Objective. There has been a strong push toward the recognition and treatment of children with behavioral health problems by primary care pediatricians. This study was designed to assess the extent to which a sample of primary care pediatricians diagnose and treat behavioral health problems and to identify factors that may contribute to their behavioral health practice.
Methods. A standard interview was conducted with 47 pediatricians who work in primary care settings in a predominantly urban setting in North Carolina. Pediatricians' responses to questions about the estimated percentage of children in their practice with a behavioral health disorder, tools used to make diagnoses, frequent and infrequent diagnoses made, comfort level with making a diagnosis, reasons for not making a diagnosis, use of psychotropic medications, types of nonmedication interventions provided, educational background, and needs involving behavioral health issues were evaluated.
Results. Pediatricians estimated that the average percentage of children in their practices with a behavioral health disorder was 15%. The study did not find significant differences in perceptions related to time in practice or gender of the pediatric provider. The most frequent behavioral health diagnosis was attention-deficit/hyperactivity disorder (ADHD), and the majority incorporated behavioral questionnaires, expressed a high level of comfort with the diagnosis, and frequently or occasionally prescribed stimulants. Variability was noted in both practice and comfort for other behavioral health disorders. Slightly fewer than half of the pediatricians frequently diagnosed anxiety and depression. Those who make these diagnoses commonly incorporated questionnaires and reported frequent or occasional use of selective serotonin reuptake inhibitors. Comfort in making the diagnosis of anxiety was highly associated with use of selective serotonin reuptake inhibitors. The vast majority (96%) of pediatricians provided nonmedication interventions, including supportive counseling, education for coping with ADHD, behavior modification, and/or stress management. Diagnosis and treatment of severe behavioral health disorders were infrequent throughout the pediatric practices. Areas of greatest educational interest included psychopharmacology, diagnosis and treatment of depression and anxiety, and updates on ADHD. The majority of pediatric providers did not identify a need for education about several high-prevalence disorders that they do not frequently diagnose or treat, including conduct disorder and substance abuse.
Conclusions. Pediatricians in this sample frequently diagnosed and treated ADHD. For all other behavioral health disorders, pediatricians reported variability in both comfort and practice. They frequently provided both pharmacologic and nonpharmacologic treatments for children and adolescents with mild to moderate behavioral health disorders but not for severe disorders. Although they identified needs for additional education for anxiety and depression, the majority did not identify educational needs for several high-prevalence behavioral health disorders, including conduct disorder and substance abuse.
Between 14% and 20% of children and adolescents experience significant mental health disorders.1–5 However, the shortage of mental health providers, stigma attached to receiving mental health services, chronic underfunding of the public mental health system, decreased reimbursement to mental health providers, and disparate insurance benefits have contributed to the fact that only 2% of these children are seen by mental health specialists. In contrast, ∼75% of children with psychiatric disturbances are seen in primary care settings,6 and half of pediatric office visits involve behavioral, psychosocial, or educational concerns.1
There are compelling reasons to encourage an expanding role for pediatricians in the behavioral health care of their patients. Services received in the child's “medical home” may be less stigmatizing and better coordinated with the child's medical care. Children and families with untreated mental health problems overutilize health services; thus, recognition and treatment of behavioral health disorders in the primary care setting may reduce health care costs.6,7 More important, pediatric providers' early treatment of mild to moderate behavioral health disorders decreases the risk of long-term disability for children and adolescents, because disturbed children are more likely to become disturbed adults.3 Furthermore, when referral to a mental health specialist is needed, pediatricians can facilitate use of these services. Pediatric referral has been found to be an important predictor of mental health service use,8 especially when a parent consults with a pediatric provider about behavioral and emotional issues.4
In 1993, the American Academy of Pediatrics (AAP) adopted a policy statement that focused on the prevention, early detection, and management of behavioral, developmental, and social problems as a main part of the scope of pediatric practice.9 This commitment to the psychosocial aspects of pediatric care was reaffirmed in 2001.10
The AAP has contributed to the development of educational programs and tools to assist pediatricians in this broadening role. In 1996, the AAP and partner organizations published the Diagnostic and Statistical Manual for Primary Care, Child and Adolescent Version,11 providing a framework for primary care physicians to recognize and diagnose a wide spectrum of children's behavioral and developmental problems.12 In 2000, the AAP developed a clinical practice guide to assist pediatricians with the diagnosis and treatment of attention-deficit/hyperactivity disorder (ADHD). In 2002, with support from the Maternal and Child Health Bureau, Jellinek et al13 edited Bright Futures in Practice: Mental Health—Volume II and its accompanying Tool Kit, providing additional practical assistance in prevention and treatment of childhood behavioral health problems.
At the same time, other factors have worked against a psychosocial focus in pediatric care. The most powerful has been economic. Managed care and other reimbursement strategies have created behavioral health carve-outs with separate provider panels, administrative procedures, and utilization controls. Primary care pediatricians have been pressed to treat more patients, with an average visit lasting 13 minutes. Lack of reimbursement for behavioral health treatment and disincentives to make referrals to mental health specialists have further discouraged pediatric providers from addressing psychosocial issues.14
Another major factor impeding behavioral health practice involves pediatric providers' perceived skills and level of confidence. The Future of Pediatric Education II Task Force report published in 2000 noted that practitioner surveys identified a lack of confidence in the ability to identify and treat developmental and psychological problems as a major barrier to care. The report commented on the implications of this lack of confidence for pediatric practice and education.15 In a study published in 2001, 90% of responding pediatricians felt responsibility for recognizing symptoms of depression in children and adolescents, whereas only 27% felt responsibility for treatment. Almost half lacked confidence in their skills to recognize depression, and few (10%–14%) had confidence in their ability to treat.16
In addition, parents are unwilling to bring up behavioral health issues in the pediatric setting. Some evidence suggests that parents may have a real reluctance to discuss or consult with a pediatrician about behavioral problems.4,7
In light of these conflicting influences, the present study was designed to evaluate the extent to which a sample of primary care pediatricians diagnose and treat behavioral health disorders. It examined factors that might have an impact on recognition and treatment, including length of time in practice, gender, and comfort level. In addition, it assessed the pediatricians' perceived practice and educational needs in relation to behavioral health diagnosis and treatment.
Subjects were pediatricians (n = 47) who work in private practices in a 3-county area in the Piedmont region of North Carolina. All 49 pediatricians who had been employed for at least 1 year in private practice in the 3-county area were invited to participate. Only 2 pediatricians declined. Forty-five of the pediatric providers were located in an urban setting. Mean time in practice was 14.7 years (range: 1–36 years). All of the pediatric providers had completed general pediatric residencies, whereas 6 had additional specialty training. Although 31 of the pediatricians received their training in 1 of the 5 residency programs in North Carolina, the remainder trained in diverse geographic areas, including New York; Florida; Connecticut; Georgia; Ohio; Kentucky; Washington, DC; Utah; and Pennsylvania. Gender of the respondents included 26 women and 21 men.
The study was approved by the Institutional Review Board, and verbal permission was given by the subjects for participation in the study. Pediatricians were interviewed individually in their offices by a clinical psychologist using a standard interview guide.
The standard interview guide included questions concerning demographics, estimated percentage of children in their practice who met criteria for a behavioral health diagnosis, tools used to make diagnoses, reasons for not making a behavioral health diagnosis, educational background, and needs involving behavioral health issues. The pediatricians first were asked to name specific behavioral health diagnoses that they frequently made, and then they were asked about diagnoses that they infrequently made. As these were open-ended questions, the same disorder, such as anxiety, could be named as a frequent or infrequent diagnosis. There was no limit to the number of disorders that the pediatricians were allowed to name. Each diagnostic response was followed by a Likert-scale concerning level of comfort in making the diagnosis ranging from 1 to 5, with 1 being “not very comfortable” and 5 being “entirely comfortable.” The interviewer read to the pediatricians a list of psychotropic medications (stimulants, atomoxetine [Strattera], clonidine/guanfacine, antidepressants [tricyclic], selective serotonin reuptake inhibitors [SSRIs], anxiolytics, anticonvulsants, and neuroleptics). The pediatricians were asked to describe the frequency with which they initiated each medication. Response choices included “never,” “rarely,” “occasionally,” and “frequently.” Pediatricians were also asked to describe types of nonmedication interventions offered for behavioral health problems.
On the basis of pediatrician perception, the mean percentage of children who were believed to have a behavioral health disorder was 15% (range: 1%–50%; Fig 1).
Factors, including time in practice and gender, were examined in terms of recognition of behavioral health disorders. Pediatricians were divided into 2 groups on the basis of years of time in practice. Those with 13 or fewer years of experience (n = 23) reported 14% of their patients to have a behavioral health disorder, whereas those with 13 or more years of experience (n = 24) reported 15% of their patients to have a behavioral health disorder. Difference between the groups was nonsignificant (t = .042; P > .96). Female pediatricians indicated that 13% of their patients had a behavioral health disorder, whereas male pediatricians indicated that 18% of their patients had a behavioral health disorder. Difference based on gender was nonsignificant (t = 1.56; P > .13).
Forty-one (87%) of the pediatricians used checklists or Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria in making behavioral health diagnoses. The Vanderbilt Rating Scales17 were used by 31 pediatric providers, the Conners Rating Scales–Revised18 were used by 19 providers, and the Attention Deficit Disorders Evaluation Scale19 was used by 8 providers. In addition, 17 of the pediatricians indicated use of depression checklists. None of the pediatricians routinely screened children for behavioral health disorders.
ADHD was named by 45 of the 47 pediatric providers as their most frequent behavioral health diagnosis. Depression (n = 19), anxiety (n = 17), and oppositional defiant disorder (n = 13) followed ADHD with some mention of adjustment (n = 7) and eating disorders (n = 3). Substance abuse and pervasive developmental disorders were named by only 1 pediatrician each, and conduct disorder was not named. The level of comfort in frequently made diagnoses was generally high (Table 1).
The less frequent behavioral health diagnoses named by the pediatricians were anxiety (n = 23), depression (n = 22), bipolar disorder (n = 12), oppositional defiant disorder/conduct disorder (n = 11), and psychosis (n = 10). Other less frequent diagnoses included pervasive developmental disorders (n = 7), adjustment disorder (n = 6), eating disorder (n = 3), and substance abuse (n = 2). Comfort level in making these diagnoses was generally within the middle range (Table 2).
The association between comfort level and frequency of making the diagnoses of anxiety and depression was examined. For those who frequently made the diagnosis of anxiety, the level of comfort (mean: 3.9) was significantly higher (t = 3.2; P > .003) than for those who infrequently made this diagnosis (mean: 2.9). For depression, the level of comfort (mean: 3.8) was significantly higher (t = 2.9; P > .005) than for those who infrequently made this diagnosis (mean: 3.0).
When asked about reasons for not making behavioral health diagnoses, the most frequent response involved uncertainty of the diagnosis. For example, a number of pediatricians were concerned about ruling out possible comorbid conditions, whereas others questioned whether a child met full diagnostic criteria. Mixed features in presentation, lack of information or knowledge of the child, and parental discrepancies concerning perceptions of the child all contributed to uneasiness in making a diagnosis. A second prohibitive factor involved perceived impact of the diagnosis on the child and family. Some pediatricians were concerned about possible negative effects of labeling a child. Other providers indicated a hesitance when parents were not ready to accept the diagnosis. A third factor involved pediatrician characteristics. A number cited lack of personal comfort in making behavioral health diagnoses, whereas others cited lack of experience or training. Notably, only 1 pediatrician mentioned reimbursement as a reason for not diagnosing, and only 1 mentioned medical-legal concerns.
Pediatricians initiated the use of stimulants for the treatment of ADHD at a high level, with 94% of them indicating frequent or occasional use. Nearly 50% initiated the use of SSRIs for the treatment of depression or anxiety. There was infrequent use of other antidepressants, clonidine/guanfacine, or anxiolytics. Four of the pediatricians indicated that they occasionally prescribed bupropion. Almost none of them used anticonvulsants or neuroleptics for behavioral health disorders (Table 3).
Additional analysis explored factors associated with the initiation of SSRIs. A stepwise regression analysis was performed with frequency of SSRI use as the dependent variable and comfort with the diagnosis of anxiety, percentage of behavioral health diagnosis, time in practice, and gender as independent variables. Results suggested that comfort with the diagnosis of anxiety disorder was highly associated with the use of SSRIs (t = 3.2; P > .006) and accounted for 40% of the variance. The other factors did not contribute significantly to the model.
The vast majority (96%) of pediatricians indicated that they provided nonmedication interventions for their patients with behavioral health problems. More than half (51%) of the physicians provided counseling, frequently for adolescents with anxiety or depression. The pediatricians were reluctant to describe their interventions as therapy, but many of the examples given or observed involved reflective listening, problem solving, contracting, repeated contact after traumatic events, and cognitive reframing. Nearly half (48%) spent time focused on issues concerning ADHD. This included information about the cause of the disorder, school modifications, organizational skills, consistency in parenting, structured activities, educational rights, self-esteem issues for the child, and parental self-care. Approximately one third (34%) provided assistance with parenting and child management issues, and another third (29%) offered behavior modification techniques to address noncompliance and oppositional behavior in children. A few (8%) pediatricians focused on stress-related issues and provided stress management skills, hypnosis, or relaxation techniques.
To accommodate increased time demands as a result of behavioral health issues, 50% of the pediatric practices indicated that they extended time on the initial visit when a behavioral health issue was anticipated, and 41% extended time for follow-up visits. One practice designated specific days to see children with behavioral health issues, and 5 of the practices set specific times, such as late afternoon, for behavioral health consultations.
When queried about changes that could result in pediatricians' increasing their capacity for behavioral health diagnoses, the most frequently cited response was more training and experience. Pediatricians also indicated a need for more personal comfort in making diagnoses. More time for evaluation, routine behavioral health screening, and greater assistance from and access to mental health providers were mentioned as needed changes. Less frequently cited responses included increased awareness of behavioral health problems by parents, enhanced reimbursement, and tools to assess psychosocial issues.
The majority of pediatricians reported a significant lack of behavioral health training during residency. When asked in which areas of behavioral health they received best training, 57% responded with “none.” Of the 27 pediatricians who made this response, 24 had been in practice for >5 years. For pediatricians who had been in practice for 5 years or less, 75% named 1 or more behavioral health diagnoses in which they felt sufficiently trained. The most frequently mentioned diagnosis was ADHD. Areas in which they most frequently felt least prepared were depression and anxiety. In terms of future training, the areas of greatest interest included psychopharmacology (n = 15), diagnosis and treatment of depression and anxiety (n = 13), updates on ADHD (n = 9), and recognizing comorbidities (n = 5). Several pediatricians also requested additional training in the diagnosis and treatment of oppositional defiant disorder, conduct disorder, and pervasive developmental disorders. Only 3 pediatric providers mentioned substance abuse.
The occurrence of behavioral health disorders in their patients, as reported by pediatricians in the present study, was similar to prevalence rates reported in the literature.4–8 Almost one third of the respondents reported that 20% or more of their patients had a diagnosable behavioral health disorder. Their perceptions did not seem to depend on experience, as measured by time in practice, or gender. Although pediatricians who had recently completed their residencies reported more exposure to behavioral health issues during training, many of the more experienced pediatricians had acquired training through continuing medical education and reading about diagnosis and treatment of these disorders.
On the whole, the evolving focus on psychosocial issues was apparent in the practices of the majority of the pediatricians in this sample. With the exceptions of substance abuse and conduct disorder, the behavioral health diagnoses most frequently made by the pediatricians were the most frequently occurring disorders: ADHD, anxiety, and depression.1 It is unclear whether the low rate of diagnosis of substance abuse and conduct disorder represents a predominance of younger children in their practices or failure to recognize features of these disorders. Study results suggested that pediatricians in this sample rarely diagnosed or initiated treatment for more serious behavioral diagnoses.
Torstenson's20 belief that the diagnosis and treatment of ADHD has changed from psychiatrists to pediatricians was supported by study results. The majority of pediatricians in this study used objective instruments to diagnose and treat ADHD, similar to findings reported by Block.21 These pediatricians reported a high level of comfort in making a diagnosis of ADHD and prescribed stimulants for treatment. Almost half of the pediatricians provided educational or other nonpharmacologic interventions for ADHD. The pediatric providers seemed to have incorporated the diagnosis of ADHD as an ongoing part of their practice, and they expressed a desire for updates, not basic information, on ADHD through continuing education.
The diagnosis of anxiety and depression seemed to be shifting more gradually to pediatric providers. There was use of semiobjective instruments to diagnose these conditions, and nearly half of the pediatricians used SSRIs frequently or occasionally to treat these disorders. Our results contrasted with an earlier report in which psychotropic medications were used nearly exclusively for children with ADHD, and little medication was used for other behavioral health disorders.22 Findings also contrasted with survey results in which 86% of pediatricians lacked confidence to treat depression with medication.16 On the basis of their qualitative responses, it did seem that the pediatricians in this sample were more likely to treat mild to moderate depression than major depression, as postulated by Torstenson.20
In light of the fact that depression in adolescents is a significant risk factor for both suicide and substance abuse, it is of concern that many pediatricians in this sample listed anxiety and depression as infrequently made diagnoses and that none used instruments to routinely screen for behavioral health problems.16 As primary care settings may be the only environment in which these adolescents are seen, their high mortality rate from accidents, homicide, and suicide suggests the critical need for pediatricians to recognize and inquire about these symptoms.6 It is also of concern as children with emotional disorders (eg, affective psychoses, depression, anxiety) have been found to be significantly less likely to receive a prescription medication than children with disruptive disorders yet incur significantly greater hospitalization expenditures.23
The pediatric provider's level of comfort seemed to affect significantly the frequency of making diagnoses of anxiety and depression and may therefore have an impact on treatment. It is encouraging that additional training in diagnosing and treating depression and anxiety was 1 of the most frequently requested continuing education experiences.
Nearly all of the pediatricians in this study provided nonmedication, behavioral health interventions for their pediatric patients and families. The pediatricians tended to discount their interventions and were reluctant to describe them as therapeutic. For example, 1 pediatrician had worked closely with several adolescents who expressed suicidal ideation after the break-up of relationships. The pediatrician seemed surprised when the interviewer suggested that the physician's supportive counseling probably saved lives. One pediatrician worked with a child who had a choking experience and developed anxiety about eating specific foods. The pediatrician pointed out to the child that he had eaten 3 meals a day for 10 years without a previous choking episode and that it was unlikely that the event would recur. The physician also suggested that the child chew gum when thoughts of choking occurred. Although the intervention was successful, words such as “cognitive reframing” or “distraction” were not used to describe the approach.
A limitation of this study was its reliance on self-reported behavior of the pediatricians. This was somewhat offset by the open-ended nature of many of the survey questions that required the pediatrician to generate a response rather than suggesting responses for them. Another limitation of the study was its regional nature. The study reflected regional practice patterns, not necessarily generalizable to other geographic areas or managed care environments. However, its findings about the education and training of pediatricians reflect a broader perspective, because the pediatricians trained in diverse residency programs. A positive aspect of the study design was its completion in a primary care environment, rather than an academic medical center, which may increase generalizability of the findings.
Perhaps the most important and most generalizable findings of this study involve the need for increased training and for continuing medical education in behavioral health. The greatest perceived needs were in psychopharmacology and in the diagnosis and management of anxiety and depression. The pediatricians in this sample were generally confident in the diagnosis and management of ADHD and did not perceive a need for basic education in this area; they were interested in updates. Conversely, there was an absence of routine behavioral health screening as well as evidence of gaps in knowledge about several prevalent childhood disorders, including conduct disorder and substance abuse. Of concern, there was limited interest expressed in education about these disorders. Strategies to enhance pediatricians' behavioral health practice will need to address their unidentified needs, as well as those that they do identify. Educational efforts should also recognize and encourage pediatricians' use of nonpharmacologic behavioral health interventions, typically used without recognition of formal nomenclature describing the interventions or their therapeutic value.
Future studies are needed to follow changes in pediatric practices over time. Objective measures of behavioral health practice will strengthen our understanding of pediatricians' roles and needs in providing behavioral health care to their patients.
Responses from the pediatricians suggested their strong interest in diagnosing and treating behavioral health disorders within their perceived limits and level of comfort. They were very concerned about the correctness of these diagnoses and considered the impact on both the child and the family. The pediatricians reflected a sense of carefulness when considering a behavioral health diagnosis, although there was little expressed concern about medical-legal issues. Logistically, many were attempting to accommodate the need for more time required by behavioral health issues in their office practices, but frequently they voiced the need for even more time for these patients. Although they used instruments for specific diagnoses, there was no reported use of routine screening measures. The pediatric providers used a variety of pharmacologic and nonpharmacologic treatments for mild to moderate behavioral health problems: most frequently for ADHD, less frequently for anxiety and depression. The pediatricians expressed interest in continuing education to address their perceived needs.
This research was supported by the Duke Endowment as part of the Primary Care-Children's Mental Health Initiative.
- ↵American Academy of Pediatrics, Committee on Psychosocial Aspects of Child and Family Health. The pediatrician and the “new morbidity.” Pediatrics.1993;92 :731– 733
- ↵American Academy of Pediatrics, Committee on Psychosocial Aspects of Child and Family Health. The new morbidity revisited: a renewed commitment to the psychosocial aspects of pediatric care. Pediatrics.2001;108 :1227– 1230
- ↵Wolraich ML, Felice ME, Drotar D, eds. The Classification of Child and Adolescent Mental Diagnoses in Primary Care: Diagnostic and Statistical Manual for Primary Care (DSM-PC) Child and Adolescent Version. Elk Grove Village, IL: American Academy of Pediatrics; 1996
- ↵Wolraich ML. Addressing behavior problems among school-aged children: traditional and controversial approaches. Pediatr Rev.1997;18 :266– 270
- ↵Jellinek MS, Patel BP, Froehle MC, eds. Bright Futures in Practice: Mental Health. Arlington, VA: National Center for Education in Maternal and Child Health; 2002
- ↵American Academy of Pediatrics, Future of Pediatric Education II Task Force. Organizing pediatric education to meet the needs of infants, children, adolescents, and young adults in the 21st century. Pediatrics.2000;105(suppl 1) :176
- ↵Wolraich ML. Vanderbilt ADHD Teacher Rating Scale (VADTRS) and the Vanderbilt ADHD Parent Rating Scale (VADPRS). Oklahoma City, OK: University of Oklahoma Health Sciences Center; 2003
- ↵Conners C. Conners' Rating Scales–Revised Technical Manual. North Tonawanda, NY: Multi-Health Systems; 1997
- ↵McCarney SB. Attention Deficit Disorders Evaluation Scale. 2nd ed. Columbia, MO: Hawthorne Educational Services, Inc; 1995
- ↵Gardner W, Kelleher KJ, Wasserman R, et al. Primary care treatment of pediatric psychosocial problems: a study from pediatric research in office settings and ambulatory sentinel practice network. Pediatrics.2000;106(4) . Available at: www.pediatrics.org/cgi/content/full/106/4/e44
- ↵Guevara JP, Mandell DS, Rostain AL, Zhao H, Hadley TR. National estimates of health services expenditures for children with behavioral disorders: an analysis of the medical expenditure panel survey. Pediatrics.2003;112(6) . Available at: www.pediatrics.org/cgi/content/full/112/6/e440
- Copyright © 2004 by the American Academy of Pediatrics