www.cuwai.com Contextual Emotion-Regulation Therapy for Childhood Depression: Description and Pilot Testing of a New Intervention MARIA KOVACS, PH.D., JOEL SHERRILL, PH.D., CHARLES J. GEORGE, M.S., MYRNA POLLOCK, M.S.W., RAMESHWARI V. TUMULURU, M.D., AND VINCENT HO, M.D. ABSTRACT Objective: To pilot test the acceptability and efficacy of contextual emotion-regulation therapy (CERT), a new, developmen- tally appropriate intervention for childhood depression, which focuses on the self-regulation of dysphoria. Method: Two samples of convenience (n = 29, n = 2) served to verify some CERT constructs; it was then operationalized in a treatment manual. To pilot test CERT, 20 children (ages 7Y12; 35% girls) with DSM dysthymic disorder (mean duration 24.4 months) entered an open, 30-session, 10-month, 4-phase trial, with 6- and 12-month follow-up. Assessments included independent clinical evaluations and self-rated questionnaires. Results: Fifteen children completed therapy, four were administratively terminated, and one dropped out. Completers did not clinically differ from the rest, but they were more likely to have better educated and less depressed mothers and intact families. At the end of treatment, 53% of the completers had full and 13% partial remission of dysthymia (remission from superimposed major depression was 80%). By 6- and 12-month follow-up, 79% and 92% had full remission of dysthymia (p G .0001). Self-reported depressive and anxiety symptoms significantly declined by the end of treatment (p G .001) and remained so throughout follow-up. Conclusions: CERTenables clinicians to Bmatch[ the intervention to children_s emotion regulatory needs and symptoms and was readily accepted by families. The promising results suggest the need for a randomized trial. J. Am. Acad. Child Adolesc. Psychiatry, 2006;45(8):892Y903. Key Words: childhood depression, dysthymia, emotion regulation, treatment. The nonsomatic treatment of depressive disorders in (Compton et al., 2004; Curry, 2001; Kendall and juveniles continues to be a mental health priority Choudhury, 2003; Michael and Crowley, 2002). However, research in this area has focused typically Accepted March 14, 2006. on adolescents 12 years of age and older and short-term Drs. Kovacs and Tumuluru are with the Department of Psychiatry, University of Pittsburgh School of Medicine and Western Psychiatric Institute and Clinic, interventions for episodes of major depressive disorder Pittsburgh; Dr. Sherrill is with the National Institute of Mental Health, (MDD; e.g., Curry, 2001). Although in terms of sheer Bethesda, MD; Mr. George and Ms. Pollock are with the University of numbers, children 12 years old and younger with Pittsburgh Medical Center; Dr. Ho is with CYKE, Inc., Atlanta. chronic depression represent a minority of the popu- This article is based on work that was completed while Dr. Sherrill was at the University of Pittsburgh. Views expressed within this article represent those of the lation of depressed juveniles, there are several reasons authors and are not intended to represent the position of National Insitute of why they should be targeted for intervention research. Mental Health (NIMH), National Institutes of Health (NIH), or the In clinically referred children, dysthymic disorder Department of Health and Human Services (DHHS). This study was supported by NIMH grant 5R21 MH55244; preparation of (DD), one form of chronic depression, lasts almost the manuscript was supported by NIMH Program Project grant 5P01 4 years on average, and about 18% of youngsters with MH56193. The authors thank Melina Orsini-Young, M.S.W., for her MDD have episodes longer than 1.5 years (Kovacs participation in the study and an anonymous reviewer for helpful suggestions. Correspondence and requests for the treatment manual, parents_ manual, and et al., 1997). Such extended periods of depression are unpublished assessment tools to Dr. Maria Kovacs, WPIC, 3811 O_Hara Street, likely to have deleterious functional and developmental Pittsburgh, PA 15213; e-mail: firstname.lastname@example.org. consequences across the elementary school years, a pe- 0890-8567/06/4508-0892Ó2006 by the American Academy of Child and Adolescent Psychiatry. riod during which various basic skills must be acquired. DOI: 10.1097/01.chi.0000222878.74162.5a In addition, protracted dysphoric mood at a young age 892 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com NEW TREATMENT FOR CHILDHOOD DEPRESSION may interfere with the structural and functional problems (e.g., acute suicidality), and includes a module maturation of cortical and subcortical brain circuits, on coping skills training. a normative process that continues into adolescence (Casey et al., 2000; Durston et al., 2001). Finally, Framework and Rationale among youngsters, DD typically predates subsequent The relationships between stress and coping and MDD (Kovacs et al., 1994) and thus represents one of between stress and depression, which have been exten- the earlier manifestations of vulnerability to major mood disorders. sively documented, serve as a general framework for the One likely reason for the lack of psychological treatment. Within this framework, CERT posits that intervention trials with clinically depressed children is problems in the adaptive self-regulation of distress and that available treatments generally are not developmen- dysphoria compromise a child_s ability to cope and fa- tally appropriate or sensitive (see Weisz et al., 1992), a cilitate the emergence of clinical depression. The specific concern specifically noted with regard to the cognitive- assumptions of CERT are that dysfunctional self- behavioral therapies (Hammen et al., 1999; Kendall and regulation of distress and dysphoria, a key characteristic Choudhury, 2003). The emphasis on short-term inter- of depressed children, developmentally precedes the ventions for MDD probably reflects various constraints onset of depressive disorder; stress, along with contextual imposed by health care insurers as well as a range of variables, interact with preexisting regulatory difficulties practical and clinical issues (e.g., Weisz et al., 1997). to enable progression of dysphoric emotion to dysphoric Thus, the relative scarcity of treatment trials with mood and then to a depressive disorder; ongoing affective chronically depressed children is likely to mirror a distress (dysphoria, irritability, anhedonia) maintains combination of factors along with the belief that brief the depressive syndrome and therefore its reduction intervention may be inappropriate for dysthymia. Con- must be a salient treatment target; and dysfunctional sequently, there is scant empirical information about regulatory responses to distress and dysphoria can be whether and how long-term depression in childhood can identified and remediated. be ameliorated. In the present article, we describe a new The CERT explanatory paradigm is presented to intervention, developed specifically for school-age de- families as the following themes: (1) the child_s Bmood pressed children, and report on its application to treat problems[ are likely to have come about gradually, along chronic depression. The goals of this open pilot testing with earlier signs of sensitivity to even small, everyday trial were to assess the intervention_s acceptability distress elicitors (i.e., has had difficulties in the adaptive (willingness to remain in treatment) and efficacy down-regulation of distress), (2) the depressive episode (diagnostic and symptom severity outcomes). has been preceded by one or more stressors or difficulties (distress elicitors), representing either a major negative event or an accumulation of minor hassles, (3) given that CONTEXTUAL EMOTION-REGULATION THERAPY this child has been historically sensitive to distress elici- (CERT) FOR DEPRESSION tors, as well as specific contextual issues, he or she has CERT, which focuses on the self-regulation of distress been unable to modulate or down-regulate the dysphoric and dysphoria, was inspired by a combination of clinical emotion, leading to protracted negative mood, which has observations, the literature on stress and coping, and adverse effects on the parentYchild and other relation- a developmental approach to emotion regulation. It is a ships, and (4) protracted negative mood and its social goal-directed and problem-focused intervention, with a repercussions negatively affect other areas of functioning, strong didactic emphasis. The intervention_s explanatory contributing to a spiral of symptoms, which culminate in paradigm and its application to chronic depression have clinical depression. been operationalized in a working treatment manual, Examples are used to reframe the child_s emotion which also addresses practical issues (e.g., basic thera- regulatory difficulties and depressive symptoms in the peutic stance, effective use of language with children), context of stress and coping, which helps families to provides Bhow to[ guidance (e.g., introducing the anchor key CERT concepts in their daily lives. The explanatory framework, selecting target symptoms and treatment goals are identified as helping the child to intervention techniques, reframing mood-related in- recover from his or her current depressive disorder cidents in terms of emotion regulation), notes special (symptom reduction) and become more skilled at J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 893 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com KOVACS ET AL. regulating dysphoric emotions and in responding to the scene, finding some distracter, or trying to engage distress-eliciting situations (Bcoping skills training[ with caregivers_ help for distress relief. The increasing so- an emphasis on adaptive emotion regulatory responses). phistication of language, cognitive, and interpersonal skills gradually enables children and adolescents to master a Development of Emotion Regulatory (ER) Responses variety of additional ER responses, which can involve their A major conceptual basis of CERT is that starting in own thought processes (cognitions), more strategic infancy and continuing thereafter, ER responses in general discourse and interaction with caregivers or peers, or and those evoked by distress and dysphoria in particular manipulation of features of their environment. In general, unfold as part of a complex normative developmental pro- the family context and the social Bregulators[ therein play cess within a social context (e.g., Kopp, 1989; Thompson, key roles in the unfolding and use of distress-specific ER 1994). How this process takes place, what its key com- responses during childhood, whereas peers as Bsocial ponents are, and how salient features are best defined or regulators[ become important in adolescence. measured have been topics of lively debates during the Overall, the developmental literature reveals several past 15 years or so. Kopp_s 1989 article on the regulation important features of the normative unfolding of distress/ of distress and negative emotions in infancy has served as dysphoria self-regulation. First, the vast majority of ER an important stimulus for research on this construct. A responses to distress, which human beings display, are monograph entirely dedicated to the development of products of learning, having been molded and shaped by emotion regulation (Fox, 1994), a special issue of the social context in general, and the family context in par- journal Child Development 10 years later (March/April ticular. Second, a wide variety of behaviors and actions 2004), as well as various books on this topic (e.g., Bradley, can serve as ER responses, which typically (but not 2000; Garber and Dodge, 1991) reveal both the invariably) become increasingly sophisticated with age. multifaceted nature of ER and the variety of approaches Correspondingly, caregiver-initiated ER responding is used to study it. gradually replaced by self-initiated responding. Third, as There is general consensus that distress or dysphoria a normal part of development, children_s ER repertoires can elicit a variety of regulatory responses that can serve to become larger, come to include mostly functional but diminish or alleviate it, that multiple interactive processes also some dysfunctional responses, and are characterized are involved (or called upon) in any given ER response, by ER response deployment, which is both flexible and and that the nature of the response or its component context sensitive. Fourth, there are considerable indi- elements vary as a function of developmental stage. There vidual differences in ER responses to distress and dys- also is considerable agreement that the ability to refocus phoria, which are likely to represent a combination of attention away from the source of distress or dysphoria, innate physiology, history of ER response acquisition and or the associated feelings, represents a primary ER re- deployment, and personal resources. Finally, ER sponse, starting in early childhood, and that refocusing of responses can occur at various points along an emotion attention can be achieved through a variety of actions and experience and can be defined in various ways. behaviors. Although ER responses are typically initiated by caregivers for distressed infants and toddlers, children CERT and the Regulation of Dysphoria become increasingly able to self-regulate such negative In CERT, emotion regulation is defined as responses, emotions. For example, during infancy, dysphoria and which may serve to interrupt, lessen, diminish and ter- distress are modulated mostly through input from minate, or maintain, accentuate, and prolong the emotion. caregivers who achieve that goal by various actions (e.g., Adaptive or functional ER responses to dysphoria serve to gentle touching, coddling, rocking, attempts to distract the down-regulate it and therefore enable the child to feel infant, feeding). With the emergence of basic skills (e.g., better, whereas maladaptive or dysfunctional responses ambulation, motor coordination, language, attention exacerbate or prolong the negative mood (and thereby focusing), toddlers and young children gradually master engender other problems). Thus, one of the goals of a repertoire of responses that they can initiate and deploy CERT is to identify the specific ways in which a given child to modulate or terminate their own feelings of sadness or has historically responded to dysphoria or distress elicitors distress. Thus, a distressed young child can alleviate that in various contexts and the types of ER responses that have emotion by self-soothing, turning to play activity, leaving been useful. 894 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com NEW TREATMENT FOR CHILDHOOD DEPRESSION Within the framework of CERT, ER responses to another. On the contrary, the individually tailored dysphoria are grouped into four interconnected domains: approach of CERT accommodates the fact that a given biological, behavioral, cognitive, and social/interpersonal. ER response (e.g., provision of physical comfort) may This concept is presented to families as Bsources of emotion help one child to regain emotional homeostasis, but have regulation[ or Btypes of control processes[ that can help a the opposite effect on another child and that there are child to feel better. The Bbiological[ domain of ER is age-related differences in regulatory competencies (e.g., a defined as the physiological infrastructure underlying 7-year-old_s ER repertoire is less likely than that of a emotions and ER, or the Bphysical processes[ in one_s 11-year-old to contain adaptive cognitive responses). body, which also involves a child_s innate sensitivity and Third, because ER responding is shaped across develop- possibly high reactivity to emotion elicitors. The Bbe- ment through learning and practice, didactics and havioral[ regulatory domain is presented to families as rehearsal are key components of CERT. Bways of acting or what you do[ to manage or control Another salient feature of CERT is the emphasis on the dysphoria, including responses such as walking away from Bcontext[ of symptoms and ER responses. The a distressing situation, turning to play activity, doing Bcontextual mapping[ of a symptom or problem scenario physical exercise, playing music, doing chores, and working (i.e., constructing a detailed account of a distress-eliciting on a project. The Bcognitive[ regulatory domain is incident) enables the therapist to select a point of described to children as Bwhat goes on in your head[ intervention (for that given complaint), which may be the that can help to feel better, including conjuring up some ER response, some nonmood depressive symptom, or image to counter the dysphoric emotion, helpful Bself- environmental features. Throughout treatment, the talk,[ focusing attention on neutral or positive topics, and ongoing goals are to identify the given child_s habitual changing how you think about what makes you sad, such adaptive and maladaptive responses to persistent dys- as minimizing its significance. The Bsocial/interpersonal[ phoria and daily events that exacerbate it (the latter being ER domain, which entails enlisting other people as the typically exemplified by problem scenarios brought into means to down-regulate one_s own dysphoria, is presented sessions); pinpoint to families ER responding that up- as Bhow other people can affect you or help you to feel regulates (rather than modulates) or has no effect on the better.[ This domain includes both the types of inter- emotion or mood; reinforce or teach age-appropriate, actions whereby others can help modulate a given child_s adaptive, dysphoria-specific regulatory responses; and distress, such as the child getting a hug or being physically reframe the child_s difficulties using a general ER and comforted, talking to a parent or teacher about the coping perspective. dysphoria, and engaging a peer in play or some project, If the existing repertoire of adaptive ER responses to and effective ways of recruiting social regulators such as dysphoria is inadequate, then the child_s inclination the use of explicit language. Consistent with the im- toward one versus another ER domain is used to portance of context in the experience and regulation of remediate deficits. For example, a child who is social by emotions (e.g., Thompson, 1994), the treatment also con- nature would be taught adaptive ways of recruiting siders environmental manipulation and the use of external Bsocial regulators[ as one way to feel better when he or resources as (behavioral) regulatory responses. she feels sad. Alternatively, with a child who has a clear preference for thinking (but may not be socially inclined), Salient Features of CERT various cognitive ER responses would be explored. To The developmental ER perspective of CERT had facilitate adaptive ER response deployment, any number several implications for its implementation. First, because of strategies can be used including social modeling, ER is Bboth an individually based and a relational differential reinforcement, paired-associate learning, and phenomenon[ (Grolnick et al., 1996) and the use of depending on the child_s age and skills, didactics focusing social agents to regulate distress is a major ER domain on cognitive processes (i.e., alternative explanations, visual (Hofer, 1994), at least one caregiver had to actively imagery). If the child_s history reveals past deployment participate in the child_s treatment. Second, because of adaptive regulatory responses to dysphoria, then the developmental perspectives highlight individual differ- goal is to reinforce them and ensure a facilitative context. ences, no assumptions were made about one ER domain For example, if a child had been able to self-regulate or set of regulatory responses being more important than distress/dysphoria in the past by engaging in physical J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 895 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com KOVACS ET AL. activity, but recent environmental changes have con- understood, it is reiterated that under some conditions strained him or her, the therapist and the family together involving dysphoric emotion, solving the given problem identify alternative venues to enable deployment of that may be the most practical and useful approach. ER response. To ensure active parental participation in CERT, the parent is asked to serve as an Bassistant coach[ and is PRETRIAL VERIFICATION OF provided with six Brules of coaching[ (e.g., Bcoaches are SELECTED CONSTRUCTS made, not born,[ Beven a coach can have a bad day[). The rules of coaching are included in a parent_s manual Two samples of convenience served to explore aspects that summarizes basic CERT concepts. As an assistant of the explanatory framework and ER responses, namely coach, the parent not only helps the child but also tries 19 psychiatrically referred (PR) children with various on a new role and new ways of interacting, which were clinical diagnoses of depression (ages 7Y12 years; 63% envisioned to have positive effects both on the child_s male; 58% living in single-parent homes), and 10 normal ER skills and the parentYchild relationship. controls (NC) with no history of psychiatric illness or In managing depressive symptoms other than mood, treatment (ages 7Y12 years; 50% male; 40% living in the therapist can select from a broad array of focused, single-parent homes). Members of the Treatment problem solvingYoriented, behavioral, interpersonal, and Development Team, using a combination of a semi- cognitive techniques that have been widely described in structured questionnaire and an open-ended interview, the literature as practical and useful. These techniques individually interviewed the children about Bcauses[ of include teaching Bsleep hygiene[ as one way to counter and ways to alleviate dysphoria in youngsters their age sleep disturbance, keeping a log to identify temporal and about means of expressing such emotions to an features of symptoms such as fatigue, assigning pleasant adult. Both groups cited aversive peer behavior (being social activities to counter withdrawal, teaching visual teased, called names) as a salient cause of dysphoria imagery to counter negative thoughts about the self or (42% and 60%, respectively), but only the PR children others, or designing behavioral experiments to explore cited highly painful experiences (e.g., someone having new ways of responding. died [11 of 19], being left or put in foster care [5 of During Bcoping skills training,[ the final phase of 19]). Some social/interpersonal response was most CERT, age-appropriate didactic aids are employed to commonly cited as a way to modulate dysphoria, reinforce what has been learned about ER. For example, including talking to someone (63% and 40%, respec- the therapist and the family together view film segments tively). The PR children also generated ER responses depicting dysphoric emotions and the protagonists_ ER such as Bdoing something fun[ (6 of 19) and someone responses (e.g., precued sections from The Lion King, buying you something nice (3 of 19). Regarding Babe). A list of a priori defined questions and probes methods to express dysphoric feelings, rated from posed by the therapist help children review appropriate Beasy[ to Bdifficult,[ 70% of the NC children opined emotion-naming, recognize putative causes, and identify that Btalking about it[ is somewhere in the middle ER responses that were adaptive (helped the protagonist range, but only 37% of the PR children said so, and to feel better) and those that were not (made the another 37% stated that it was Bdifficult.[ By rank protagonist feel more distressed or dysphoric). ParentY order, both groups noted that talking about one_s child pairs also are helped to catalog dysphoria-eliciting emotion was the hardest, and drawing pictures of faces contexts that have been historically troublesome for the to show your emotion was relatively easy, but for the child, and how to prepare for them by the child_s PR children, selecting an emotion-expression face from a ongoing adaptive use of ER responses. The importance chart of faces was the easiest way to identify their emotion of contexts and strategies that help to initiate or maintain to an adult. According to both groups, keeping a written positive emotions also are discussed. In short, the notebook can help children track their daily emotions emphasis is on practicing adaptive regulatory responses and moods (63% of the PR group; 80% of the NC), but to dysphoria to be able to prevent such an emotion from only a minority stated that this would be easy (42% of Bgetting out of hand.[ In addition, by reviewing the PR; 20% of NC), citing various barriers (e.g., Bhe_ll general coping framework within which problems can be forget the paper[). 896 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com NEW TREATMENT FOR CHILDHOOD DEPRESSION In addition, during brief therapy with two psychi- CLINICAL TRIAL METHOD atrically referred children (a 9-year-old boy and an 11-year-old girl, both with 2-year histories of depressed Subjects mood and associated problems), we piloted enrolling Children had to meet the following study entry criteria: ages the parent as an Bassistant coach[ or Bcotherapist[ and between 7 and 12 years; no evidence of major systemic medical techniques such as the young patients keeping mood and disorder, mental retardation, or psychosis; presence of chronic depressive disorder; not receiving pharmacotherapy; and willingness activity logs, and involving parentYchild pairs in be- of at least one parent to actively participate in the treatment. havioral experiments. Parents were found to be most Recruitment efforts (at our child psychiatric outpatient referral units, receptive to intervention strategies that were developed in various other child mental health clinics, and via community advertisement) resulted in 101 initial phone screens; 36 children collaboration with them as Bco-therapists.[ The children then had a full evaluation; 20 met criteria and entered the treatment were able (with reminders) to complete mood logs that protocol. Treatment was provided free of charge. Families signed served to pinpoint diurnal variations (which, interestingly informed consents, and children provided assents. The sample of 20 was 35% female, had a mean age at entry of 10.4 years (SD = were found to differ from initial parental perceptions), 2.2 years), 90% were white, and 10% were African American or and activity logs that helped to identify behaviors that biracial; 90% lived in two-parent households and 80% lived in diminished dysphoria as well as parentYchild interactions intact families of origin. For 65% of the sample, educational that exacerbated it. attainment of the head of household was a college degree (or higher). Of the children, 50% had a history of outpatient mental health In summary, the informal survey suggested that treatment; none had a history of psychotropic medication use or children recognize the relationship between dysphoria inpatient hospitalization. and stressful events. Many of the ER responses they Chronic depression was defined as the diagnosis of DD according to either DSM-IV or DSM-III, in part because the validity of DSM-IV generated map onto those in CERT, with one exception DD criteria for youth has not yet been documented. Three cases with (traditional Bcognitive[ responses, e.g., thinking about clear Bdysthymia-like[ presentation (two were somewhat short of the the distress differently), were rare. However, this 1-year duration at intake and one with chronic illness but one symptom short) were formally diagnosed as depressive disorder not finding is age appropriate (M.K., unpublished manu- otherwise specified, but were considered for practical purposes as script). The results also suggested the advisability of having DD. At study entry, 16 subjects met criteria for both DSM-IV using visual aids to facilitate discourse about dysphoric and DSM-III DD and 4 met only DSM-III criteria. Duration of the emotion. All of these sources of information were used index episode was 24.4 months (SD = 11.9). Six children had a superimposed MDD at intake (two others had previous MDD to produce a working treatment manual, subsequent to episodes) and 70% had nonaffective (anxiety and oppositional which we initiated the clinical trial. defiant) comorbid disorders (Table 1). TABLE 1 Clinical Characteristics of the Sample at Four Time Points All Subjects at Intake Treatment Completers Variables (N = 20) Intake (n = 15) Tx End (n = 15) 6-Mo FU (n = 14) 12-Mo FU (n = 13) Diagnosis present (no., %) DD 20 (100) 15 (100) 7 (47) 3 (21) 1 (8) MDD 6 (30) 5 (33) 1 (7) 0 0 Anxiety disorder 8 (40) 7 (47) 5 (33) 3 (21) 1 (8) ODD 9 (45) 6 (40) 6 (40) 2 (14) 2 (15) Children_s self-rated symptoms (mean, SD) Depression (CDI) 16.3 (10.1) 16.6 (10.7) 4.3 (6.8) 5.0 (7.6) 5.3 (7.2) Anxiety (RCMAS) 15.0 (6.7) 15.3 (6.7) 4.7 (5.9) 4.5 (5.4) 5.3 (5.6) Mothers_ clinician-rated symptoms (mean, SD) Depression (HRS-D) 11.8 (6.5) 9.9 (5.6) 7.0 (3.9) 3.8 (3.4) 6.0 (3.2) Anxiety (HRS-A) 10.7 (5.0) 9.5 (4.5) 7.7 (3.2) 4.2 (3.3) 6.5 (5.4) Note: Results of statistical analyses are in the text. Tx = treatment; FU = follow-up; DD = dysthymic disorder; MDD = major depressive disorder; ODD = oppositional defiant disorder; CDI = Children_s Depression Inventory; RCMAS = Revised Children_s Manifest Anxiety Scale; HRS-D = Hamilton Rating Scale for Depression; HRS-A = Hamilton Rating Scale for Anxiety. J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 897 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com KOVACS ET AL. Assessment and Diagnostic Procedures treatment, follow-up evaluations included independent (but not blind) K-SADS interviews, and parent- and child-rated questionnaires. Psychiatric diagnoses at intake, end of treatment, and 6- and Participating were three therapists, whose psychotherapeutic 12-month follow-up were derived via the Schedule for Affective experiences ranged from 5 to 7 years (their highest academic degrees Disorders and Schizophrenia for School-Age Children (K-SADS; were Ph.D., M.D., and M.S.W.). The treatment room included Kaufman et al., 1997), which was administered by trained, in- visual aids that were age appropriate for children (e.g., wall poster dependent clinicians to the parent about the child and then separately depicting the regulatory domains as rooms of a house with pertinent to the child about him- or herself. The results were reviewed by at ER Btools[ noted in each ER room) and a poster that listed the least one other experienced clinician to ensure consensus diagnoses. At Brules of coaching.[ Each session involved the parentYchild pair. the end of treatment and follow-up evaluations, remission of a given Sessions were tape recorded and reviewed in weekly meetings by the disorder was defined operationally (Kovacs et al., 1984) as being free Treatment Development Team to ensure adherence to the basic of clinically significant and impairing symptoms (i.e., no more than explanatory framework, identify protocol features that needed to be one symptom and few, if any, subclinical symptoms of the particular clarified, and discuss any modifications that may be warranted. disorder remaining), and maintenance of this essentially asymptom- Intersession telephone contacts were allowed if needed to verify or atic state continuously for a minimum of 2 months. For behavior confirm some element of treatment. disorders, the minimum time frame for remission was at least 3 months. Partial remission was defined as no longer meeting full diagnostic criteria for the specific disorder but the persistence of a subsyndromal clinical picture. RESULTS During treatment, independent evaluators conducted monthly symptom assessments of the child via the Follow-up Depression and Characteristics of Completers Versus Noncompleters Anxiety Scales for Youths (FDS-Y, and FAS-Y, respectively). These clinician-rated scales were derived from the Interview Schedule for Of the 20 cases, 15 completed the protocol. Four were Children and Adolescents (Sherrill and Kovacs, 2000) and entail administratively terminated owing to too many missed standard probes regarding symptom frequency and intensity. Based sessions (defined a priori as three out of four consecutively on separate interviews with the child and with the parent about the child, the clinician provides an overall rating for each symptom scheduled sessions, without extenuating circumstances) ranging from 0 = none to 3 = severe (pervasive/constant, impairing/ and one dropped out (noncompleters were provided with disrupts functioning). The FDS-Y (26 items) has a potential score clinical referrals as needed). Completers averaged 30.6 range of 0 to 78; the FAS-Y (11 items) has a potential score range of (SD = 6.5) sessions across a mean interval of 12.2 months 0 to 33. Clinical evaluators were aware that the children were participating in a treatment protocol. (SD = 2.6); noncompleters averaged 6.2 (SD = 4.7) Children completed the Children_s Depression Inventory (CDI; sessions. Children who completed the protocol and those Kovacs, 2003) and the Revised Children_s Manifest Anxiety Scale who did not were demographically similar, with two (RCMAS; Reynolds and Richmond, 1985) about themselves at preset time points. Parents (typically mothers) completed the Beck exceptions: completers were more likely to be living in Depression Inventory (Beck et al., 1988) about themselves and intact families of origin (p = .03) and have better-educated other questionnaires and were evaluated by independent clinicians parents (p G .01). via the Hamilton Rating Scales for Depression (HRS-D; Hamilton, 1967a) and Anxiety (HRS-A; Hamilton, 1967b) at start of The two groups did not differ with regard to past treatment, its end, and at follow-up. Per the protocol, clinically treatment, diagnostic composition, and illness dura- depressed parents were referred for treatment, which was available tion at intake, rates of comorbid psychiatric diagnoses, gratis by prior arrangement. or severity of self-rated depressive and anxiety symp- toms. However, mothers of noncompleters versus com- pleters had higher levels of depression at study entry, Treatment Protocol both by self-rating (mean Beck Depression Inventory As an intervention for chronic depression, we designed a 30-session, score: 17.2, SD = 5.6 versus 7.7, SD = 6.4, respectively, multiphase, 10-month protocol as follows: phase IVmonth 1: intense p = .02) and clinical evaluation (mean HRS-D score: treatment, one to two sessions per week (up to eight sessions); phase 17.6, SD = 6.0 versus 9.9, SD = 4.5, respectively p = .03). IIVmonths 2Y 4: regular treatment, one session per week (up to 13 sessions); phase IIIVmonths 5 Y 6: tapered treatment, one session In the set of completers, three mothers and two fathers every 2 weeks (up to five sessions); phase IVVmonths 7Y10: were in treatment (for a variety of problems) when their maintenance treatment, one session/month (up to four sessions). children entered the protocol (two of these mothers The first three phases were limited to 26 sessions owing to past research (with adults) that suggested that by session 26, the vast majority of and one of these fathers continued their treatment dur- patients show some improvement (Howard et al., 1986). The ing CERT). Subsequent to the start of the protocol, maintenance phase (phase IV) was spaced across 4 months to provide two other mothers and two other fathers entered se- time for the consolidation of newly learned skills. The protocol parate treatment for themselves; of these, three were required evidence of symptomatic improvement at the 5- and 7-month points (defined via specific CDI cutoff scores for boys and girls) to self-initiated and one was via referral by the CERT advance a child to the respective phase. At 6 and 12 months after end of study coordinator. 898 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com NEW TREATMENT FOR CHILDHOOD DEPRESSION Changes in Children_s Diagnostic Status As an alternative approach, we used longitudinal At end of treatment, eight of the 15 completers modeling procedures (restricted maximum likelihood (53.3%) had full remission of the chronic depression models) on completers_ CDI and RCMAS scores at and two (13.3%) were in partial remission (by K-SADS). four time points: intake, end of treatment, and 6- and Five of these 10 patients also had had superimposed 12-month follow-up (Table 1). Similar to the results MDD, which was in full remission in four (80%). From from paired t tests, the results were highly significant among the seven patients with initial comorbid anxiety both for the model of depressive symptoms (F3,38 = 7.41 disorder, three were in full remission, but another patient p G .001) and the model of anxiety symptoms (F3,38 = developed an anxiety disorder postintake and was in 11.57 p G .001). episode at end of treatment (Table 1). All six patients We also examined clinician-rated symptom severity with oppositional defiant disorder continued to meet full patterns across the course of therapy at key treatment diagnostic criteria. transition points (1, 4, and 7 months and treatment Remission of dysthymia continued across the follow-up, end). Overall, repeated-measures analyses of variance with 79% of the available completers remitted by suggested significant changes across time in children_s the 6-month follow-up and 92% by the 12-month clinician-rated depression (FDS-Y ratings; F3,12 = 6.81, follow-up. For comorbid anxiety disorders, the cumu- p = .006) and anxiety (FAS-Y ratings; F3,12 = 7.49, lative rates of remission were 79% by the 6- and 92% p = .004). Profile contrasts, comparing evaluations at by the 12-month follow-up (Table 1). Mantel-Haenszel each given time point with those at the next time point, x2 analyses of the distribution of cases meeting full revealed that levels of depressive and anxiety symptoms diagnostic criteria at intake, end of treatment, and 6- and dropped significantly from the 1-month to the 4-month 12-month follow-up confirmed statistically significant assessments (F1,14 = 13.07, p = .003 for FDS-Y; F1,14 = decreases in the rates of DD (x2 = 27.85, p G .0001), 10.64, p = .006 for FAS-Y) and evidenced more gradual MDD (x2 = 10.26, p = .001), Bany depressive disorder[ and statistically insignificant declines from 4 to 7 months, (x2 = 27.85, p G .0001), and Bany anxiety disorder[ (x2 = and from 7 months to the end of treatment (F1,14 = 3.88 5.46, p = .02). and 0.04 for FDS-Y; and 0.92 and 0.05 for FAS-Y; all p > 0.05). Thus, the most dramatic decrements in symp- toms occurred during the first 4 months of the treatment. Changes in Children_s Symptom Severity Using clinically rated depressive symptoms, Figure 1 illus- Using the 20-patient intent-to-treat sample (and trates the changes during treatment relative to pretreat- carrying forward the last scores of noncompleters), ment levels. To generate the figure, the intake K-SADS we found that self-rated depressive symptoms (CDI) de- (current ratings) was used to estimate symptom severity clined from being in the clinical range at entry (mean = at entry by summing across 22 items that had counter- 16.3, SD = 10.1) to normal levels (mean = 5.2, SD = 6.7) parts in the FDS-Y. In turn, the relevant FDS-Y items by end of treatment (paired t test = 4.81, p G .001). were rescaled onto a 3-point scale. There was a parallel, statistically significant pretreatment (mean = 15.0, SD = 6.7) to posttreatment (mean = 6.7, SD = 7.8) decrease in levels of anxious symptoms (paired Changes in Maternal Symptoms t test = 5.48, p G .001) as quantified by the RCMAS. The We used longitudinal statistical procedures (restricted findings were similar for the 15 completers (Table 1): maximum likelihood models) to examine data from the there were highly significant pre/post changes in their clinical evaluations of mothers of completers (HRS-D levels of depression (CDI scores, paired t test = 4.45, and HRS-A scores; Table 1) at intake, end of treatment, p G .001) and anxiety (RCMAS scores, paired t test = and 6- and 12-month follow-up. According to the results, 6.97, p G .001). Treatment gains were maintained across levels of maternal depression (HRS-D scores) and anxiety time, as suggested by comparing end-of-treatment scores (HRS-A scores) significantly diminished across time to those at the 6- and 12-month follow-up for the CDI (F3,31 = 4.71, p = .008, and F3,32 = 3.90 p = .018, (t = j0.34, and t = j0.14, respectively, both not respectively). However, because four mothers had been significant) and the RCMAS (t = 0.58 and t = j0.64, receiving their own treatment separately from the CERT respectively, both not significant). protocol, we reran the analyses using only the remaining J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 899 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com KOVACS ET AL. DISCUSSION CERT is a new intervention for childhood depres- sion that provides therapists and families with an ex- planatory framework that is conceptually meaningful, clinically sensible, and developmentally appropriate. It builds on emotion self-regulation and focuses on self- regulatory responses to distress and dysphoria, which unfold across development, are important for adaptive functioning, and appear to be dysfunctional in individuals with depressive disorders across the life span. It highlights the importance of parental involve- ment in treatment and offers one way to facilitate it. Because it incorporates commonly used intervention techniques and is manual based, CERT can be taught to therapists. The generally positive overall results of this pilot study underscore the feasibility of our approach and suggest the need for a larger, randomized trial of the efficacy of CERT. The importance of adaptive self-regulation of dys- Fig. 1 Clinician-rated severity of children_s depressive symptoms during phoria and distress, which is the ultimate target of CERT, treatment. Error bars represent the 95% upper confidence interval of the mean. is supported by extensive developmental research address- ing the years of childhood (Fox, 1994; Kopp, 1989) and ER skills remain critical across the life span (Gross, 1998). There is evidence that clinically depressed youngsters have 11 mothers. In this subgroup, changes in levels of maternal fewer and less effective ER skills for managing negative symptoms from intake and through the follow-up were emotions than do their nonaffected peers (Garber et al., not significant (HRS-D: F3,24 = 1.99, p = .143; and HRS- 1995). Because dysregulated affect is a salient feature of A: F3,25 = 2.24, p = .109). depressive disorders, all empirically supported treatments of depression have components that target aspects of emotion regulation. However, cognitive-behavioral inter- Parental Treatment and Changes in Children_s ventions and interpersonal psychotherapy, which have Symptom Severity been specifically used with depressed youngsters (e.g., Can parents_ own treatment account for the signifi- Brent et al., 1997; Mufson et al., 1999; Stark et al., 1987) cant symptomatic improvement of the children? To represent Bdownward extensions[ of treatments that were examine this issue, we reanalyzed children_s depressive originally designed for adults. Furthermore, none of them and anxiety symptom scores, using only cases whose emphasizes ER as the key organizing principle in its mothers were not receiving psychiatric treatment during delivery or case formulation, defines ER from a develop- the CERT study. The improvement in symptoms of mental perspective, takes into account the diverse ways in depression and anxiety in this subgroup of young which dysphoria can be regulated, or strategically engages patients remained statistically significant across the four the parent as an emotion regulatory agent and treatment time points (restricted maximum likelihood model on ally. In contrast, by viewing dysregulated mood as the key 11 subjects CDI: F3,28 = 6.73, p = .002; and RMAS: feature of depressive disorders and highlighting that ER F3,27 = 11.13, p G .001). Significant symptom im- responses are developmentally acquired, CERT guides provement also was observed for the subset of cases clinicians to focus on a child_s history of ER response who did not have either parent in treatment during acquisition and deployment. By acknowledging the di- the CERT protocol (n = 8; CDI: F3,20 = 4.88, p = .011; versity of responses that serve to down-regulate dysphoria and RMAS: F3,19 = 4.20, p = .019). and individual (including age-associated) differences in 900 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com NEW TREATMENT FOR CHILDHOOD DEPRESSION that regard, CERT makes it possible to match the delivery (e.g., Brent et al., 1997). Thus, our study is among the of the treatment to a given child_s needs and competen- first to enlist parents in each session of an intervention, cies. By organizing ER responses into meaningful do- with a clearly defined role that is focused on the mains, CERT provides one useful way to characterize ER depressed child_s ER needs. repertoires for distress management. Several preliminary findings are worthy of note. First, Although partly for practical reasons, CERT_s presen- at the end of treatment, the rate of remission of super- tation of ER, ER domains and responses, and contextual imposed MDD (80%) was high, suggesting the useful- factors that affect them is somewhat simplified, its ness of CERT for nonchronic depression. Second, the emphasis is supported by recent research directions on finding that the rate of remission of DD was 53% at mood disorders (Davidson et al., 2002). Furthermore, end of treatment and that it took an additional 6 to some of the initial premises that guided CERT have been 12 months for another 40% of cases to achieve full extended into a more comprehensive model of vulner- remission underscore the recalcitrant nature of chronic ability to juvenile-onset depression (M.K., unpublished depression. This finding also may partly mirror the fact manuscript). In the extended model, the Bbiological[ that our patients already had illness durations somewhat regulatory domain is redefined as the neurophysiological longer than 2 years, on average, at study entry. Survival underpinning of ER, with an emphasis on relevant brain analyses of time to recovery from depression generally cortical and subcortical circuits. A fourth (Bsomatic- suggest that the probability of recovery in any given time sensory[) regulatory domain has been defined to interval is typically higher earlier in the illness and accommodate ER responses that rely on physical-sensory becomes less likely as time goes by (e.g., Kovacs et al., modalities to self-regulate dysphoria. The extended model 1994); in other words, at the start, we had a sample with a acknowledges more fully the relations of ER responding low likelihood of recovery. Notably, even our lengthy and brain Bemotion circuits[ and that individual differ- protocol did not accommodate the needs of about 20% ences in brain circuitry may contribute to why some of the youths, whose treatments were somewhat extended children find it more difficult than others to down- during the tapering or maintenance phases. Third, al- regulate dysphoria, and why a given ER response may though comorbid disorders were not targeted by CERT, Bwork[ for some children but not for others. the somewhat higher rate of recovery from anxiety Given that the field of psychotherapy for depressed disorders by the end of treatment (two of seven) children has been generally bereft of developmentally compared to that from oppositional defiant disorders based interventions (Weisz et al., 1992), it is hoped that (zero of six) is worthy of note. This may possibly suggest CERT represents only one of several new initiatives to that the ER responses and problem-solving approaches design treatments for that age group. Results of this considered in CERT may have some relevance to con- open trial provide initial evidence of the feasibility and ditions that have dysregulated mood, such as persistent apparent clinical efficacy of CERT for chronic depres- anxiety, as a salient feature. This conclusion must be sion in childhood. The finding that 75% of the enrolled viewed with caution owing to the small overall sample families completed this lengthy protocol suggests its size and the number of those with comorbid disorders. acceptability. It is notable that the retention rate across Underscoring the importance of monitoring parental our 10-month intervention was comparable to that psychopathology among clinically referred depressed reported for the 12-week, open, pilot trial of Inter- youngsters (Ferro et al., 2000; Hammen et al., 1999), personal Psychotherapy for Adolescents (79%; 11 of depression was a notable feature of noncompleter 14 patients; Mufson et al., 1994). mothers, in spite of the availability of free treatment for Parental involvement is a well-established component parents (separate from the CERT study). Other char- of evidence-based interventions for children with acteristics of noncompleter mothers (e.g., lower level of disruptive behavior disorders (see Chorpita et al., 2005) education, more likely to be a single parent) echo and increasingly so for those with anxiety disorders (e.g., findings from various treatment studies (e.g., Curry, Ginsburg et al., 2004). In contrast, the involvement of 2005) that socioeconomic resources of families affect parents in manual-based treatment trials for depressed children_s treatment responses. Ways in which CERT youths has entailed either adjunctive group sessions (e.g., can be modified to accommodate parents with sub- Clarke et al., 1999) or systemic family therapy optimal resources certainly warrants attention. J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 901 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com KOVACS ET AL. In most research trials, decisions about length of at least a 4-month treatment period as advisable for treatment, which is typically brief, appear to be guided children with chronic depression. by considerations other than clinical need (Kovacs and Sherrill, 2001). This presents a challenge when patients Disclosure: The Children_s Depression Inventory (CDI) is published by suffer from a chronic disorder and when an intervention Multi-Health Systems, Inc., for which Dr. Kovacs receives royalties. The seeks to apply a developmentally based paradigm other authors have no financial relationships to disclose. because both imply the need for time. However, it can be argued that the lengthy protocol we piloted is REFERENCES likely to be above and beyond the resources of most Beck AT, Steer RA, Garbin MG (1988), Psychometric properties of the Beck clinical settings. Therefore, it is noteworthy that the Depression Inventory: twenty-five years of evaluation. Clin Psychol Rev 8:77Y100 most dramatic reduction in depressive symptoms in our Bradley SJ (2000), Affect Regulation and the Development of Psychopathology. sample occurred during the first 4 months of CERT, New York: The Guilford Press suggesting that 4 months could be a Bcompromise Brent DA, Holder D, Kolko D et al. (1997), A clinical psychotherapy trial for adolescent depression comparing cognitive, family, and supportive duration[ for treating chronic depression. However, therapy. Arch Gen Psychiatry 54:877Y885 this finding also could signify that treatment gains from Casey BJ, Giedd JN, Thomas KM (2000), Structural and functional brain development and its relation to cognitive development. Biol Psychol 54: less frequent sessions (months 5Y10) require more time. 241Y257 In any case, CERT is sufficiently flexible to accommo- Chorpita BF, Daleiden EL, Weisz JR (2005), Identifying and selecting the common elements of evidence based interventions: a distillation and date various protocol lengths. In addition, along with matching model. Ment Health Serv Res 7:5Y20 the explanatory framework, the coping skills module of Clarke GN, Rohde P, Lewinsohn PM, Hops H, Seeley JR (1999), Cognitive-behavioral treatment of adolescent depression: efficacy of CERT can be Buncoupled,[ and with minor modifica- acute group treatment and booster sessions. J Am Acad Child Adolesc tions, used in prevention trials. Psychiatry 38:272Y279 Compton SN, March JS, Brent D, Albano AMV, Weersing R, Curry J (2004), Cognitive-behavioral psychotherapy for anxiety and depressive Limitations disorders in children and adolescents: an evidence-based medicine By its very nature, an open pilot clinical trial such as review. J Am Acad Child Adolesc Psychiatry 43:930Y959 Curry JF (2001), Specific psychotherapies for childhood and adolescent ours is limited by the small sample size, which depression. Biol Psychiatry 49:1091Y1100 constrains its generalizability, and the unfeasibility of Curry JF (2005), Predictors and moderators of acute treatment outcome in TADS. Presented at the Joint Annual Meeting of the American Academy blind clinical evaluations, which may introduce rater of Child and Adolescent Psychiatry, Toronto, October 18Y23, 2005 bias. We sought to counter one other source of limi- Davidson RJ, Pizzagalli D, Nitschke JB, PutnamK (2002), Depression: perspectives from affective neuroscience. Annu Rev Psychol 53: tation in trials of this sort, namely, a single therapist 545 Y574 who designed the treatment, by having three therapists, Durston S, Pol HEH, Casey BJ, Giedd JN, Buitelaar JK, van Engeland H not including the originator of the intervention (M.K.). (2001), Anatomical MRI of the developing human brain: what have we learned? J Am Acad Child Adolesc Psychiatry 40:1012Y1020 Furthermore, owing to the absence of a comparison Ferro T, Verdeli H, Pierre F, Weissman MM (2000), Screening for treatment, we cannot unequivocally state that CERT depression in mothers bringing their offspring for evaluation or treatment of depression. Am J Psychiatry 157:375 Y379 was the agent responsible for our patients_ improve- Fox NA ed. (1994), The development of emotion regulation: biological and ment and remission. Finally, there were no Hispanic behavioral considerations. Monogr Soc Res Child Dev 59(2Y3 Serial No. 240) Garber J, Braafladt N, Weiss B (1995), Affect regulation in depressed and families in our sample. Thus, future studies need to nondepressed children and young adolescents. Dev Psychopathol 7: examine the usefulness of CERT with children who had 93 Y115 Garber J, Dodge KA, eds (1991), The Development of Emotion Regulation and had culturally diverse ER histories. Dysregulation. New York: Cambridge University Press Ginsburg GS, Siqueland L, Masia-Warner C, Hedtke KA (2004), Anxiety Clinical Implications disorders in children: family matters. Cogn Behav Pract 11:28Y43 Grolnick WS, Bridges LJ, Connell JP (1996), Emotion regulation in two- The chronicity of childhood DD, the prevalence of year-olds: strategies and emotional expression in four contexts. Child Dev comorbid diagnoses, and the likelihood of parental 67:928 Y941 Gross JJ (1998), The emerging field of emotion regulation: an integrative emotional distress present treatment challenges. How- review. Rev Gen Psychol 2:271Y299 ever, the preliminary indications are that CERT, with Hamilton M (1967a), Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 6:278Y296 its individually tailored approach and focus on ER, is Hamilton M (1967b), Diagnosis and rating of anxiety. Br J Psychiatry 3: well accepted by families and may help children to 76 Y79 Hammen C, Rudolph K, Weisz J, Rao U, Burge D (1999), The context of recover, although it does require commitment from the depression in clinic-referred youth: neglected areas in treatment. J Am family and the clinician. Clinicians may wish to consider Acad Child Adolesc Psychiatry 38:64 Y71 902 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited. www.cuwai.com NEW TREATMENT FOR CHILDHOOD DEPRESSION Hofer MA (1994), Hidden regulators in attachment, separation, and loss. Developmental and Clinical Perspectives, Goodyer IM, ed. New York: Monogr Soc Res Child Dev 59:192Y207 Cambridge University Press Howard KI, Kopta SM, Krause MS, Orlinsky DE (1986), The dose-effect Michael KD, Crowley SL (2002), How effective are treatments for child and relationship in psychotherapy. Am Psychol 41:159 Y164 adolescent depression? A meta-analytic review. Clin Psychol Rev 22:247Y269 Kaufman J, Birmaher B, Brent D et al. (1997), Schedule for affective disorders Mufson L, Moreau D, Weissman MM, Wickramaratne P, Martin J, and schizophrenia for school-age childrenVpresent and lifetime version Samoilov A (1994), Modification of interpersonal psychotherapy with (K-SADS-PL): initial reliability and validity data. J Am Acad Child Adolesc depressed adolescents (IPT-A): phase I and II studies. J Am Acad Child Psychiatry 36:980Y988 Adolesc Psychiatry 33:695 Y705 Kendall PC, Choudhury MS (2003), Children and adolescents in cognitive- Mufson L, Weissmann M, Moreau D, Garfinkel R (1999), Efficacy of behavioral therapy: some past efforts and current advances, and the interpersonal psychotherapy for depressed adolescents. Arch Gen challenges in our future. Cogn Ther Res 27:89Y104 Psychiatry 56:573Y579 Kopp CB (1989), Regulation of distress and negative emotions: a de- Reynolds CR, Richmond BO (1985), Revised Children_s Manifest Anxiety velopmental view. Dev Psychol 25:343Y354 Scale (RCMAS): Manual. Los Angeles: Western Psychological Services Kovacs M (2003), Children_s Depression Inventory (CDI): Technical Manual Sherrill JT, Kovacs M (2000), The Interview Schedule for Children and Update. North Tonawanda, NY: Multi-Health Systems Adolescents (ISCA). J Am Acad Child Adolesc Psychiatry 39:67Y75 Kovacs M, Akiskal HS, Gatsonis C, Parrone PL (1994), Childhood-onset Stark KD, Reynolds WM, Kaslow NJ (1987), A comparison of the relative dysthymic disorder: Clinical features and prospective naturalistic efficacy of self-control therapy and a behavioral problem-solving therapy outcome. Arch Gen Psychiatry 51:365Y374 for depression in children. JAbnorm Child Psychol 15:91Y113 Kovacs M, Feinberg TL, Crouse-Novak MA, Paulauskas SL, Finkelstein R Thompson RA (1994), Emotion regulation: a theme in search of definition. (1984), Depressive disorders in childhood: I. A longitudinal prospective Monogr Soc Res Child Dev 59:25Y52 study of characteristics and recovery. Arch Gen Psychiatry 41:229Y237 Weisz JR, Rudolph KD, Granger DA, Sweeney L (1992), Cognition, Kovacs M, Obrosky DS, Gatsonis C, Richards C (1997), First-episode major competence, and coping in child and adolescent depression: research findings, depressive and dysthymic disorder in childhood: clinical and socio- developmental concerns, therapeutic implications. Dev Psychopathol demographic factors in recovery. J Am Acad Child Adolesc Psychiatry 4:627Y 653 36:777Y784 Weisz JR, Thurber CA, Sweeney L, Profit VD, LeGagnoux GL (1997), Brief Kovacs M, Sherrill JT (2001), The psychotherapeutic management of treatment of mild-to-moderate child depression using primary and major depressive and dysthymic disorders in childhood and adoles- secondary control enhancement training. J Consult Clin Psychol 65: cence: issues and prospects. In: The Depressed Child and Adolescent: 703Y707 Scaling Up Promising Interventions: Feasibility of Screening Adolescents for Suicide Risk in BReal-World[ High School Settings Denise Hallfors, PhD, Paul H. Brodish, MSPH, Shereen Khatapoush, PhD, Victoria Sanchez, DrPH, Hyunsan Cho, PhD, Allan Steckler, PhD Objectives: We evaluated the feasibility of a population-based approach to preventing adolescent suicide. Methods: A total of 1323 students in 10 high schools completed the Suicide Risk Screen. Screening results, student follow-up, staff feedback, and school responses were assessed. Results: Overall, 29% of the participants were rated as at risk of suicide. As a result of this overwhelming percentage, school staffs chose to discontinue the screening after 2 semesters. In further analyses, about half of the students identified were deemed at high risk on the basis of high levels of depression, suicidal ideation, or suicidal behavior. Priority rankings evidenced good construct validity on correlates such as drug use, hopelessness, and perceived family support. Conclusions: A simpler, more specific screening instrument than the Suicide Risk Screen would identify approximately 11% of urban high school youths for assessment, offering high school officials an important opportunity to identify young people at the greatest levels of need and to target scarce health resources. Our experiences from this study show that lack of feasibility testing greatly contributes to the gap between science and practice. American Journal of Public Health 2006; 96(2):282Y287. J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:8, AUGUST 2006 903 Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.