Parental Substance Abuse and Suspected Child Abuse/Maltreatment Predict Outcome in Children's Inpatient Treatment STEWART GABEL , M.D .,
AND
RICHARD SHINDLEDECKER, M.A .
Abstract. There has been relatively little empirical study of outcome for school aged children treated in psychiatric hospitals. In earlier studies, four preadm ission variables were operationally defined and studied as hypothe sized predictors of poor outcome in psychiatr ic day treatment by the criterion of recommended in-home or out-of-home placements on discharge. The preadmission variables are recent histories of severe aggression and suicidal ideation/behavior in the child and histories of parental substance abuse and reported suspicions of child abuse/maltreatment. In this study, the work has been extended to determine if these four preadmission factors relate to outcome of children treated in inpatient hospitalization. The findings indicate that the parent/family related variable s, and not the child variables (nor child diagnosis ), are significantly related to out-of-home recommended placements . J . Am . Acad. Child Adolesc. Psychiatry, 1990, 29 ,6:919-924. Key Words: parental substance abuse , child abuse , treatment outcome. There has been relatively little empirical study devoted to an assessment of outcome for severely disturbed children treated in inpatient hospital settings (Blotcky et al. , 1984) despite the apparently increasing and at times controversial use of inpatient treatment for children and adolescents (American Academy of Child and Adolescent Psychiatry, 1989). In earlier work (Gabel et al., 1988), a model was developed, operationally defined, and tested empirically to determine whether specific preadmission factors in the child and/or in the parents or family could be used to predict outcome in children's day treatment . Two preadmission factors in the child , severe aggression and suicidal ideation/ behavior, and two preadmission factors in the parents/family, reported suspicions of child abuse/maltreatment and parental substance abuse, were hypothesized as predictors of poor outcome. The outcome criterion used was based on an important clinical dimension in the psychiatric treatment of children and adolescents : the determination of whether a child should remain in the commun ity or should be recommended for out-of-home placement in residential or hospital settings . In the initial study , Gabel et al. (1988) found that each of the four factors was associated with recommended outof-home placements from hospital based day treatment , with the child factor of severe aggression and the parent/family
factor of parental substance abuse being most important in predicting outcome. In addition, the four factors in the model were additive , such that increasing numbers of preadmission factors in the child 's history were significantly associated with poor outcome . In a second study at the same day treatment program, after several programmatic changes had been made to address factors such as severe aggression in the child and family involvement in the program , Gabel et al. (in press) found that outcome for aggressive children in the group, defined in a manner similar to that of the earlier study , showed significant improvement, with aggression no longer being a determinant of poor outcome , although parental substance abuse continued to be significant as a predictor of poor outcome . In another study, Gabel et al. (1989) questioned whether the earlier findings of Gabel et al. (1988) would generalize to other day treatment settings and reviewed the outcome of two additional day treatment programs. One program was hospital based (as was the original center) and one was community based. The findings indicated that the model was not supported when considering the other two treatment centers individually . When the three day treatment centers' findings were considered together , however , the four preadmission factors were each significantly associated with outcome, with severe aggression and a history of reported suspicions of child abuse/maltreatment being most important in predicting outcome . The current study tests the hypothesis that these four preadmission factors , severe aggression in the child , suicidal ideation/behavior in the child, reported suspicions of child abuse/maltreatment, and a history of parental substance abuse as indicators of parent/family dysfunction also predict poor outcome in children's inpatient treatment .
Accepted June 7, 1990. Dr . Gabel is Unit Chief . Children' s Day Hospital, New York Hospital-Cornell Medical Center , Westchester Division, and Assistant Professor of Psychiatry, Cornell University Medical College. Mr . Shindledecker is Head of Data Management, Department of Psychiatry, St . Luke' s-Roosevelt Hospital Center, College of Physicians and Surgeons, Columbia University. . The authors wish to acknowledge Cynthia R . Pfeffer, M.D .,for her review of an earlier draft ofthis paper. Lisa Christopher, M .A. , served as research assistant. Request reprintsfromDr. Gabel, New York Hospital-CornellMedical Center, WestchesterDivision, 21 Bloomingdale Road, White Plains,
Method Sample
All children discharged from the Child Inpatient Unit (CIPU) at New York Hospital-Comell Medical Center, Westchester Division from January 1, 1986, to June 30,
NY 10605.
0890-8567/90/2906-0919$02.00/0© 1990 by the American Academy of Child and Adolescent Psychiatry .
919
GABEL AND SHINDLEDECKER TABLE
1. Demographic, Diagnostic, and Primary Variables with Outcome Significancea
Recommended Disposition Variable Sex Male Female Race White Nonwhite Medicaid Yes No Primary Axis I DSM-IIl-R diagnosis Conduct disorder Depressive disorder, (Major depression or dysthymic disorder) Adjustment disorder Schizophrenia Primary preadmission variables Severe aggressive/ destructive behavior Suicidality Suspected child abuse/maltreatment Parental substance abuse Age Length of stay (days)
In Home
25 4
22 7
p
Out of Home 1.72
0.19
1.98
0.16
6.12
0.02
33 12 27 18
22
24 21
9 6
16 11
0.16 0.14
0.69 0.71
3 2
3 4
0.32 0.09
0.57 0.76
18 17
27 24
0.03 0.20
0.86 0.66
4
16
4.24
0.04
7
11 9.48 ± 2.2 133.3 ± 86.4
29 9.92 ± 2.3 166.7 ± 84.3
4.99 t = 0.80 t = 1.61
0.03
df = 72 p
AND
RICHARD SHINDLEDECKER, M.A .
Abstract. There has been relatively little empirical study of outcome for school aged children treated in psychiatric hospitals. In earlier studies, four preadm ission variables were operationally defined and studied as hypothe sized predictors of poor outcome in psychiatr ic day treatment by the criterion of recommended in-home or out-of-home placements on discharge. The preadmission variables are recent histories of severe aggression and suicidal ideation/behavior in the child and histories of parental substance abuse and reported suspicions of child abuse/maltreatment. In this study, the work has been extended to determine if these four preadmission factors relate to outcome of children treated in inpatient hospitalization. The findings indicate that the parent/family related variable s, and not the child variables (nor child diagnosis ), are significantly related to out-of-home recommended placements . J . Am . Acad. Child Adolesc. Psychiatry, 1990, 29 ,6:919-924. Key Words: parental substance abuse , child abuse , treatment outcome. There has been relatively little empirical study devoted to an assessment of outcome for severely disturbed children treated in inpatient hospital settings (Blotcky et al. , 1984) despite the apparently increasing and at times controversial use of inpatient treatment for children and adolescents (American Academy of Child and Adolescent Psychiatry, 1989). In earlier work (Gabel et al., 1988), a model was developed, operationally defined, and tested empirically to determine whether specific preadmission factors in the child and/or in the parents or family could be used to predict outcome in children's day treatment . Two preadmission factors in the child , severe aggression and suicidal ideation/ behavior, and two preadmission factors in the parents/family, reported suspicions of child abuse/maltreatment and parental substance abuse, were hypothesized as predictors of poor outcome. The outcome criterion used was based on an important clinical dimension in the psychiatric treatment of children and adolescents : the determination of whether a child should remain in the commun ity or should be recommended for out-of-home placement in residential or hospital settings . In the initial study , Gabel et al. (1988) found that each of the four factors was associated with recommended outof-home placements from hospital based day treatment , with the child factor of severe aggression and the parent/family
factor of parental substance abuse being most important in predicting outcome. In addition, the four factors in the model were additive , such that increasing numbers of preadmission factors in the child 's history were significantly associated with poor outcome . In a second study at the same day treatment program, after several programmatic changes had been made to address factors such as severe aggression in the child and family involvement in the program , Gabel et al. (in press) found that outcome for aggressive children in the group, defined in a manner similar to that of the earlier study , showed significant improvement, with aggression no longer being a determinant of poor outcome , although parental substance abuse continued to be significant as a predictor of poor outcome . In another study, Gabel et al. (1989) questioned whether the earlier findings of Gabel et al. (1988) would generalize to other day treatment settings and reviewed the outcome of two additional day treatment programs. One program was hospital based (as was the original center) and one was community based. The findings indicated that the model was not supported when considering the other two treatment centers individually . When the three day treatment centers' findings were considered together , however , the four preadmission factors were each significantly associated with outcome, with severe aggression and a history of reported suspicions of child abuse/maltreatment being most important in predicting outcome . The current study tests the hypothesis that these four preadmission factors , severe aggression in the child , suicidal ideation/behavior in the child, reported suspicions of child abuse/maltreatment, and a history of parental substance abuse as indicators of parent/family dysfunction also predict poor outcome in children's inpatient treatment .
Accepted June 7, 1990. Dr . Gabel is Unit Chief . Children' s Day Hospital, New York Hospital-Cornell Medical Center , Westchester Division, and Assistant Professor of Psychiatry, Cornell University Medical College. Mr . Shindledecker is Head of Data Management, Department of Psychiatry, St . Luke' s-Roosevelt Hospital Center, College of Physicians and Surgeons, Columbia University. . The authors wish to acknowledge Cynthia R . Pfeffer, M.D .,for her review of an earlier draft ofthis paper. Lisa Christopher, M .A. , served as research assistant. Request reprintsfromDr. Gabel, New York Hospital-CornellMedical Center, WestchesterDivision, 21 Bloomingdale Road, White Plains,
Method Sample
All children discharged from the Child Inpatient Unit (CIPU) at New York Hospital-Comell Medical Center, Westchester Division from January 1, 1986, to June 30,
NY 10605.
0890-8567/90/2906-0919$02.00/0© 1990 by the American Academy of Child and Adolescent Psychiatry .
919
GABEL AND SHINDLEDECKER TABLE
1. Demographic, Diagnostic, and Primary Variables with Outcome Significancea
Recommended Disposition Variable Sex Male Female Race White Nonwhite Medicaid Yes No Primary Axis I DSM-IIl-R diagnosis Conduct disorder Depressive disorder, (Major depression or dysthymic disorder) Adjustment disorder Schizophrenia Primary preadmission variables Severe aggressive/ destructive behavior Suicidality Suspected child abuse/maltreatment Parental substance abuse Age Length of stay (days)
In Home
25 4
22 7
p
Out of Home 1.72
0.19
1.98
0.16
6.12
0.02
33 12 27 18
22
24 21
9 6
16 11
0.16 0.14
0.69 0.71
3 2
3 4
0.32 0.09
0.57 0.76
18 17
27 24
0.03 0.20
0.86 0.66
4
16
4.24
0.04
7
11 9.48 ± 2.2 133.3 ± 86.4
29 9.92 ± 2.3 166.7 ± 84.3
4.99 t = 0.80 t = 1.61
0.03
df = 72 p
