OBJECTIVE. Primary care clinics have become the “de facto” mental health clinics for teens with mental health problems such as depression; however, there is little guidance for primary care professionals who are faced with treating this population. This study surveyed experts on key management issues regarding adolescent depression in primary care where empirical literature was scant or absent.
METHODS. Participants included experts from family medicine, pediatrics, nursing, psychology, and child psychiatry, identified through nonprobability sampling. The expert survey was developed on the basis of information from focus groups with patients, families, and professionals and from the research literature and included sections on early identification, assessment and diagnosis, initial management, treatment, and ongoing management. Means, standard deviations, and confidence intervals were calculated for each survey item.
RESULTS. Seventy-eight of 81 experts agreed to participate (return rate of 96%). Fifty-three percent of the experts (n = 40) were primary care professionals. Experts endorsed routine surveillance for youth at high risk for depression, as well as the use of standardized measures as diagnostic aids. For treatment, “active monitoring” was deemed appropriate in mild depression with recent onset. Medication and psychotherapy were considered acceptable options for treatment of moderate depression without complicating factors such as comorbid illness. Fluoxetine was rated as the most appropriate antidepressant for use in this population. Finally, experts agreed that patients who are started on antidepressants should be followed within 2 weeks after initiation.
CONCLUSIONS. Survey results support the identification and management of adolescent depression in the primary care setting and, in specific situations, referral and co-management with specialty mental health professionals. Even with the recent controversies around treatment, experts across primary care and specialty mental health alike agreed that active monitoring, pharmacotherapy with selective serotonin reuptake inhibitors, and psychotherapy can be appropriate under certain clinical circumstances when initiated within primary care settings.
Adolescent depression is a typically recurring disorder with serious morbidity and mortality.1,2 It occurs commonly, affecting between 3% and 7% of adolescents.2,3 Although the majority of primary care professionals (PCPs) believe that it is their responsibility to diagnose depression in their adolescent patients, many of these teens go unrecognized and untreated.4,5 In fact, research shows that only 50% of adolescents with depression receive a diagnosis before reaching adulthood.4
In the past few decades, much new information has come forth from the research literature regarding the management of depression in adolescents; however, few studies have examined several aspects of depression identification and management, including strategies for assessment and initial management after diagnosis, such as the use of mnemonics, standardized assessment tools, and frequency of initial follow-up. Furthermore, the translation of what has been learned through research efforts into clinical practice has been limited by the shortage of mental health professionals throughout the United States and Canada. This shortage increased the need for information and strategies on the management of depression in primary care (PC) settings, although few PCPs have adequate training in the identification and management of depression in adolescents.
PCPs' ability to manage adolescent depression is also limited by the lack of research conducted specifically in PC settings. Many issues pertaining to optimal management of adolescent depression in PC have not been addressed by research to date; neither are some critical questions likely to be answered by future research, in part because of feasibility challenges and lack of funding, to name just a few obstacles.6–8 Such issues include recommended strategies for early identification and frequency of follow-up.
To fill in these knowledge gaps, a steering committee of US and Canadian experts in depression and in PC developed a survey for administration to a larger group of PC and specialty mental health professionals to gather and develop consensus expert opinions regarding critical issues in the management of adolescent depression in the PC setting. This steering committee developed the survey on behalf of a larger guidelines development effort, known as the Guidelines for Adolescent Depression–Primary Care (GLAD-PC).9,10
The expert survey was developed on the basis of information from 2 sources: (1) focus groups with patients, families, and professionals and (2) the research literature. We conducted 16 focus groups (both for this project and for other associated projects) in Canada (n = 6) and the United States (n = 10) with PCPs (n = 90) and youth patients (n = 12) and their families (n = 38) to review issues that are pertinent to the management of depression in PC. The focus groups were conducted in PC clinics, including pediatric clinics and family medicine clinics, in Canada (Montreal, and Toronto) and the United States (New Jersey, Utah, and Florida). Findings from the focus groups were used to identify clinicians' views about critical issues and clinical concerns arising from the management of adolescent depression in PC and to guide the development of the survey. In-depth results from the focus groups will be available in a separate publication. Briefly, focus group results indicated that clinicians' questions, concerns, and need for guidance about adolescent depression management in PC fell into 5 areas: early identification, assessment/diagnosis, initial management, treatment, and ongoing management.
We then searched the literature to find what evidence was available to address these 5 problem areas where research data might provide clinical guidance. The areas reviewed included (1) major depressive disorder identification/assessment in PC settings; (2) efficacy of psychopharmacologic and psychotherapeutic treatments (with almost all data available only via clinical trials conducted in specialty mental health settings, rather than PC settings); (3) efficacy of brief counseling and psychosocial supportive interventions when offered in PC; and (4) efficacy and impact of ongoing clinical management and follow-up to include mental health referral, establishing other community linkages when needed, and organization of care.
In areas where the literature was silent but clinical guidance was nonetheless needed, we developed questions to ascertain expert opinion to fill in these evidence gaps. Where common sense dictated, no questions were developed. One such example is the merit of providing education about depression to patients and families.
The initial draft of the survey was reviewed by members of the GLAD-PC steering group, including experts in PC and mental health (see Acknowledgments), and modifications were made on the basis of their feedback. The final draft was edited and finalized by 2 expert PCPs (Drs Stein and Laraque) to ensure relevance and applicability for PCPs and that items encompassed the full range of sensible options for the PC setting. Additional review was done with a small group of PCPs to ensure clarity and to estimate the time required for completion. The final questionnaire was organized into several sections, early identification, assessment and diagnosis, initial management, treatment, and ongoing management, and consisted of 34 general question categories and 220 response options.
The survey, using a 9-point scale (1 = most inappropriate treatment and 9 = extremely appropriate treatment), is a modified survey format developed by the Rand Corporation for examining expert consensus.11 The large number of response options allowed us to gather expert opinion regarding depression management under a variety of clinical circumstances. For example, experts were asked to rate the appropriateness of specific strategies for identifying depression during different types of visits, such as well-child and urgent care. We were also able to seek expert opinion on the appropriateness of specific management strategies for different clinical presentations, varying by the presence or absence of a specific symptom, such as psychosis, in patients with depression. In addition, the 9-point scale provided detailed responses regarding the appropriateness of individual strategies and treatment choices in specific clinical scenarios. Finally, a small proportion of items in the survey asked for experts to estimate timelines (eg, how many weeks would you wait before starting treatment?). A copy of the survey is available on request.
Survey participants were chosen when they met 1 of 4 criteria: (1) membership in child and adolescent psychiatric organizations in Canada and the United States, including their Academies of Child and Adolescent Psychiatry; (2) recipients of federal grants for related research; (3) lead author of at least 2 articles on clinical research in the area from 1999 to 2004 based on Medline citations; or (4) key PC clinical and research leaders with expertise in the area of guideline development and/or emotional and behavioral disorders that present in PC settings. As is typical of activities in which expert opinion is sought, selection of experts was nonrandom and likely does not represent the diversity of community-based PCPs for children. We deliberately sought experts in PC because their practices, although not representative, might serve as a standard for more “representative” PCPs. In addition, we attempted to include a broad range of experts from several disciplines, including family medicine, pediatrics, nursing, psychology, and child psychiatry. This method of nonprobability sampling is frequently used in social science research.12 The survey and procedures were approved by the institutional review board of New York State Psychiatric Institute.
The mean, SDs, and 95% confidence intervals were calculated for each item. For making responses more clinically applicable, a categorical tag such as “treatment of choice,” “first-line,” “second-line,” and “third-line” treatment were designated on the basis of the lowest category in which the confidence interval fell, with boundaries of ≥8 as treatment of choice, ≥7 as first-line, and ≥4 as second-line. Items with responses <4 were considered generally inappropriate in the PC setting.
Seventy-eight of 81 invited experts agreed to participate (return rate of 96%). On average, the experts had been in practice for 20 years (SD: 8.6 years). Fifty-three percent of the experts (n = 40) were PCPs; the remainder were specialty mental health experts in child and adolescent psychiatry and in psychology. Only 76 of the 78 completed surveys were available for data analysis, with one completed survey lost in the mail and another completed survey misplaced prior to data entry.
Overall, findings indicate that experts thought that the strategies used for early identification of emotional problems should vary on the basis of the circumstances of the presentation by an adolescent (Fig 1). In well-child visits, the intervention that was rated as most appropriate was “asking general questions” (average: 8.41; SD: 1.25), described as “surveillance” by the American Academy of Pediatrics. Of the several options that were rated highly, answers were not exclusive; that is, one could indicate that multiple particular approaches were acceptable, but one cannot determine from these responses in isolation which of the acceptable responses was optimal. In the case of an urgent care visit for some medical chief complaint where the physician has another, potentially urgent, problem to attend to, only asking general questions was rated as a first-line practice.
In contrast, when experts were asked about how to gather information regarding depression during an outpatient visit for adolescents who present with an emotional problem as the chief complaint, all choices were rated as strategies of choice or first-line with the exception of depending solely on youth and/or parental complaints. These endorsed strategies included the use of youth- or parent-reported psychosocial or depression checklists, use of the screener interview HEADSS-home, and education/employment, activities, drugs, sexuality, suicide/depression, use of the screener Guidelines for Adolescent Preventive Services questionnaire, and asking the youth both general and depression specific questions (Fig 1). In such instances, both face-to-face interview questions and structured checklists were thought to be of use.
Given the scenario that the PCP already suspects depression during the assessment, experts agreed that interviews should be conducted with the adolescent alone (average: 8.9; SD: 0.4), with the parent/caregiver alone (average: 7.74; SD: 2.05), and jointly with both (average: 8.04; SD: 1.28). Experts also agreed that the use of a youth-report depression questionnaire, such as the Beck Depression Inventory, was appropriate (average: 7.88; SD: 1.53).13
As part of an evaluation to confirm the diagnosis of depression, experts agreed that each of the following was a critical component of the assessment: obtaining history of presenting illness (average: 8.96; SD: 0.2), previous history and treatment for depression and/or emotional problems (average: 8.96; SD: 0.2), medical history (average: 8.66; SD: 0.81), substance use history (average: 8.93; SD: 0.34), psychosocial history (ie, current stressors) (average: 8.92; SD: 0.32), current functioning at school or with friends (average: 8.95; SD: 0.22), and family history (average: 8.78; SD: 0.51). To confirm the diagnosis of depression, the experts agreed that the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) diagnostic criteria should be used (average: 8.34; SD: 0.96). Investigations and examinations that were rated as essential to rule out other causes of depressive symptoms included specific (ie. focused) physical examination as indicated by history (average: 8.19; SD: 1.36), a complete physical examination (average: 8.09; SD: 1.70), and tests for thyroid function (average: 7.35; SD: 2.04).
Experts' responses regarding recommended treatment options for mild and moderate/severe depression are shown in Figs 2 and 3. As seen in these figures, experts' recommendations differed on the basis of severity of depressive symptoms, duration of illness, and the occurrence of comorbidity and/or suicidality or psychosis.
For all cases of confirmed depression, experts agreed that it would not be appropriate to have no follow-up or to ask the patient to return only if their symptoms worsen (Figs 2 and 3). Active monitoring was deemed acceptable but only appropriate in “mild” depression (for discussion, see “Limitations”) lasting <4 weeks. In contrast, experts generally agreed that if a patient has moderate depression regardless of duration, then he or she should be referred to mental health for assessment and/or treatment and the initiation of antidepressants should be considered. When questioned about whether the PCP should consider starting an antidepressant for moderate depression lasting 2 to 4 months, the overall rating was 6.89, just shy of the average of 7 (first line). For patients who present with either psychosis or suicidality, referral to crisis settings or to mental health for assessment and treatment was rated highly, essentially as the treatment of choice. As seen in these figures, the use of antidepressant treatment is viewed as more appropriate the longer a patient has been experiencing symptoms of depression with suicidality. Finally, in patients with comorbid illnesses (eg, substance abuse, conduct disorder, attention deficit hyperactivity disorder, anxiety disorders), experts were less likely to recommend the initiation of an antidepressant by the PCP. In these patients, referral to mental health for assessment and treatment were considered more appropriate.
Treatment of choice among the antidepressants were selective serotonin reuptake inhibitors (average: 8.46; SD: 1.1), specifically fluoxetine (average: 8.54; SD: 0.76). Sertraline (average: 7.83; SD: 1.68), citalopram (average: 7.55; SD: 1.72), and escitalopram (average: 7.26; SD: 2.23) were rated as first-line treatments. Both tricyclic antidepressants (average: 2.88; SD: 2.24) and monoamine oxidase inhibitors (average: 1.86; SD: 1.35) were considered inappropriate in this age group. When the PCP is prescribing antidepressant treatment, experts generally concluded that patients should be followed up within 2 weeks after the initiation of antidepressant treatment (mean: 1.98 weeks: SD: 1.04) and within 7 weeks for those who have stabilized on medication (mean: 7.2 weeks; SD: 4.53). Finally, in terms of management of adverse events, the experts consider all of the following procedures as highly appropriate: assessing for adverse effects after every new medication is started (average: 8.89; SD: 0.39), asking about common adverse effects on every visit (average: 8.61; SD: 0.95), talking to family/caregivers (average: 7.85; SD: 1.53), and examining the patient (average: 7.08; SD: 1.92).
For those with partial improvement with antidepressant treatment, the experts thought that the maximum therapeutic dosage should be used (average: 8.14; SD: 1.17) but not exceeded (average: 2.91; SD: 2.17). If partial improvement is achieved with maximum dosages of an antidepressant, then referral to mental health for additional assessment and treatment recommendations (average: 8.18; SD: 1.2) and/or for implementation of treatment (average: 7.88; SD: 1.37) were deemed as the most appropriate. Similarly, for those who do not improve on antidepressant treatment, again, the maximum therapeutic dosage should be used (average: 7.72; SD: 1.94); however, it should not be exceeded (average: 2.73; SD: 2.06). If there is no response with the maximum dosage of antidepressant treatment, then experts endorsed referral to mental health for additional assessment and treatment recommendations (average: 8.22; SD: 1.22), referral to mental health for actual treatment delivery (average: 7.88; SD: 1.37), or switching to another antidepressant (average: 7.43; SD: 2.22). Finally, in cases in which a mental health referral has been made, experts recommended that PCPs continue to provide support to the patient and see the patient again within 4 weeks (mean: 3.51 weeks; SD: 2.34) even though the patient has been referred.
The results of this survey suggest remarkable consistency among experts in the field on the need for particular assessment, treatment, and management strategies. The findings from the expert survey support the use of strategies for early identification of patients who are at high risk for developing depression. Several strategies were rated as extremely appropriate during well-child visits and appointments at which the chief complaint is an emotional problem. The somewhat lower rating for “depression-specific checklists” in these clinical encounters may be because such a checklist might miss a youth with some other type of behavioral or emotional problem. In urgent care visits, only asking general questions was rated as extremely appropriate (strategy of choice). In contrast to the expert opinion, there is emerging evidence from the literature supporting the use of standardized tools to identify adolescents with emotional problems such as depression even during urgent care visits.14,15 This highlights a common problem with relying solely on expert opinion: expert opinion may not have incorporated the most up-to-date research evidence, or experts may be limited in their willingness to do so because of the constraints of practice settings, which are not usually present when auxiliary research staff conduct studies. However, these contradictory results may also represent the strength of the research evidence (ie, the sum of the current research evidence may not be strong enough to change current expert opinion). A recent example of this is the use of antidepressants in children and adolescents. Although the Food and Drug Administration (FDA) reanalyses of safety data showed an increased risk for adverse effects in patients who were treated with antidepressants compared with those who were given placebo, expert panels from both the American Academy of Child and Adolescent Psychiatry and the American Psychiatric Association remain supportive of the use of antidepressants in children and adolescents, citing that benefits from the medications outweigh the risks.16
With regard to diagnosis, the results of the expert survey support the use of the DSM-IV criteria by PCPs to diagnose depression in adolescents. The experts also agreed that interviews should include meeting with the adolescent alone and with the parents/caregivers. Although this was rated as extremely appropriate, it is not always possible to include family members in the assessment of an adolescent. Factors that may limit the involvement of the family include geographic distance in cases of separated families or the adolescent's reluctance to have his or her family contacted.
The experts agreed that active monitoring, pharmacotherapy with selective serotonin reuptake inhibitors, and psychotherapy can be appropriate under certain clinical circumstances. The choice of treatment strategy varied on the basis of the severity and the duration of depressive symptoms and the occurrence of comorbidity. Thus, findings generally supported the premise that it may be acceptable (and essential) for the PCP to provide treatment for adolescent depression, in instances of moderate to severe depression lasting months, and in the face of difficulties obtaining access to child and adolescent psychiatry. In terms of the question pertaining to whether the PCP should consider starting an antidepressant for moderate depression lasting 2 to 4 months, although the overall rating (6.89) fell short of the average of 7 (first line), there were interesting differences among respondents, with 65% of child and adolescent psychiatrists rating this as a first-line strategy but only 43% of PCPs rating this as an acceptable first-line strategy. These findings might suggest the concerns that PCPs have about taking on this responsibility without additional support and training.
The survey also provided guidance regarding ongoing treatment of adolescents with antidepressants. Strategies for addressing partial or no improvement on antidepressants were recommended including using the maximum therapeutic dosage and referral to mental health if there is only partial improvement with the maximum dosage. Although the experts agreed that PCPs are able to treat adolescents with depression in their practices, they also agreed that in many clinical scenarios, referral to mental health was the strategy of choice, such as with a suicidal or psychotic adolescent; however, many factors influence a successful mental health referral, including patient and parent/caregiver acceptance and availability of mental health professionals. Because it is not always feasible for adolescents to be referred to mental health, guidance for the management of more complicated cases in the PC setting may also be needed.
One aspect of treatment, the appropriate frequency of follow-up when initiating antidepressant treatment, was addressed by the FDA. The monitoring guide provided by the FDA gives clear direction regarding the frequency of face-to-face follow-up for the first weeks of antidepressant treatment. The FDA expert panel recommended that follow-up occur within 1 week of initiation of antidepressant treatment. Similarly, our expert survey recommended <2 weeks as the maximum time before the first follow-up. The lack of research studies that have addressed this issue makes expert guidance particularly important regarding this aspect of treatment.
Finally, in this survey, the definitions of mild, moderate, and severe depression were left up to the clinical judgment of the survey completers; however, for the results of this report to be clinically useful, theses definitions must be clarified. According to the DSM-IV, severity of depressive disorders can be based on symptom count.17 This commonly used method to define depression severity has been used in large population-based studies18 and may be particularly relevant in PC settings, where less severe clinical presentations of depression may be more common. Thus, mild depression may be characterized on the basis of lower scores in standardized depression scales, shorter duration of symptoms, or meeting minimal criteria for depression. Following the DSM-IV, mild depression might be defined as 5 to 6 symptoms that are mild in severity. Furthermore, the patient might experience only mild impairment in functioning. In contrast, depression might be deemed to be severe when a patient experiences all of the depressive symptoms listed in the DSM-IV. Depression might also be considered severe if the patient experiences severe impairment in functioning. Moderate depression falls between these 2 categories. In general, however, even if not all 9 DSM-IV–defined symptoms of depression are present, an adolescent with at least 5 criteria of depression should be considered to be in the severe category if presenting with a specific suicide plan, clear intent, or recent attempt; psychotic symptoms; or severe impairment in functioning (eg, unable to leave home).
There are several limitations to this study. One important limitation is the reliance of respondents' reports about their recommended practices under idealized, perhaps artificial clinical situations that can be captured and described in specific survey questions. Although the survey questionnaire format is an accepted approach for gathering expert opinion on complex clinical situations that commonly present, for purposes of brevity and feasibility, it cannot fully capture the range of possible clinical scenarios that present in the PC setting. Furthermore, the experts could rate several strategies or treatments in each question as appropriate; therefore, it is not clear from the survey results whether these strategies or treatments should be used in combination or would also be appropriate if used alone. To provide appropriate clinical guidance, the survey results cannot serve as freestanding recommendations but need to be developed into coherent guidelines for PCPs. Nonetheless, our findings do provide support for the benefits of expert consensus approaches in addressing questions that would otherwise remain unaddressed in other types of empirical studies.
The high degree of consistency on several recommendations suggests that PCPs can be given clear guidance on the management of depression, especially in settings where mental health services are accessible. Other findings that could provide additional guidance, such as a comparison of responses from those who are and are not PC providers and from providers who work in different health care systems (eg, Canadian versus American experts), are beyond the scope of this article and will be presented in subsequent publications.
We acknowledge the financial support of the Center for the Advancement of Children's Mental Health and Sunnybrook Health Sciences Centre.
The GLAD-PC Steering Committee consists of the following (listed alphabetically): Boris Birmaher, MD, Western Psychiatric Institute and Clinic, University of Pittsburgh; John Campo, MD, Ohio State University; Greg Clarke, PhD, Center for Health Research, Kaiser Permanente; Dave Davis, MD, University of Toronto; Angela Diaz, MD, Mount Sinai School of Medicine; Allen Dietrich, MD, Dartmouth Hitchcock Medical Center; Graham Emslie, MD, University of Texas Southwestern Medical School; Bernard Ewigman, MD, Department of Family Medicine, University of Chicago; Eric Fombonne, MD, McGill University; Sherry Glied, PhD, Columbia University; Kimberly Eaton Hoagwood, PhD, Office of Mental Health, New York State, and Columbia University; Charles Homer, MD, National Initiative for Children's Healthcare Quality; Miriam Kaufman, MD, Hospital for Sick Children, University of Toronto; Kelly J. Kelleher, MD, Ohio State University; Stanley Kutcher, MD, Dalhousie Medical School; Danielle Laraque, MD, American Academy of Pediatrics New York Chapter 3, District II, and Mount Sinai School of Medicine; Michael Malus, MD, Department of Family Medicine, McGill University; James Perrin, MD, Massachusetts Medical School, Harvard Medical School; Harold Pincus, MD, Columbia University; Brenda Reiss-Brennan, APRN, Intermountain Health, UT; Diane Sacks, MD, Canadian Pediatric Society; Ruth E.K. Stein, MD, Forum for Child Health, New York Academy of Medicine, Albert Einstein College of Medicine; Bruce Waslick, MD, Baystate Health Systems, MA.
- Accepted May 29, 2007.
- Address correspondence to Amy H. Cheung, MD, Department of Psychiatry, University of Toronto, 33 Russell St, Third Floor Tower, Toronto, Ontario, M5S 2S1, Canada. E-mail:
Financial Disclosure: Dr Cheung is on the Speakers' Bureau of Eli Lilly; Dr Jensen has received several unrestricted educational grants from Eli Lilly, McNeil, and Janssen-Ortho and is a consultant for Shire-Richwood, UCB Pharma, McNeil, and Janssen-Ortho and is also on the Speakers' Bureau for UCB Pharma, McNeil, and Janssen-Ortho. The other authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Fleming JE, Offord DR, Boyle MH. Prevalence of childhood and adolescent depression in the community: Ontario Child Health Study. Br J Psychiatry.1989;155 :647– 654
- ↵Cheung A, Dewa C. Canadian Community Health Survey: depression and suicidality among adolescents aged 15–18. Healthcare Policy.2006;2(2) :76– 89
- ↵Kessler RC, Avenevoli S, Ries Merikangas K. Mood disorders in children and adolescents: an epidemiologic perspective. Biol Psychiatry.2001;15 :1002– 1014
- ↵Zuckerbrot RA, Cheung AH, Jensen PS, Stein REK, Laraque D; GLAD PC Steering Group. Guidelines for Adolescent Depression in Primary Care (GLAD-PC): part I—identification, assessment, and initial management. Pediatrics.2007;120(5) . Available at: www.pediatrics.org/cgi/content/full/120/5/e1299
- ↵Cheung A, Zuckerbrot RA, Jensen PS, et al. Guidelines for Adolescent Depression in Primary Care (GLAD-PC): part II—treatment and ongoing management. Pediatrics.2007;120(5) . Available at: www.pediatrics.org/cgi/content/full/120/5/e1313
- ↵Kahn DA, Frances A, Docherty JP. Expert consensus guidelines series: treatment of bipolar disorder. J Clin Psychiatry.1996;57(suppl 12A) :3– 89
- ↵Mutchnick RJ, Berg BL. Research Methods for the Social Sciences. Needham Heights, MA: Allyn and Bacon; 1996
- ↵Beck AT, Steer RA. Manual for the Beck Depression Inventory. San Antonio, TX: Psychological Corporation; 1987
- ↵Borowsky IW, Mozayeny S, Ireland M. Brief psychosocial screening at health supervision and acute care visits. Pediatrics.2003;112(pt 1 ):129– 133
- ↵Zuckerbrot RA, Maxon L, Pagar D, Davies M, Fisher PW, Shaffer D. Adolescent depression screening in primary care: feasibility and acceptability. Pediatrics.2007;119 :101– 108
- ↵American Academy of Child and Adolescent Psychiatry. Psychiatric medication: March 2004 FDA Advisory. Available at: www.aacap.org/page.ww?section=Psychiatric=Medication&name=March+2004+FDA+Advisory. Accessed December 4, 2007
- ↵American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994
- Copyright © 2008 by the American Academy of Pediatrics