Nonpharmacological Response in Hospitalized Children With Conduct Disorder

Nonpharmacological Response in Hospitalized Children With Conduct Disorder

Nonpharmacological Response in Hospitalized Children With Conduct Disorder RICHARD P. MALONE, M.D., JAMES F. LUEBBERT, M.D., MARY ANNE DELANEY, M.D., ...

5MB Sizes 0 Downloads 44 Views

Nonpharmacological Response in Hospitalized Children With Conduct Disorder RICHARD P. MALONE, M.D., JAMES F. LUEBBERT, M.D., MARY ANNE DELANEY, M.D., KRISTA A. BIESECKER, B.A., BRIDGET L. BLANEY, B.A., AMY B. ROWAN, M.D., AND MAGDA CAMPBELL, M.D.

ABSTRACT ObJective: There is a paucity of research regarding the effects of hospitalization and/or the response to placebo in

children with conduct disorder who are hospitalized for chronic and severe aggression. However, many children with this problem are hospitalized and immediately begin pharmacotherapy. In this report, the effects of hospitalization and placebo administration were examined. Method: Subjects were forty-four children (37 males, 7 females) with conduct disorder, aged 9.83 to 17.14 years, who were hospitalized for chronic and severe aggression. This was a 4-week double-blind and placebo-controlled study with a 2-week single-blind placebo lead-in period. During the 2-week placebo baseline period, aggression was measured on a 24-hour basis, using the Overt Aggression Scale. Only subjects meeting a specific aggression criterion were randomized to the treatment period of the trial. Results: Of the 44 subjects enrolled, 23 (52.3%) met the aggression criteria for entering the treatment period (baseline nonresponders), while 21 (47.7%) did not (baseline responders). Thus, almost half of the subjects, while taking no active medication, benefited from the inpatient milieu/structure and/or placebo. Conclusion: This finding has important treatment and research implications. Medication to treat aggression should not be initiated immediately upon hospitalization because improvements associated with hospitalization may be attributed inaccurately to pharmacotherapy, resulting in unnecessarily medicating children. A placebo baseline period is essential to decrease the risk of a type II error in pharmacological research concerning aggression. J. Am. Acad. Child Ado/esc. Psychiatry, 1997, 36(2):242-247. Key Words: conduct disorder, aggressive behavior, child, hospitalization, pharmacological treatments, placebo.

Aggressive behavior in children and adolescents is a major public health concern (U.S. Department of Health and Human Services, I99I) and is one of the most common reasons for referral to psychiatric ser-

Acaptrd [une 26. 1996. Drs. Malone, Luebbert, Delaney, and Rowan and Ms. Biesecker and Ms. Blaney art with the Department ofPsychiatry. Medica] Colugt ofPennsylvania and Habnemann University, Philadelphia. Dr. Campbell is with the Department of Psychiatry. Neui York University Medical Center, Neu/ York. Part of this papl' was presented at th« 35th Annual Muting of tht Neui Clinical Drug Evaluation Unit. Orlando. FL. May 31 to [unr 3. 1995. This work was supported in part by USPHS grant MH-00979 (R.P.M) and th« Allegheny-Singer Research Institute (R.P.M.). The authors thank SchoolofMedicin«for his htlp in rtvitwing Gtorgt M Simpson. MD.. of this manuscript. and the staff of the Child and Adolescent Inpatient Service of Hahnemann University Hospital. 16 South. Reprint rtqutsts to Dr. Malone, Drpartment ofPsychiatry. Eastern Pennsyluania Psychiatric Institute, 3200 Henry Avtnut. Philadelphia, PA 19129; fax: (215) 849-6673; e-mail: malonl'@alltghtny.tdu

usc

0890-8567/97/3602-0242$03.00/0©1997 by the American Academy of Child and Adolescent Psychiatry.

242

vices. Although a number of approaches for reducing aggressivebehavior, such as psychopharmacological and behavioral, have been studied (Campbell et al., 1982, 1984, 1995; for review see Kazdin, 1987; Werry and Wollersheim, 1989), there are few treatments that are generally accepted as effective. Clearly, there remains a need to develop and test effective treatments aimed at decreasing aggressive behavior. Psychotropic medications are administered clinically, often for prolonged periods of time, to children and adolescents who are aggressive. To date, there have been a limited number of controlled trials of the safety and efficacy of psychotropic medication in the treatment of aggressive behavior in conduct disorder (Campbell et al., 1992). This is not surprising because, except for trials of stimulants in the treatment of attention deficit disorder, there are few controlled clinical trials for most psychiatric disorders in children (Jensen et al., 1994). There is pressure from many sources to decrease the rate of psychiatric hospitalization and the length of

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY. 36:2. FEBRUARY 1997

PHARMACOTHERAPY AND HOSPI TALIZA TION

stay for those patients who are hospitalized. While this pressure is not new, there is a perception along with suggestive evidence that this pressure will lead to increased and, by implication, unnecessary administration of medication to patients (Carroll et al., 1980; Pi et al., 1983). In all likelihood, this will be accompanied by increased concern in a number of quarters that children and adolescents are being medicated unnecessarily. While there is a pauciry of research regarding the effect of hospitalization on psychiatric patients, evidence from the literature on adult patients suggests that hospitalization can be therapeutic, even with seriously disturbed psychotic patients (Pi et al., 1983). The effect of psychiatric hospitalization on the aggressive behavior demonstrated in children and adolescents remains an understudied area, though it is clinically important becau se aggressive behavior is often the reason for hospitalization. If hospiralizarion itself greatly affects the rate of aggressive behavior, then it will be a confounding variable for any judgments made about the indications for and benefits of other interventions, such as psychotropic medication. The purpose of this article is to report on findings regarding aggressive behavior demonstrated during the placebo baseline period of a double-blind and placebocontrolled study of lithium for the treatment of aggressive behavior in children with conduct disorder. All subjects were inpatients who had been hospitalized with histories of uncontrolled aggressive behavior and poor response to other treatments. The history of severe aggressive behavior in these children and adolescents with conduct disorder was thought to justify a trial of psychotropic medication. METHOD Setting The setting for this stud y was the Ch ild and Adolescent Psychiatry Inpati ent Service of Hahnemann Universiry Hospital. This is an acute care unit in a tertiary care hospit al. which serves as a major teaching hospital for the Medical College of Pennsylvania and Hahnemann Universiry Program.

Subjects The subjects were 44 child ren and adolescents (37 males and 7 females) whose ages ranged from 9.83 to 17.4 years (mean = 12.62 :!: 1.93) and who met DSM-II/-R (American Psychiatric Association . 1987) criteria for conduct disorder. Ethni c distribution was as follows: African-American, 43.2%; Caucasian , 25%; His-

panic. 25%; and other ethnic groups , 6.8%. Socioeconomic status was as follows: I, 0%; II, 4.5 %; III, 6.8%; IV, 15.9%; V, 68 .2% ; and undetermined, 4.5 % (H ollingshead. 1957). None met DSMII/-R criteria for mental retardation, major depressive disorder, mania. recent drug dependence, or any psychotic disorders including schizophrenia or schizophreniform disorder . None had any major med ical problem s such as seizure disorder or renal, cardiac, or thyroid disorders. nor did they have any contraindications for lithium.

Procedure This is an ongoing study. The study was approved by the institutional review board of Hahnemann Universiry. Consent was obta ined from the guard ian and assent from the child before participation in the study. All subjects had been admitted to the hospital with histories of severe and chronic aggressive behavior. They were enrolled in a double-blind and placebo-controlled clinical trial of lithium for the treatment of aggressive behavior in children with conduct disorder. After a 2-week single-blind placebo baseline period, subjects who met a threshold level for aggression (described below) entered a 4-week double-blind and placebo-controlled trial and were randomized to lithium or placebo. Those who did not display or had reduced aggressive behavior were not random ized. A variety of instruments were used, including the Overt Aggression Scale (OAS) (Yudofsky et al., 1986); the Child Behavior Checklist (Achenbach , 1991a); the Youth Self-Report (Achenbach, 1991b); the Conners Teachers Questionnaire (Psychophannacology Bulletin, 1973); the Inattentive Overactivity With Aggression Conners Rating Scale (Loney and Milich . 1982); the Clinical Global Impressions (G uy, 1976); the Glob al Clinical Consensus Rating (Campbell et al., 1984. 1995); the Prior Medicat ion Record (Psychopharmacology Bulletin, 1973) ; and the Diagnostic Interview for Children and Adolescents-Revised (DICA) (Reich and Welner. 1990). Because this report concerns aggressive behavior demonstrated during the 2-week baseline period, the OAS data will be presented, as the OAS was the measure of aggressive behavior.

Overt Aggress ion Scale The OAS was developed by Yudofsky and associates (1986 ) to measure aggressive behavior in both adult and child psychiatric inpat ients. This measure has been used in a number of drug treatment studies of aggressive children and adolescents (Cueva et al., 1996; Kafantaris er al., 1992; Malone er al., 1994b; Rifkin et al., 1989). The OAS measures the frequency and severiry of aggressive behavior. Aggressive behavior is divided into four types: (I ) aggression against self, (2) aggression against others , (3) aggression against objects, and (4) verbal aggression. Within each type of aggressivebehavior . severity is indicated by a number of increasingly serious criteria . All ward staff were trained in the use of this measure at the beginning of the study, and new staff members were trained as part of their orientation to the ward. Each subject was monitored for aggressive behavior 24 hours a day, 7 days a week, by the ward staff, which includes nurses, activity workers, child care workers, social workers, and physicians. The frequency and severity scores on the OAS ratings during the placebo baseline period determined whether or not a subject had met the aggressive behavior criterion to enter the treatment period and be random ized to lithium or placebo. The threshold aggressive behavior criterion for random ization was to have three episodes of aggressive behavior per week, two of which included physical aggression, and to have

J. AM . AC AD . C H I LD AD OL ESC. PSYCHI ATR Y. 36 : 2. FE BRUA RY 19 9 7

243

MALONE ET AL.

a mean total severity score of 18 per week (for scoring see Silver and Yudofsky, 1987).

Analysis Independent t tests were used to investigate differences between placebo baseline responders and nonresponders for demographic variables including age, sex, race, and socioeconomic status, as well as for differences in baseline GAS scores and duration of aggression and conduct disorder. Chi-square was used to test for differences between the same groups for categorical variables regarding history of past treatment. Repeated-measures analysis of variance was used to test whether the groups differed in weekly GAS scores over the baseline period.

RESULTS Subjects

In all, 44 subjects were enrolled in the study, Only 23 (52.3%) of these 44 subjects met the aggressive behavior criterion to enter the 4-week treatment period (baseline nonresponders), while 21 (47.7%) failed to meet the aggression criteria, not displaying sufficient frequency and severity of aggressive behavior (baseline responders). Baseline responders and nonresponders were not significantly different in age, sex, race, IQ, socioeconomic group, or histoty of prior treatment (Table I). Data from the parent version of the OleA were used to investigate differences among subjects in the chronicity of aggressive behavior and conduct disorder (Table 2). Baseline nonresponders had a significantly longer history of aggressive behavior than did baseline responders (4.94 ± 2.88 versus 3.13 ± 1.67 years; t = -2.88, df = 27.78, P = .03), but the groups did not differ in the duration of conduct disorder (4.14 ± 2.52 versus 3.22 ± 1.73 years).

Overt Aggression Scale

The OAS severity scores and incident frequency scores are presented (Table 3). The baseline nonresponders had a total mean severity score of 33.7 ± 12/week, while the baseline responders had a mean score of 8.6 ± 6.4/week. The difference between these scores was significant (t = -8.76, df = 34.08, P < .001). The baseline nonresponders had a mean of 7.48 ± 2.5 incidents of aggression per week, while the baseline responders had a mean of 2.36 ± 1.6 incidents of aggression per week. The difference between these scores was significant (t = -7.97, df= 42, P < .001). Likewise, the two groups showed significant differences in various categories of aggressive incidents: incidents of physical aggression (4.04 ± 1.6 versus 1.02 ± 0.8; t= -7.94, df= 31.8, P < .001), incidents with severity at or above the second level (2.39 ± 1.4 versus 0.74 ± 0.56; t= -5.31, df= 29.7, P < .001), and incidents of verbal aggression alone (3.43 ± 1.4 versus 1.33 ± 1.1; t = - 5.52, df = 42, P < .001). Thus, the baseline nonresponders had more severe aggression and more aggressive acts than did the baseline responders. There was no significant effect for time; that is, the number of incidents and the severity of aggression remained stable over the 2-week baseline period in both groups. Likewise, there was no interaction between group and time.

DISCUSSION

While aggressive behavior is one of the most common reasons for psychiatric hospitalization for children with conduct disorder, there has been little systematic study

TABLE 1 History of Prior Treatment Baseline

Baseline Responders (n = 21)

Nonresponders (n = 23)

Treatment Prior hospitalizations" Pharmacotherapy"

Behavior modification b Group psychotherapy" Individual psychotherapy" Remedial education"

n

0/0

12 16 9 9 22 17

52.2 69.6 39.1 39.1 95.7 73.9

n

0/0

X'

P

6

28.6 47.6 28.6 19.0 85.7 71.4

2.53 2.19 0.54 2.13 I.31 0.03

.11 .14 .46 .14 .25 .85

lO

6 4 18 15

" By history. b From Prior Medication Record.

244

J.

AM. ACAD. CHILD ADOLESC. PSYCHIATRY. j6:2, FEBRUARY 1997

PHAR MA C O THERAPY AND HO SPITALIZATION

TABLE 2 C hronicity of Aggressive Behavior and Conduct Disorder Duration of Histo ry (Years) Variable Aggressive behavior" Co nduct disorder "

Baseline Nonrespond ers

Baseline Responde rs

(n =23)

(n=21 )

Statistic (t )

p

4.94 ± 2.88 4. 14 ± 2.52

3.13 :!: 1.67 3.22 ± 1.73

-2.28 -1 .31

.03 .20

" Data from Diagnosti c Interv iew for C hildren and Adolescents-Parent version, item 64D . " Data from Diagnostic Interv iew for Ch ildren and Adolescent s-Parent version . item 74B.

of the effect of ho spitalization itself on aggressiveness. In addition. knowledge is limited concerning the impact of placebo treatment in this population. Other stu d ies have shown th at children and adolescents have a high rate of placebo response in other disorders, such as ma jor depression (for review see Ambrosini er al., 1993) and aut ism (Ernst et al., 1990; Locascio et al., 1991). The results of th is study suggest that many children who are hospital ized for serious aggressive behavior may benefit by remission of aggressive beha vior. All subj ects who met criteria for initial enrollment in the study were exposed to the same hospital milieu and were ad ministered placebo med ication while baseline measur es were obtained. Only 23 (52.3%) of the 44 subjects enrolled demonstr ated sufficient aggressive beh avior to continue into the active treatment period of the trial. The remaining 2 1 subjects, comprising 47.7% of the sample, responded beneficially to nonpharmacological aspects of th e hospital milieu and/or placebo medication . This high rate of baseline response is in agreement with other inpatient treatment stud ies of aggressive beh avior in conduct disorder. Reports of other investigators suggest that many children hospitalized for aggressive behavior have shown a reduction in or absence of thi s behavior during th e baseline period of controlled clinical trials (Campbell et aI.,

1984, 1995; Cueva et al., 1996). Furthermore, it was reported that placebo response during the treatment per iod was related to psychosocial stressors and the presence of hyperactivity (C ampbell et al., 1995 ; Sanchez et al., 1994) . In thi s study, it was found that placebo baseline nonresponders had a longer history of aggressive beh avior than d id placebo baseline responders. It is also conceivable that placebo baseline responders and nonresponders differ on a number of variables. Previously, we reported that placebo baseline responders and nonresponders differed significantly on scores from th e Youth Self-Report (M alone et al., 199 4a) . Further research should be undertaken to identify children who are likely to respond to hospitalization and/or placebo med ication. Such research would be helpful in det ermining the type of children who require psycho social treatments and/or environment al change including hospitalization. Prior reports with adult inpatients indicate that psychiatric hospitalization can have a therapeutic effect (Ca rroll et al., 1980; Pi er al., 1983). Pi and associates (1983) reported on 67 pat ients with mixed d iagnoses (schizophrenia, 28 % ; person ality disorder, 16% ; depression , 9.7%; drug psycho sis, 5%; and mania, 4%) who were followed up, medication-free. for 1 week after ad missio n to the ho spital. These aut ho rs found

TABLE 3 O vert Aggression Scale Scores Mean Scores per Weekly Period Severity" To tal incide nts Physical incident s Serious physical incidents" Verbal incide nts

Baseline N onr esponders (n

33 .7 7.48 4.04 2.39 3.43

= B) ::': 12 ::': 2.5

± 1.6 ::': 1.4 ::':

1.4

Baseline Respond ers ( n = 2 1) 8.6 2.36 1.02 0.74 1.33

::': ::': ::': ::': ::':

6.4 1.6 0.8 0.56 1.1

Statist ic ( t)

p

-8.76 - 7.9 7 - 7.94 - 5.31 -5 .52

.00 1 .001 ,DOl .00 1 .00 1

" Score based on tot al nu mber of aggression points. "Physical incidents with aggression at or above the second level of severity.

J.

AM . ACAD . CHI!.D ADOl.ESC. PSYCH IAT RY..\(,,2 . FEBR UA RY ) ')') 7

245

MALONE ET AI..

that 20.9% of patients' symptoms remitted without the use of medication. It should be noted that these adult patients had a substantial rate of disorders in which medications are clearly indicated and have established effectiveness. In conduct disorder with severe aggression, psychoactive agents are not as clearly indicated or effective. While lithium has shown efficacy in reducing aggressive behavior in inpatients with conduct disorder(Campbell et al., 1984, 1995), efficacy has not yet been established for outpatients (Klein, 1991). In the present study, the subjects had received the diagnosis ofconduct disorder, and all had long-standing histories of serious aggressive behavior for which they required hospitalization. This behavior had failed to respond to other treatments and could not be controlled in their home situations. Many clinicians assumed that, because of its chronicity, this behavior would be difficult to treat and that treatment with placebo would not be successful. However, nearly half of these children did not meet the criterion for aggressive behavior during the 2-week placebo baseline period. Because these children left their home environments and were hospitalized, it is difficult to attribute the amount of aggression demonstrated during the baseline period to the administration of placebo alone. The children's behavior was undoubtedly affected by the change in environment and the hospital milieu and structure. Our data indicate that in clinical practice with inpatients, medication should not commence immediately upon admission. Time should be allowed for the possible response of the child to hospitalization. If a child begins to take medication soon after admission, there is a high probability that response will be incorrectly attributed to drug, while the response may be non pharmacological. Because aggression in conduct disorder is viewed as a chronic condition, these children may continue to take the medication on a long-term basis (Werry, 1993), meaning that close to 50% of such children could continue to take medication with unnecessary exposure to side effects. It is of concern to many clinicians that third-party reviewers equate the need for hospitalization with a need for treatment with medication. Our data suggest that hospitalization itself may have a therapeutic effect on aggressive behavior in children with conduct disorder. Perhaps children who respond to hospitalization will have a different long-term outcome than those who do not.

246

J.

The finding that there was such a high "placebo response" in this study has important implications for pharmacological research in child psychiatry. It is essential to have a placebo baseline period so that baseline responders can be excluded from entering the treatment period of the study. By having a baseline period, one may avoid prescribing medication to children where this is not indicated. The placebo baseline period is important not only to avoid giving medication needlessly to children, but also to avoid type II errors. At enrollment for this study, there was no way to identify which of the subjects would respond during the placebo baseline period. Yet, close to 50% of the subjects in the study were placebo baseline responders. Given the sample size, only extremely powerful treatments would show differential effects for drug versus placebo if such a large number of baseline responders were to be randomized to lithium versus placebo. In this sample of patients, the ratings of aggression were fairly constant during the 2-week placebo baseline period. Therefore, it is conceivable that a single week of baseline would be sufficient to eliminate some of the placebo baseline responders. Further research should assess whether the low rate ofaggressivebehavior in baseline responders remains beyond the 2-week period studied. It will be important to determine whether the aggressive behavior returns. What we do not know is what would happen to the subjects if they returned to the same home environment without pharmacotherapy. Isolating the specific therapeutic factors at work in the hospital environment may make it possible to incorporate them into the home environment. Follow-up studies of this population are required.

REFERENCES Achenbach TM (l991a), Manual fOr the Child Behavior ChtCklistl4-18 and 1991 Profil«. Burlington: University of Vermont Department of Psychiatry Achenbach TM (l991b), Manual for the YouthSr/fRrportand 1991 Projik. Burlington: University of Vermont Department of Psychiatry Ambrosini PJ, Bianchi MD, Rabinovich H, Elia J (1993), Antidepressant treatments in children and adolescents I. Affective disorders. jAm Acad Child Adolesc Psychiatry 32: 1-6 American Psychiatric Association (1987), Diagnostic and Statistic Manual ofMental Disorders. 3rd edition, reoised (DSM-lll-R). Washington, DC: American Psychiatric Association Campbell M, Adams PB, Small AM er al. (1995), Lithium in hospitalized aggressive children with conduct disorder: a double-blind and placebocontrolled study. jAm Acad Child Adolesc Psychiatry 34:445-453 Campbell M, Cohen IL, Small AM (1982), Drugs in aggressive behavior. jAm Acad Child Psychiatry 21:107-117

AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 36:2, FEBRUARY 1997

PHARMACOTHERAPY AND HOSPITALIZATION

Campbell M, Gonzalez NM, Silva RR (1992), The pharmacologic treatment of conduct disorders and rage outbursts. Psychiatr Clin North Am 15:69-85 Campbell M, Small AM . Green WH er al, (1984), Behavioral efficacy of haloperidol and lithium carbonate: a comparison in hospital ized aggressive child ren with conduct disorder. Arch Gm Psychiatry 41:650- 656 Carroll RS. Miller A, Ross B. Simpson GM (1980), Research as an impetus to improved treatment. Arch Gen Psychiatry 37:377-380 Cueva JE, Overall JE . Small AM. Armenteros JL. Perry R. Campbell M (1996) . Carbamazep ine in aggressive ch ildren with conduct disorder: a double-blind and placebo-controlled study . J Am Acad Child Adolesc Psychiatry 35:480- 490 Ernst M, Locascio J] , Small AM. Lynch N, Campbell M (1990), Placebo response in autistic children. In: Sciemific Proceedings of th« Annual Muting. American Academy of Child and Adolescent Psychiatry, Vol 6. p 47 Guy W (1976) . ECDEU Assessment Manual for Psychophannacology (D H EW Publication ADM 76- 338). Rockville. MD: US Department of Health. Education. and Welfare Holl ingshead AB (1957) . Two Factor Index ofSocial Position. New Haven . CT: Yale Universiry Department of Sociology Jensen PS, Vitiello B, Leonard H, Laughren TP (1994) . Design and methodology issues for clinical treatment trials in children and adolescents. Psychophannacol Bull 30:3-8 Kafanraris V. Campbell M. Padron-Gayol MV. Small AM. Locascio JJ. Rosenberg CR (1992), Carbarnazepine in aggressive conduct disorder children: an open pilot study. Psychopbarmacol Bull 28: 193-199 Kazdin AE (1987). ConductDisorders in Childhoodand Adolescence. Newbury Park, CA: Sage Klein RG (1991), Prelim inary results: lithium effects in conduct disorders. In : CME Syllabus and ProcudingsSummary. Symposium 2. 144th Annual Muting of tht American Psychiatr ic Association, New Orleans. pp . 119-120 Locascio J]. Malone RP, Small AM et al, (1991). Factors related to haloperidol response and dyskinesias in autistic children. Psychopbarma-

col Bull 27:119-126 Loney J. Milich RS (198 2). H yperactiviry, inattention. and aggression in clinical pract ice. In: Aduances in Behavioral Pediatrics, Vol 2. Wolraich M . Routh D. eds. Greenwich. CT: JAI Press

J.

Malone RP, Biesecker KA, Delaney MA. White MA. Luebbert J (I994a), Youth Self-Report in aggression: baseline placebo responders and nonresponders. In : Scimtific Proceedings, 4 Jst Annual Muting oftht American Acatkmy of Child and Adolescen: Psychiatry, New York, p 49 Malone RP. Luebbert J, Pena-Ariet M. Biesecker K. Delaney MA (I994b). The Overt Aggression Scale in a study of lithium in aggressive conduct disorder . Psychopbarmacol Bull 30:215-218 Pi EH, Miller A. Rosenberg MR, Schultz CL, Simpson GM (1983) , Therapeutic effects of psychiatr ic hospitalization. J Psychiatr Treat

EvaI5:135-142 Psychophannacology Bulletin (1973). Pharmacotherapy of Children. Special issue

Reich W. Weiner Z (1990), The Diagnostic Interview for Children and Adoltscmts-(l) Revised. St Louis: Washington University Rifkin A, Karajgi B, Perl E. Dicker R, Boppana V, Hasan N (1989), Lithium treatment of conduct disorders in adolescents . Paper presented at the 29th annual meeting of the New Clinical Drug Evaluation Unit (NCDEU) . Key Biscayne. FL. June Sanchez LE. Arrnenreros JL. Small AM. Campbell M. Adams PB (1994) . Placebo response in aggressive children with conduct disorder. Psycho-

pharmacal Bull 30:209-213 Silver JM, Yudofsky SC (198 7). Documentation of aggression in the assessment of the violent pat ient . Psychiatr Ann 17:375-384 US Department of Health and Human Services (1991) . Healthy People 2000: National Health Promotion and Disease Prevention Objectiues (Publication PHS 91-50213). Washington, DC: US Department of Health and Human Services Werry JS (1993). Long-term drug use in psychiatric disorders in children. Facts, controversies, and the future . Acta Paedopsycbiatr 56: 113-118 Werry JS, Wollersheim JP (1989) . Behavior therapy with children and adolescents : a twenty- year overview. JAm Acad Child AdoltJC Psychia-

try 28:1-18 Yudofsky SC. Silver JM. Jackson W , Endicott J, Williams D (1986) , The Overt Aggression Scale for the objective rating of verbal and physical aggression. Am J Psychiatry 143:35-3

AM . ACAD. C H IL D A DO L ESC. PSYCHIATRY . 3 6:2 . FEBRUARY 199 7

247