rebecca Bitsko, PhD, CDC National Center on Birth Defects and Development Disabilities, Atlanta, GA; Teri Brister, PhD, Programs for Young Families, Brandon, MS; Jana Davidson, MD, FRC, BC Mental Health & Addictions Research Institute, Vancouver, BC; Debra Dihoff, MA, National Alliance on Mental Illness, Raleigh, NC; Tiffany Farchione, MD, Food and Drug Administration, Division of Psychiatry Products, Silver Spring, MD; Laurence Greenhill, MD, New York State Psychiatric Institute, New York, NY; Fay Kagan, MD, Hathaway Children’s Services, Sylmar, CA; Penelope Knapp, MD, Department of Psychiatry and Behavioral Sciences, UC Davis M.I.N.D. Institute, Sacramento, CA; Laurel Leslie, MD, MPH, Tufts Clinical and Translational Science Institute, Boston, MA; Mark Olfson, MD, MPH, Columbia University Medical Center, New York, NY; Benedetto Vitiello, MD, National Institute of Mental Health, Bethesda, MD; Julie Zito, PhD, University of Maryland School of Pharmacy, Baltimore, MD. We also thank Jennifer Seida, MPH, and her colleagues at the University of Alberta EPC, authors of the 2010 AHRQ comparative effectiveness review.

January 2015


35

USE OF ANTIPSYCHOTIC MEDICATIONS IN PEDIATRIC AND YOUNG ADULT POPULATIONS References 1.

2.

3.

4.

5.

6.

7.

8.

9.

36

Leucht S, Corves C, Arbter D, et al. Second-generation versus first-generation antipsychotic drugs for schizophrenia: a meta-analysis. Lancet. 2009;373(9657):31–41. Haddad PM, Das A, Keyhani S, et al. Antipsychotic drugs and extrapyramidal side effects in first episode psychosis: a systematic review of head-head comparisons. J Psychopharmacol. 2012;26(5 Suppl):15–26. Zito JM, Safer DJ, DosReis S, et al. Trends in the prescribing of psychotropic medications to preschoolers. JAMA. 2000;238:1025–1030. Zito JM, Safer DJ, DosReis S, et al. Psychotropic practice patterns for youth: A 10 year perspective. Arch Pediatric Adolesc Med. 2003;157:17–25. Zito JM, Safer DJ, de Jong-van den Berg LT, et al. A threecountry comparison of psychotropic medication prevalence in youth. Child Adolesc Psychiatry Ment Health. 2008;121: 26. Zito JM, Safer DJ, Sai D, et al. Psychotropic medication patterns among youth in foster care. Pediatrics. 2008;121: e157–e163. Pathak S, Arszman SP, Danielyan A, et al. Psychotropic utilization and psychiatric presentation of hospitalized very young children. J Child Adolesc Psychopharmacol. 2004;14:433–442. Seida JC, Schouten JR, Mousavi SS, et al. First- and Second-Generation Antipsychotics for Children and Young Adults. Comparative Effectiveness Review No. 39. Rockville, MD: prepared by the University of Alberta Evidence-based Practice Center under Contract No. 2902007-10021; February 2012 (available at www.ncbi.nlm.nih.gov/books/NBK84643, accessed December 20, 2014). Gaynes B, Christian R, Saavedra L, et al. Future Research Needs for Attention Deficit Hyperactivity Disorder: Effectiveness of Treatment in At-Risk Preschoolers, LongTerm Effectiveness in All Ages, and Variability in Prevalence, Diagnosis, and Treatment. Future Research Needs No. 9. Rockville, MD: prepared by RTI-UNC Evidence-based Practice Center under Contract No. 290-

January 2015

10.

11.

12.

13.

14.

15.

2007-10056-I; May 2012 (available at www.ncbi.nlm .nih.gov/books/NBK83972, accessed December 20, 2014). Carey TS, Sanders GD, Viswanathan M, et al. Framework for Considering Study Designs for Future Research Needs. Methods Future Research Needs Paper No. 8. Rockville, MD: prepared by the RTI–UNC Evidence-based Practice Center under Contract No. 290-2007-10056-I; March 2012 (available at www.ncbi.nlm.nih.gov/books/NBK95273, accessed December 20, 2014). Carey TS, Crotty KA, Morrissey JP, et al. Future Research Needs for the Integration of Mental Health/Substance Abuse and Primary Care. Future Research Needs Paper No. 3. Rockville, MD: prepared by the RTI International– University of North Carolina at Chapel Hill Evidencebased Practice Center under Contract No. 290-200710056-I; September 2010 (available at www.ncbi.nlm .nih.gov/books/NBK51247, accessed December 20, 2014). Owens DK, Lohr KN, Atkins D, et al. Grading the Strength of a Body of Evidence When Comparing Medical Interventions. Methods Guide for Effectiveness and Comparative Effectiveness Reviews. Rockville, MD: Agency for Healthcare Research and Quality (US); 2008 (available at www.ncbi.nlm.nih.gov/books/NBK47091, accessed December 20, 2014). Christian R, Saavedra L, Gaynes B, et al. Future Research Needs for First- and Second-Generation Antipsychotics for Children and Young Adults. Future Research Needs Paper No. 13. Rockville, MD: prepared by the RTI-UNC Evidence-based Practice Center under Contract No. 290 2007 10056 I; February 2012 (available at www.ncbi.nlm .nih.gov/books/NBK84660, accessed December 20, 2014). Agency for Healthcare Research and Quality Effective Health Care Program. How Are Research Topics Chosen? (available at www.effectivehealthcare.ahrq.gov/index.cfm /submit-a-suggestion-for-research/how-are-research-topics-chosen, accessed January 26, 2012). Whitlock EP, Lopez SA, Chang S, et al. AHRQ series paper 3: Identifying, selecting, and refining topics for comparative effectiveness systematic reviews: AHRQ and the Effective Health-Care Program. J Clin Epidemiol. 2010; 63:491–501.



Use of antipsychotic medications in pediatric populations: what do the data say?

Medicaid prior authorization policies for pediatric use of antipsychotic medications.

Differences in Antipsychotic-Related Adverse Events in Adult, Pediatric, and Geriatric Populations.

Future directions and research needs.

Future directions in antipsychotic drug research.

Research needs of pediatric hospitalists.

Recommendations for Cancer Epidemiologic Research in Understudied Populations and Implications for Future Needs.

Antipsychotic medications for schizophrenia.

Antipsychotic medications for schizophrenia.

Research on drug use: a review of problems, needs and future perspectives.

Antipsychotic medications for schizophrenia--reply.

Treatment of Young People With Antipsychotic Medications in the United States.

Treatment of molluscum contagiosum in adult, pediatric, and immunodeficient populations.

Nonmedical use of prescription medications in young adults.

Venous Thromboembolism Risk Assessment and Prophylaxis Use in Pediatric, Adolescent, and Young Adult Hematology Oncology Patients.

Herbicide-resistant weeds: from research and knowledge to future needs.

Stakeholder assessment of comparative effectiveness research needs for Medicaid populations.

Survivorship of hip and knee implants in pediatric and young adult populations: analysis of registry and published data.

Pharmacy switch of antipsychotic medications: patient's perspective.

Variation in use of antipsychotic medications in nursing homes in the United States: A systematic review.

Assessing pediatric and young adult substance use through analysis of prehospital data.

Cocaine use in individuals with schizophrenia: impact on doses of discharge antipsychotic medications.

The Formation, Elimination, Interpretation, and Future Research Needs of Phosphatidylethanol for Research Studies and Clinical Practice.

Forecasting Medicaid Expenditures for Antipsychotic Medications.