Return to Previous page |
| PDF Version, B&W |  |
NSCAW Publications
| Title | Authors | Source/ Publication Info/ Year | Data Set | Keywords | Sample | Analysis | Results |
|---|---|---|---|---|---|---|---|
| Gender and ethnic differences for Hispanic children referred to child protective services | Alzate, M.M. Rosenthal, J. A. | Children and Youth Services Review Available online 2008 | Caregivers interview Caseworkers interview Child Interview Baseline Wave 3 Wave 4 |
Child protective services; Hispanic/Latino; gender differences; ethnic differences | NSCAW CPS sample (n= 5,501) | Examined differences between Hispanic and non-Hispanic children. Univariate Bivariate Logistic regression Linear regression Ordinal probit regression. |
Hispanic children: 18% Hispanic boys more likely than non-Hispanic to be reported for physical abuse (43% vs. 29%), and less likely to be reported for neglect or other (51% vs. 66%). Among children in placement a higher percentage of Hispanic are in the birth to 2 years old group (Hispanic; 50%, non-Hispanic: 24%). Multivariate analysis: Hispanic children less likely to be reported for neglect or other (OR: 0.47), and less likely to have behavioral problems (B: -3.2). |
| Depression among Latino children in the public child welfare system. | Ayón, C. Marcenko, M. O. | Children and Youth Services Review, Vol.30, pp. 1366-1375. 2008 | Caseworker interview Caregiver interview Child interview Baseline Wave 3 Wave 4 |
Latino families; mental health; child welfare | Children of Latino origin 7 years and older (n=385) | Examined rate and severity of depression longitudinally, and the relationship with family immigration status (length of residency in the U.S less than 10 years vs. more). Univariates Bivriates Latent growth model analysis (Mplus 4.0) |
Latino composition: Mexican: 64%; Puerto Rican 15%; Other 21%. Levels of depression were low, children experienced less depression symptoms over time: Wave 1 mean 10.5, range 0-39, clinical 11%; wave 3 mean 9.1, range 0-33, clinical 6%; wave 4 mean 7.8, range 0-31, clinical 4%. LGM: at baseline older children had higher levels of depression than younger children. Mexican and Puerto Rican children experienced higher levels of depression than other children. Over time, Mexican children has a slight decrease in depression, while Puerto Rican children experienced a slight increase in depression compared to other Latino children that experienced a sharp decline in depression. |
| Health care use of children whose female caregivers have intimate partner violence histories | Bair-Merritt, M.H. Feudtner, C. Localio, A.R. Feinstein, J.A. Rubin, D. Holmes, W.C. | Archives of Pediatrics and Adolescent Medicine Vol.162, pp. 134-139 2008 | Caregivers interview Caseworkers interview Baseline Wave 3 Wave 4 |
Maternal depressive symptoms; domestic violence; young children; women; emergency experiences; prevalence; injuries; exposure; asthma | Children living in home with a female primary caregiver. Baseline-wave 3 (n=2,689) Wave 3-wave 4 (n=2,546) | Examined baseline report of IPV and children’s subsequent ED use and hospitalization from baseline to Wave 4 Frequencies, bivariates, negative binomial multivariate regression (Stata 9.1) | Mean number of ED visits and hospitalizations: 1.4 first interval (compare to no IPV, IRR: 2.0; 95% CI, 1.3-2.9) and 0.9 second interval (compare to no IPV, IRR: 1.9; 95% CI, 1.2-3.0) among children whose caregiver disclosed severe IPV; 0.9 and 0.5 children whose caregiver disclosed minor IPV; 0.7 and 0.5 children whose caregiver disclosed no IPV. |
| Propensity score matching strategies for evaluating the success of child and family service programs |
Barth, R.P. Guo, S. McCrae, J.S. |
Research on Social Work Practice Vol. 18, pp. 212-222 2008 | Caregivers interview Caseworkers interview Baseline Wave 2 Wave 3 |
Child welfare; propensity score matching; evaluation; services research. | Study 1: in-home children whose primary caregiver was female (n= 2,758) Study 2: children in kin care at least 75% of time between baseline and wave 3 (n=169), children in foster care for at least 75% of time between same period (n=398) |
Examined the utility of applying PSM to the analysis of NSAW through 3 studies: (1) re-report rates among caregivers who received substance abuse services, (2) well-being of children in kin-foster care, and (3) psychopathology symptoms in children with mental health and family based service receipt. Propensity score matching, survival analysis, -test of mean differences in change scores, OLS regression, analysis of covariance. | Study 1: 10% of female caregivers received substance abuse treatment. After (PSM) the group with SA treatment had a significantly higher rate of re-reports (18.7% among caregivers with SA treatment and 10.0% among caregivers without treatment). Study 2: Overall decline in self care as measured by the Vineland by about 10 points, no significant differences by placement before or after PSM. Social skills score remained stable, with no significant differences by placement before or after PSM. Children in kinship care had lest externalizing behavioral problems and those in foster care had more externalizing behavioral problems from baseline to wave 3, differences were significant before and after PSM. No significant differences were found related to academic achievement or intelligence measures before or after PSM. Study 3: Before PSM, children who received mental health and family based services improved significantly less in internalizing and total problem behavior than other children. Post PSM these differences were not significant, indicating that childrren receiving services are not worse off. |
| Developmental Status and Early Intervention Service Needs of Maltreated Children. | Barth, R. P., Scarborough, A., Lloyd, E. C., Losby, J., Casanueva, C., & Mann, T. | Report Office of the Assistant Secretary for Planning and Evaluation. U.S. Department of Health and Human Services 2008 | Caseworker interview Caregiver interview Child assessments Baseline Wave 2 Wave 3 Wave 4 |
Development; early intervention services | Infants less than 36 months at baseline with a maltreatment case classified as substantiated or high risk (no n disclosed) | Compares developmental needs and early intervention services use by infants in NSCAW and infants in the National Early Intervention Longitudinal Study (NEILS). Univariate Bivariates Logistic regression |
Measured Delay at Baseline: 49%. Of those with measure delay at baseline, 51% had MD at wave 3 and 44% at wave 4. Of those with measure delay at Wave 3, 57% had measure delay at wave 4. IFSP at Wave 2: 28% |
| Behavioral problems following reunification of children in long-term foster care | Bellamy, J.L. | Children and Youth Services Review Vol. 30 , pp. 216-228 2008 | Caregiver Interview Caseworker Interview Child Interview Baseline Wave 3 Wave 4 Long term foster care sample |
Foster care; reunification; behavior problems; multiple imputation; propensity score matching | Children 2 and older who had experienced at least 8 months of foster care.(n=604) | Examined the relationship between reunification and behavioral outcomes for children. Bivariate, logistic regression, propensit score matching (PSM)(STATA0 9.0) | Reunification: 59.8% had no reunification, 9.5% had reunification at all 3 waves, 30.7% were in and out of home in following waves. Children reunified at all waves were compared to a sub-group not-reunified matched by PSM. No significant associations were found between reunification and behavioral outcomes. |
| Parenting services for mothers involved with child protective services: Do they change maternal parenting and spanking behaviors with young children | Casanueva, C. Martin, S. L. Runyan, D. K. Barth, R. P. Bradley, R. H. | Children and Youth Services Review Vol. 30, pp. 861-878 2008 | Caregiver Interview Caseworker Interview Baseline Wave 3 | Parent training; parent support; program evaluation; maternal parenting; spanking; Child Welfare System; Child Protective Services; national survey of child and adolescent well-being; child maltreatment; prevention | Female caregivers of children 0-10 years who received parenting services and a comparison group obtained trough propensity score matching (n=588) | Examined changes in parenting (parental responsiveness and learning stimulation) and spanking behavior among women who received parenting services with changes among women who did not. Univariate Bivariate Logistic regression Propensity Score matching Multiple linear regression. |
Parenting services: median 12 sessions (mean = 23.9, SD = 37.7; range: 1–446); 36% received only parenting skills training, 30% received parenting skills training and family counseling; 21% received only family counseling; 11% received only parent aide services; and 2% received all three services. Bivariates: Among children 3- to 5-year-old, mothers who received services improved responsiveness while those without services deteriorated (improvement: 0.93 to 0.96; deterioration: 0.94 to 0.86). Multivariate analysis did not show any significant differences between those who receive and did not receive services in any of the measures of parenting, neither on changes in spanking behaviors. |
| Developmental Needs and Individualized Family Service Plans Among Infants and Toddlers in the Child Welfare System | Casanueva, C. E. Cross, T. P. Ringeisen, H. | Child Maltreatment Vol. 13, pp. 245-258 2008 | Caregiver Interview Caseworker Interview Child interview Baseline Wave 2 Wave 3 Wave 4 Wave 5 |
child development; longitudinal research; measure development; children in child welfare; infants; logistic re | 1845 children aged 0 to 36 months at baseline with information on substantiation status (n=1,845) | Examined levels of developmental need, early intervention service use, and variations in need and service use during the following 5 to 6 years by substantiation status. Univariate Bivariates Logistic regression |
Developmental need defined as one measure at least 2 SD under the mean or two measures at least 1.5 SD under the mean. Developmental need: 35.2% baseline; 39.2 wave 3; 41.1% wave 4; 42.0% wave 5. No significant differences between substantiated and unsubstantiated children except at baseline (need among unsubstantiated: 38.2%, need among substantiated: 27.6%). Early intervention (IFSP): 12.7% Predictors of IFSP: substantiation (OR: 1.9); in home active CWS case (OR: 3.8); out of home (OR: 5.4). |
| Quality of Maternal Parenting among Intimate-Partner Violence Victims Involved with the Child Welfare System | Casanueva, C. Martin, S. L. Runyan, D. K. Barth, R. P. Bradley, R. H. | Journal of Family Violence Vol. 23, pp. 413-427 2008 | Caregiver Interview Caseworker Interview Baseline | Parenting behaviors; intimate partner violence; national study; child abuse and neglect | Female caregivers of children 0-10 years living in home at baseline (n= 1,943) | Examined the relationship between women’s experiences with IPV and the quality of maternal parenting. Univariate Bivariate Linear regression |
Current physical IPV: 31% Past physical IPV: 16% Women who had experienced IPV in the past had significantly higher (i.e., better) parenting scores based on the Total Home-SF (B = 0.04; p < .05) than women who were currently experiencing IPV. There were no significant differences between women who were currently experiencing IPV and women who had never experienced IPV. |
| Repeated reports for child maltreatment among intimate partner violence victims: Findings from the National Survey of Child and Adolescent Well-Being | Casanueva, C. Martin, S. L. Runyan, D.K. | Child Abuse & Neglect In Press 2008 | Caseworker Interview Caregiver Interview Baseline Wave 2 Wave 3 |
Re-reports; Intimate partner violence | 1,236 children not placed in out-of-home care whose caregivers were: (1) the alleged perpetrators of the child maltreatment at baseline (independently of substantiation status) and (2) biological mothers (98.6%), adoptive mothers (1%), or stepmothers (0.3%). | Examined whether mothers’ experiences of intimate partner violence (IPV) were associated with repeat reports (re-reports) for child maltreatment to Child Protective Services (CPS). Univariates Bivariates Logistic regression Survival analysis. |
Almost half (44%) of the mothers reported for alleged child maltreatment at baseline had experienced physical violence by their partner. Children of mothers physically abused by an intimate partner during the last 12 months or previously at the intake interview were twice as likely as children of mothers who had not experienced such violence to be re-reported to CPS (OR:.2.0) Re-reports occurred almost twice as quickly for children of mothers who experienced IPV compared to children of mothers who had not experienced IPV (HR: 1.9). |
| Caseworkers Judgments and Substantiation | Cross, T. Casanueva, C.E. | Child Maltreatment Available online 2008 | Caseworker interview Caregiver interview Baseline | Maltreatment; substantiation; evidence | Children with information on substantiation (n=4,514) | Examined the relationship of harm, risk and sufficiency of evidence to substantiation. Univariate Bivariate Logistic regression |
Substantiation rate : 29.9% Predictors of substantiation: harm (OR: 1.6); risk (OR: 1.9); evidence (OR: 3.1); female (OR: 1.9); age 6-10 (OR: 2.2). Predictors of substantiation of physical abuse: harm (OR: 1.8); risk (OR: 2.5); evidence (OR: 3.2); female (OR: 2.2). Predictors of substantiation of sexual abuse: harm (OR: 2.4); evidence (OR: 5.0); age 3-5 (OR: 0.1); Hispanic (OR: 0.2); income less than $15,000 (OR: 7.1). Predictors of substantiation of failure to provide: evidence (OR: 4.4); age 6-10 (OR: 3.2). Predictors of substantiation of failure to supervise: harm (OR: 1.6); risk (OR: 1.9); evidence (OR: 3.5); female (OR: 2.3). |
| Caseworker assessments of risk for recurrent maltreatment: association with case-specific risk factors and re-reports | Dorsey, S. Mustillo, S.A. Farmer, E.M. Elbogen, E. | Child Abuse & Neglect Vol. 32, pp. 377-91. 2008 | Caseworker interview Caregiver Interview Baseline Wave 2 Wave 3 |
Risk factors; recurrent maltreatment; risk assessment; decision-making | Children with a report of child physical abuse or neglect who were not placed in out-of-home care (N=2,139) | Examined associations between caseworkers' risk assessments and demographic, child, parent and family-level risk factors; and agreement between caseworkers' risk assessments and subsequent reports of maltreatment. Bivariates (chi square) logistic regression (Stata 9.0) |
Recurrent maltreatment: 20.2%; 16.1% among those classified as low risk, 28.1% among those classified as high risk. Predictors of recurrence: Caregiver history of being abused or neglected and prior reports of maltreatment associated with recurrent maltreatment among those with physical abuse at index investigation, only prior report associated with recurrent maltreatment among those with neglect at index investigation |
| Behaviors of youth involved in the child welfare system | Grogan-Kaylor, A. Ruffolo, M. C. Ortega, R. M. Clarke, J. | Child Abuse & Neglect Vol. 32, pp. 35-49. 2008 | Child Interview Caregiver Interview Caseworker Interview Baseline Wave 3 | Delinquent; youth behaviors; child welfare system | Youth ages 11 and older (n= 1,180 at baseline, n= 1,273 at wave 3) | Examined protective and risk factors associated with youth delinquent behaviors over time. Univariates Tobit regression |
Mean number of delinquent behaviors was 12.3 (SD=2.4). Many have 0 delinquent behaviors. Predictors of delinquency: Older youth, males, youth who received CWS services, and physically abuse youth were more likely to engage in delinquent behaviors than other youth. Increases in caregiver monitoring and in the quality of relationship with caregivers were associated with decreases in delinquent behaviors. |
| Self-reported disciplinary practices among women in the child welfare system: Association with domestic violence victimization. | Kelleher, K. J. Hazen, A. L. Coben, J. H. Wang, Y. McGeehan, J. Kohl, P. L. Gardner, W. P. | Child Abuse & Neglect Vol. 32, pp.811-8. 2008 | Caseworker Interview Caregiver Interview Child Assessment Baseline | Domestic violence; child abuse; parenting | Permanent female caregivers (n= 1,305) | Examined the association between physical domestic violence victimization (both recent and more than a year in past) and self-reported disciplinary practices . Descriptive statistics (univariate), univariate logistic regression analysis, and multivariate logistic regression. | 55.5% reported no domestic violence exposure , 28.9% reported prior year domestic violence, 15.6% reported domestic violence but not in the past 12 months. In bivariate comparisons any prior domestic violence exposure was associated with higher rates of psychological aggression, physical aggression and neglectful disciplinary behaviors as compared to those with no domestic violence victimization. In multivariate analysis those with remote and recent domestic violence victimization employed more psychological aggression, while only caregivers with recent DV reported more physical aggression and neglectful behaviors. |
| Profiles of victimized women among the child welfare population: Implications for targeted child welfare policy and practices | Kohl, P.L. Macy, R.J. | Journal of Family Violence Vol, 23, pp. 57-68 2008 | Caregiver Interview Caseworker Interview Baseline Wave 3 | Domestic violence; child welfare; child maltreatment; latent class analysis | Female permanent caregivers with at least one incident of IPV in the 12 months prior to baseline (n= 1,229) | Identified patterns of needs and resources among sub-groups of victimized women, and associations with domestic violence at wave 3 and child maltreatment re-reports. Latent class analysis (Mplus 3.11), ANOVA, Bivariate (Chi-square). |
Profiles of needs and resources among victimized caregivers: (1) 13% Multi problem (major depression, childhood history of abuse and neglect, high stress), 2% Substance problems (alcohol and/or drug abuse, high arrest, social support), 20% Stressed-low support ((high levels of stressors, some history of child abuse and neglect, some arrest, low support), and 65% low support only (low in all variables). Although severe violence decline for all groups by wave 3, women in the multi problem group and low support group had significantly higher rates of DV. New child maltreatment reports were significantly different across groups: 46% multi-problem, 4% substance problems, 21% stressed-low support, 29% low support only. |
| Effects of Family Violence on Psychopathology Symptoms in Children Previously Exposed to Maltreatment | Maikovich, A.K. Jaffee , S.R. Odgers, C.L. Gallop, R. | Child Development Vol, 79, pp.1498 - 1512 2008 | Caregiver Interview Caseworker Interview Baseline Wave 3 Wave 4 |
Child psychopathology, physical discipline, witnessing violence, CBCL internalizing and externalizing | Children 5 years or older at baseline (n= 2,925) | Examined the effects of physical discipline and witnessing violence in trajectories of internalizing and externalizing symptoms. Univariate analysis, correlations, Latent Difference Score (LDS) modeling using SEM. | Children externalizing symptoms based on the caregiver report on the CBCL declined from a mean of 16.1 at baseline to 14.6 at wave 4. Internalizing symptoms declined from 10.4 at baseline to 9.5 at wave 4. Harsh physical discipline at baseline predicted changes in score of externalizing symptoms at wave 3, while harsh physical discipline at wave 3 predicted changes in score of externalizing symptoms at wave 4. Witnessing violence at baseline predicted changes in score of internalizing symptoms at wave 3, witnessing violence at wave 3 predicted changes in score of internalizing symptoms at wave 4. |
| Emotional and Behavioral Problems Reported in Child Welfare Over 3 Years | McCrae, J.S. | Journal of Emotional and Behavioral Disorders In press (available online) 2008 | Caseworker interview Caregiver interview Children Interview Baseline Wave 3 Wave 4 |
Emotional and behavioral problems, child welfare services; NSCAW; depression; posttraumatic stress | Children ages 2-14 at baseline with complete 3-wave data (n= 2,852). | Examined child emotional and behaviorla problems, depression and post traumatic stress. Bivariate analysis (chi square tests, conditional probabilities). | Behavioral problems, clinical level (at or above 64): 33.9% baseline, 29.6% wave 3, 26.6% wave 4., total any time: 48.5%. Depression: 15.7% baseline, 6.4% wave 3, 5.8% wave 4, total any time: 19.5%. Posttraumatic stress: 12.8% baseline, 5.9% wave 3, 4.9% wave 4, total any time: 18.8%. Any reported problem: 46.2% baseline, 39.0% wave 3, 36.2% wave 4, total any time: 62.2%. |
| Using cumulative risk to screen for mental health problems in child welfare | McCrae, J.S. Barth, R.P. | Research of Social Work Practice Vol. 18, pp. 144-159 2008 | Caregiver Interview Caseworker Interview Child Interview Baseline | Cumulative risk; NSCAW; mental health screening; child welfare | Children ages 2 to 14 whose primary maltreatment type was physical abuse, sexual abuse, and neglect (n= 3,022) | Examined risk assessment items and developed a cumulative risk score to screen for children mental health problems. Bivariate (chi square and t tests). Logistic regression. |
CBCL borderline or clinical range: 45.9% children 2-6 years old, 64.3% children 7-14 years old, 57.0% total. Cumulative risk score based on caseworker report: range: 0-34; mean: 12.6. The odds of children scoring in the borderline-clinical range increased by .07 with each 1 point increase in the total score. Among children 2-6, cumulative risk scores of 9 or more showed 75% sensitivity to identify children with mental health concerns and 34% specificity. Among children 7-14, cumulative risk scores of 8 or more showed 79% sensitivity and 35% specificity. Using cumulative risk scores based on 34 risk items produced significantly better identification of children with mental health needs than asking the caseworker if the child had major special needs or behavioral problems and/or was refereed to mental health services. |
| Placement preferences among children living in foster or kinship care: A cluster analysis | Merritt, D. H. | Children and Youth Services Review Vol. 30, pp. 1336-1344 2008 | Caseworker interview Caregiver interview Child interview Baseline Wave 3 Wave 4 |
Foster care; Permanency; Placement outcomes; Cluster analysis. | Children ages 6-14 in foster or kinship care (n=1,313) | Examined children placement preferences over time. Univariate Bivariates Cluster analysis |
Identified groups: 1) I don't want to be here/going home (expect to reunite with birth family); 2) I can stay here, but I don't want to/going home (expect to reunite with birth family); 3) I can stay here and I want to, but no adoption; and 4) I can stay here and I want to, with adoption. |
| Differences in patterns of maternal arrest and the parent, family, and child problems encountered in working with families | Phillips , S.D. Erkanli, A. | Children and Youth Services Review Vol. 30, pp. 157-172 2008 | Child Interview Caregiver Interview Caseworker Interview Baseline | Parental incarceration; children of incarcerated parents; criminal careers; delinquency; NSCAW; child and adolescent outcomes | Biologicla mothers with arrest histories (n=959) and no arrest histories (n=2,428). Total n=3,387 | Examined arrest histoires of mothers and categorize mothers into groups of arrest history that are compare in terms of demographics, risk factors, and child problems. Latent class analysis (Mplus 2.13), Dirichlet Process Priors (DPP, in WinBugs). Between group differences (chi square) | Arrest history groups: (1) 3.5% mothers with dated arrest histories (no arrest for last 5 years, substance abuse) , (2) 3.5% mothers with protracted arrest histories (arrested multiple times , 72% last arrest within last 2 years, substance abuse, domestic violence), and (3) 20% mothers who were first arrested at a relatively late age (brief arrests, only arrested 1 or 2 times, overrepresentation of Black, domestic violence, extreme poverty). Children 2-10 years old of mothers in groups 1 and 3 had higher percentages of emotional and behavioral problems than children of non-arrested mothers. Children 11-14 years old of mothers in groups 1 had higher percentages of arrest than children of non-arrested mothers. |
| More than Parents in Prison: The Broader Overlap between the Criminal Justice and Child Welfare Systems | Phillips, S. D. Dettlaff, A. J. | Journal of Public Child Welfare In Press 2008 | Caseworker Interview Caregiver Interview Child Interview Baseline | Parental arrest; Substance abuse; Domestic violence; Child behavioral problems | Children who were in in-home settings at baseline (n=4,285) | Examined prevalence of criminal justice involvement among caregivers and outcomes among children. Univariates Bivariates ( Chi Square) . |
1 in every 3 children in in-home settings had primary caregivers who had been arrested at least once. 1 in every 100 children were living with a primary caregiver who was previously incarcerated 1 in every 10 with a caregiver sentenced to probation 1 in every 5 with a caregiver who had an outcome of arrest other than probation or prison. Substance abuse and domestic violence were more prevalent among primary caregivers with arrest records than among never-arrested caregivers. No statistically significant differences in levels of emotional and behavioral problems. Children whose parents were sentenced to probation reported the lowest levels of child or youth arrest, although children of never arrested caregivers reported similar levels of arrest than children of formerly incarcerated caregivers (about 10%). |
| Longitudinal Patterns of Health Insurance Coverage Among a National Sample of Children in the Child Welfare System | Raghavan, R . Aarons, G.A . Roesch, S.C. Leslie, L.K. | American Journal of Public Health Vol. 98, pp. 478-484 2008 | Caregiver Interview Caseworker Interview Child Interview Baseline Wave 2 Wave 3 Wave 4 |
Medicaid manage care, Foster care; Service use; Policy needs; Maltreatment | Children 2 years and older at baseline (n= 2,501) | Examined stability over time of health insurance. Bivariate (chi square) and Logistic regression (Stata) Latent Class Analysis (grwth mixture modeling, Mplus 4.1) | Uninsured children: 10% baseline and wave 2, 8% wave 3, 6% wave 4. Of children uninsured at baseline, 62% had Medicaid by wave 4. Latent insurance classes: (1) children who gained health insurance over time (7.6%), (2) children who stably maintained coverage over time (92%). History of sexual abuse, and race/ethnicity “other” associated with membership in the "gainer" class. Foster care placement at baseline and poorer health status associated with membership in the "Maintainer" class. |
| Special Health Care Needs Among Children in the Child Welfare System | Ringeisen, H. Casanueva, C. E. Urato, M. P. Cross, T. P. | Pediatrics Vol. 122, pp. 232-241 2008 | Caseworker interview Caregiver interview Child interview Baseline Wave 3 Wave 4 |
Child development; chronic conditions; maltreatment; special needs; survey | Children aged 0 to 15 at baseline (n= 5,496) | Examined levels of special health care needs (SHCN) and how these needs may affect children’s functioning. Univariate Bivariate Logistic regression |
SHCN: 31.1% baseline; 31.4# wave 3; 33.5% wave 4; 50.3% at any point (ever). Predictors of SHCN: male (OR: 1.8); young age (OR: 0.6); poverty (OR: 1.5); adopted (OR: 2.6); foster (OR: 2.1). |
| Rates of Part C Eligibility for Young Children Investigated by Child Welfare | Rosenberg , S.A. Smith, E.G. | Topics in Early Childhood Special Education , Vol. 28, pp.68-74. 2008 | Caseworker interview Caregiver interview Children Assessment Baseline | Developmental delays; early intervention; child maltreatment | Children who were younger than 3 years old at baseline (n= 1,997). | Examined association between child eligibility to Part C early intervention services (based on child performance on developmental assessment) and substantiation status, type of maltreatment, and placement setting. Univariate and bivariates (chi squares) analysis. | Developmental scores less or equal to -1 SD for 44.6% of the children in the Battelle Developmental inventory, 39.6% on the Preschool Language Scale-3, and 30.0% on the Vineland Adaptive Behavior Scale. Using the criteria of at least 2 scores <= -1.0 SD or one score <= -1.5 SD, 47% of children were classified as in need of Part C early intervention services. No significant differences in eligibility were found by substantiation, main type of maltreatment, and placement setting. |
| Impact of kinship care on behavioral well-being for children in out-of- home care |
Rubin, D.M. Downes, K.J. O'Reilly, A.L.R. Mekonnen, R. Luan, X. Localio, R. |
Archives of Pediatric and Adolescent Medicine Vol. 162, pp. 550-556 2008 | Caseworker interview Caregiver interview Baseline Wave 3 Wave 4 |
Family foster care, health service use, national survey, Welfare, kin, predictors, attitudes, youths | Children entering out-of-home care following maltreatment report (n=1,309) | Examined predicted probabilities of behavioral problems among kin and foster care children Bivariates, logistic regression (Stata). | 50% of children started in kinship care and 17% of children who started in foster care later moved to kinship care. Children in kinship care were at baseline more likely to those in foster care to live below the poverty line (44% vs. 23%), less likely to have behavioral problems (33% vs. 42%), less likely to use prescription medication (0.6% vs. 3.2%), less likely to use mental health services (24% vs. 35%). At follow-ups children in kin care compared to those in foster care were less likely to have behavioral problems (defined as standardized scores over 59; Wave 3: 31% vs. 47%; Wave 4: 29% vs. 48%). Controlling for a child's baseline risk, placement stability, and attempted reunification to birth family, the estimate of probability of behavioral problems at 36 months was significantly lower for children in early and late kin care compared to foster care: 32% for children assigned to early kinship care, 39% for children in late kin care and 46% for children assigned to foster care only (P=.003). Children with unstable placements had higher probability of behavioral problems than children in stable placements : 49% unstable, 32% stable (P= .007) |
| Maltreated Infants: Reported Eligibility for Part C and Later School-Age Special Education Services. | Scarborough, A. A. McCrae, J. S | Topics in Early Childhood Special Education Vol.28, pp.75-89 2008 | Caregiver Interview Caseworker interview Children Assessment Baseline Wave 2 Wave 5 |
Maltreated infants; Part C services; special education services | Children who were 0 to 12 months at baseline (n= 1,196). IFSP analysis restricted to 630 children with caseworkers information on IFSP. IEP analysis restricted to 612 children with teachers information on IEP | Examine factors associated with the report of the child having an IFSP at Wave 2 and an IEP at wave 5. Analysis: Bivariates (chi squares) and hierarchical logistic regression analysis to examine assotiatins with special education services at wave 5. |
31% of infants were reported by caseworkers to have an IFSP. Children who had an IFSP were more likely to have substantiated maltreatment than those who did not have an IFSP (65% vs. 42%). No significant differences were found between those who had an IFSP and those who didn’t on measures of development 54 months post baseline. 20% were reported to have an IEP at 54 months. IEP was associated with being male (OR=3.2, 95% CI= 1.3-7.8), having a caregiver with cognitive impairment (OR=15.3, 95% CI= 2.2-105.4), and having a caregiver with mental illness or depression (OR=0.23, 95% CI= 0.1-0.9). |
| A longitudinal study of the effects of child maltreatment on later outcomes among high- risk adolescents | Tyler, K.A. Johnson, K.A. Brownridge, D.A. | Journal of Youth and Adolescence Vol. 37, pp. 506-521 2008 | Caseworker interview Caregiver interview Child interview Baseline Wave 3 Wave 4 |
Child maltreatment; running away; adolescents; well-being | Children 11 to 14 at baseline living with a permanent caregiver and less than 17 and enrolled in school at Wave 4 (n=360) | Examined the effects of child maltreatment, parenting and disadvantaged neighborhood on victimization, delinquency and well-being via running away and school engagement. Univariate Bivariates (correlations) Path analysis (MPlus 3.13) |
Positive parenting at Wave 1 associated with school engagement (B:.15, p < .001) and negatively associated with running away at Wave 3 (B: -.42, p < .01). Running away associated with delinquency at wave 4 (B:.39, p < .001). School engagement associated with less delinquency (B:-.61, p < .001) and better well-being (B:1.90, p < .01). Positive parenting at wave 1 had a direct effect on well-being at wave 4 (B:.58, p < .01). Disadvantage neighborhood at wave 1 had a direct negative effect on well-being at wave 4 (B:-.48, p < .05). Gender significantly moderated these associations: effect of positive parenting on well-being was stronger for females. |
| Kinship Care and Nonkinship Foster Care: Informing the New Debate | Barth, R.P. Guo, S. Green, R.L. McCrae, J.S. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker interview Caregiver interview Child interview Baseline Wave 3 | Kinship; foster care; child well-being; parenting | Children in foster and kin care who remained in out of home care at least 75% of the time from baseline to wave 3 (n=567). | Examined differences in child development outcomes, and parenting among children in foster and kin care. Propensity score matching, Bivariates (Chi square and t-tests) |
Before PSM: children in kin care had better scores than children in foster care in social skills (101.4 vs. 83.9), daily living skills (11.7 vs. 91.7), and mathematics (102.1 vs. 98.6). After PSM: children in kin care had lower clinical scores than children in foster care related to trauma. Change after 18 months: using the PSM sample, children in kinship care improve their externalizing behavioral problems while children in foster care worsened. . |
| Predictors of placement moves among children with and without emotional and behavioral disorders | Barth, R. P. Lloyd, E. C. Green, R. L. James, S. Leslie, L. K. Landsverk, J. | Journal of Emotional and Behavioral Disorders Vol. 15, pp. 46-55 2007 | Caseworker Interview Child Interview Caregiver Interview Baseline Wave 2 Wave 3 Wave 4 |
Child welfare; Out-of-home placement; Placement stability; Mental health; Emotional and behavioral disorders | Children aged 7-14 who were in out-of-home care at baseline (n= 725) | Explored number of out-of-home care placements for children with and without emotional and behavioral disorders (EBD). Univariates Bivariates (Chi square) Poisson regression | Emotinal and behavioral Disorders: 50.4% Predictors of placement among children with EMD: depression (IDR: 1.3); not living with siblings (IDR: 1.5) Predictors of placement among children without EMD: child 11 or more (IDR: 1.37); female (IDR: 1.5) |
| Racial Disproportionality, Race Disparity, and Other Race-Related Findings in Published Works Derived from the National Survey of Child and Adolescent Well-Being | Dunbar, K. Barth, R.P. | Report The Annie E. Casey Foundation; Casey Family Services } 2007 | NA | Race/ethnicity | NA | Summarizes published and in-press articles and chapters based on NSCAW in order to examine the relationship between race/ethnicity and child-well-being, permanency, safety, and services. | Race/ethnicity not a significant predictor of services receipt for children remaining at home, nor of whether children would be placed in out of home care For the whole group, there were significant differences in reunification and services received. African American infants were less likely to experience reunification than white infants. African American and Other youth over 10 years of age were less likely to be reunified than white youth. Black children are less likely to receive developmental services than white even when controlling for need. African American and Hispanic children less likely to receive specialty mental health services than white children. |
| Physical abuse and adolescent development | Eckenrode, J. Izzo, C. Smith, E. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker interview Caregiver interview Child interview Baseline Wave 2 Wave 3 Wave 4 |
Physical abuse; adolescents; developmental outcomes | Youth eleven and older at baseline (n=1,179) | Developed a definition of physical abuse based on multiple sources of information. Examined consistency across reporters and developmental outcomes of physical abuse. Univariates Bivariates Regression analysis (linear and logistic regression) |
Prevalence on physical abuse: 9.2% based on caseworker report; 15.8% based on parent report; 27.1% based on youth report; 40.7% based on all sources combined. Agreement among sources about absence of abuse: 62.9% to 70.7%; agreement about presence of abuse: 16.0% to 32.9%. Caseworker report of physical abuse was associated with youth externalizing problems at wave 1. Counter to expectations, caseworker report of physical abuse was associated with greater math achievement. Caregiver report of physical abuse was associated with youth externalizing and internalizing problems; tobacco, drug, and alcohol use, and sexual activity at wave 1. Youth report of physical abuse was associated with youth externalizing and internalizing problems; delinquency; tobacco, drug, and alcohol use; sexual activity; trauma symptoms; depression; lower future expectations; and lower school engagement at wave 1. Effects of physical abuse varied by age and gender: older children had more developmental areas compromised; boys were more likely to have sexual activity and girls more likely to have internalizing problems. Stability of effects: most of the associations became attenuated over time. Associations remain significant only for behavioral problems and sexual activity by Wave 4. |
| Using High-Quality Research to Improve Child Protection Practice: An overview | Haskins, R. Wulczyn, F. Webb, M.B. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | NA | Child maltreatment, services, NSCAW | NA | NA | Describes history of NSCAW and the survey methods, and summarizes findings presented in the book on child and family problems, services and interventions, and outcomes of child protection. |
| Intimate Partner Violence in the Child Welfare System: Findings from the National Survey of Child and Adolescent Well-Being | Hazen, A.L. Connelly, C.D. Kelleher, K. Landsverk, J.A. Barth, R.P. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caregiver interview Caseworker interview Baseline | Intimate partner violence, child behavioral problems | Female caregivers of in-home children (n= 3,612) Children 2 and older (n=2,020) | Examined correlates of physical IPV, and the association between IPV and child behavioral problems, and the moderating effects of parenting. Univariates Logistic regression |
Lifetime IPV: 44.8% Past year IPV: 29.0% Correlates of severe IPV: Caregiver major depression (OR: 2.6), caregiver drug dependence (OR:2.4) and history of prior reports of child maltreatment (OR:1.8). Caregiver age was associated with decrease of IPV (OR:0.96). Correlates of less severe IPV: Presence of male partner (OR:2.0) and caregiver major depression (OR:1.88). Caregiver alcohol dependence was associated with decrease of IPV (OR:0.27). Child externalizing (B=2.17, p < .05) and internalizing (B=2.37, p < .01) problems were associated with severe IPV. |
| Building on Strengths: Current Status and Opportunities for Improvement of Parent Training for Families in Child Welfare | Hurlburt, M.S. Barth, R.P. Leslie, L. Landsverk, J.A. McCrae, J. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Agency child welfare key informants (CCCW) Caseworkers interview Baseline Wave 2 | Parent training | In-home children in families that received some type of child welfare services (n= 2,308) | Examined parent training. Bivariates. Logistic regression. |
Parent training: 94% of counties delivered parent training to families with identified need; 50% primary delivered in family’s home; 46% primarily delivered in agency; 84% received parent training in groups that included families not in the CWS; among the most commonly used programs were not EBP (1.4% Incredible Years, 0.2% Parent-Child Interaction Therapy). Predictors of Parent training: poor parenting practices (OR: 1.4); active domestic violence (OR: 1.9); Abandonment (OR: 6.9); higher risk/harm/evidence (OR: 1.2) |
| Sensitive, stimulating caregiving predicts cognitive and behavioral resilience in neurodevelopmentally at-risk infants | Jaffee, Sara R. | Development and Psychopathology Vol.19, pp. 631-647 2007 | Caregiver Interview Child Assessment Baseline Wave 3 | Child maltreatment; Maternal behavior; Prenatal exposure; Foster care; Neglect; Stress |
children aged 3 to 24 months at Wave 1 (n= 1,720 | Explored whether sensitive, stimulating caregiving would promote positive behavioral and cognitive outcomes among children who were at risk. At risk status was based on the results of a neurodevelopmental screener and a temperament inventory. Correlations, logistic and hierarchical regression analysis were used. | Resilient children were identified as those who were doing better than expected with language development and behavior problems at Wave 3 given early-emerging neurodevelopmental and temperamental vulnerabilities. Improvements in the amount of cognitive stimulation provided to children by caregivers were associated with better than expected language scores at Wave 3. Similarly, improvements in caregivers' emotional support were associated with lower than expected levels of problem behaviors in children at Wave 3. |
| Social, Emotional, and Academic Competence Among Children Who Have Had Contact With Child Protective Services: Prevalence and Stability Estimates | Jaffee, S. R. Gallop, R. | Journal of the American Academy of Child and Adolescent Psychiatry Vol. 46, pp. 757-765 2007 | Caregiver interview Caseworker interview Child interview Teacher interview Baseline Wave 3 Wave 4 |
Maltreated children, resilience; adolescence; reliability; outcomes; welfare; abuse |
Children 8 years old or older at baseline (n= 2,065) | Explored the prevalence and stability of social, emotional, and academic competence. Bivariates, unadjusted logistic regression (Stata 8.0) | 38% to 46% had a normal value in at least one area at wave 1, 44% to 46% at wave 3, and 37% to 49% at wave 3. At all waves, 14% to 22% had a normal value in at least one area. 11% to 14% had a normal value in all areas at all waves. |
| Unsuccessful In-Home Child Welfare Service Plans Following a Maltreatment Investigation: Racial and Ethnic Disparities | Kohl, P. | Report The Annie E. Casey Foundation; Casey Family Services 2007 | Caseworker interview Caregiver interview Child interview Baseline Wave 2 Wave 3 Wave 4 |
Race/ethnicity; recurrent maltreatment; out of home placement | Children who remain at home at baseline (n= 3,900) | Examined racial/ethnic differences in recurrent maltreatment and subsequent placements in out of home care. Univariate Bivariates (chi square) Logistic regression |
Re-reports by Wave 4: 27.8% Unsuccessful CWS plan (defined as re-report or new OH placement): 33.5%. Hispanic children more likely to have a rereport than African American or White children (35.5% vs. 27.0% and 24.6%). Compared to Kids Counts estimates, African American children were over represented in OH placement (they are 15.0% of the general population of children, but represent 35.4% of the children in OH). Caregivers of African American children more likely to report use of severe violence than caregivers of White children (baseline: 18.1% vs. 5.6%; ober 36 months: 18.9% vs. 9.0%) Predictors of unsuccessful CWS plan: mental health problems of caregiver reported by caseworker (OR: 1.4); family living at or below poverty level (OR: 1.4); urbanicity (OR: 1.4); previous maltreatment report at baseline (OR: 2.4). Predictors of OH placement: race/ethnicity not significant; previous maltreatment report at baseline (OR: 2.9). Predictors of Caregiver report of severe violence: child race African American (OR: 2.3); mental health problems of caregiver reported by caseworker (OR: 2.2); urbanicity (OR: 2.0). |
| Child maltreatment recurrence among children remaining in-home: predictors of re-reports | Kohl, P. Barth, R.P. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker interview Caregiver interview Child interview Baseline Wave 2 Wave 3 |
Re-reports; domestic violence; child behavior | Children who remained at home after index investigation (n= 3,143) | Examined predictors of maltreatment recurrence. Univariate Bivariate Logistic Regression |
Re-reports baseline to Wave 3: 23.0%; range: 1-10. Among all children 13% had one re-report, 6.8% had two, 1.6% had three, and 1.6% had more. Among those with a re-report, 38.4% were substantiated. Re-reports for families receiving CWS services: 32.4%, significantly higher than re-reports for families with closed cases (19.6%). Children in urban areas more likely to be reported than those in rural areas (25.1% vs. 16.9%). Logistic regression model. Predictors of re-reports: In-home CWS services (OR: 1.5); child behavioral problems (OR: 1.03); prior reports of maltreatment (OR: 2.3); domestic violence (OR: 5.0); caregiver history of abuse and neglect (OR: 5.4); family financial difficulties (OR: 1.6); urbanicity (OR: 1.8); receipt of parenting support services (OR: 1.2) |
| Systems Integration and Access to Mental Health | Landsverk, J.A. Hurlburt, M.S. Leslie, L.K. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker Interview Caregiver Interview Agency key informant (CCCW) 1999 Area Resource File (ARF) Baseline Wave 2 | Mental health; Mental health services; Behavioral problems | Agencies with data for CCCW (undisclosed n) | Presents findings from two previously published papers that used data from NSCAW and CCCW. | Assessment of physical health problems among children entering foster care: 94% Assessment of mental health problems among children entering foster care: 48% Provision of comprehensive physical, mental health, and developmental examination: 43%. Almost none agency reported assessment protocols for children not placed in out-of-home care. Children with behavioral or emotional problems more likely to receive services if there was increased coordination between the CWS and mental health agencies. |
| Alcohol, Drug and Mental Health Service Need for Caregivers and Children Involved with Child Welfare | Libby, A.M. Orton, H. D. Barth, R. P. Burns, B.J. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Child Interview Caregiver Interview Caseworker Interview Baseline Wave 3 Wave 2 data used to fill in gaps in Wave 3 data. |
Behavior problems; Mental health; Substance use; Mental health services; Services for substance use |
Caregivers and Children 2-14 years at baseline whose caregivers were the same in both rounds (n = 1,876) | Examined co-occurrence of caregiver alcohol, drug, and mental health (ADM) problems with children’s behavioral problems as well as the relationship between children’s behavioral problems and caregiver services for ADM. | Prevalence of children’s baseline behavioral problems as a function of caregiver ADM problems was 54.8% in comparison to 28.6% for children whose caregivers had no ADM problems. The highest rate of children’s behavioral problems was found for children of caregivers with baseline substance dependence as measured by the CIDI-SF — 51.9% had elevated externalizing CBCL scores. Highest prevalence of children with elevated internalizing scores (34%) was found for caregivers with serious mental health problems at baseline as judged by caseworker. Multivariate analyses indicated that children with significantly elevated externalizing scores at baseline were more than 3 times as likely to have a caregiver with a baseline ADM problem — probably mediated through impaired parenting. Youngest children were 1.8 times more likely to have a caregiver with a baseline ADM than oldest children. Caregivers with impaired parenting skills were more than 3 times more likely to have baseline ADM problems. Analyses examining service receipt among caregivers who had baseline ADM problems indicated: caregivers of children with externalizing problems were more than 3 times as likely to receive mental health services; caregivers of children who were in-home at baseline were 70% less likely to receive services for substance use; caregivers of the youngest children were 80% less likely to receive substance use services; caregivers of female children were 50% less likely to receive mental health services; Hispanic caregivers were 11 times more likely than caregivers from other ethnic groups to receive substance use services; Black caregivers were 80% less likely to receive mental health services than caregivers from other ethnic groups. The finding that caregivers of children at home were less likely to receive mental health services is a cause for concern. |
| Mental Health and Substance Abuse Services to Parents of Children Involved with Child Welfare: A study of Racial and Ethnic Differences for American Indian Parents | Libby, A. M. Orton, H. D. Barth, R. P. Webb, M. B. Burns, B. J. Wood, P. A. Spicer, P. | Administration and Policy in Mental Health and Mental Health Service Research Vol. 34, pp. 150-159 2007 | Caregiver Interview Caseworker Interview Baseline Wave 2 | Child welfare; Mental health; Substance abuse; American Indian; Disparities | Caregivers who were the same between baseline and wave 2 (n = 3,425) | Examined the extent of mental health and substance abuse problems as well as disparities in health care among caregivers as a function of ethnicity. Analysis: Multivariate logistic regression was used to examine the effect of baseline caregiver and child characteristics and caregiver risk factors on caregiver service receipt of substance use and mental health problems at 18 months post-baseline. | In comparison to other racial/ethic groups, American Indian (AI) caregivers had the highest prevalence of mental health (MH) and emotional problems (23.4%). AI caregivers had lower prevalence (7.5%) of substance use (SU) than Whites or Blacks. Among caregivers with MH problems at baseline, only 5.2% of AI caregivers received a formal assessment, which was the lowest of any group; of those who had an assessment, nearly all were considered to have a serious or moderate impairment (98.9%). AI caregivers were the least likely to receive services of any group (3.6%) though they were not the least likely to be referred for services. Hispanic caregivers were the most likely to receive a formal assessment, have a referral, and receive MH services. AI caregivers were the most likely to receive an assessment for a SU problem and to receive a referral and services—almost twice as many AI caregivers were referred to SU services as compared to White, Black, and Hispanic caregiver. Overall, 36% received some kind of SU service, usually outpatient services. Unlike other racial groups, no AI caregivers received inpatient services for either MH or SU problems, intensive day treatment for MH problems, or detox for SU problems. Among those with baseline MH and/or SU problems, race and age of child were associated with receipt of MH services. AI caregivers were significantly less likely to receive MH services than were White (OR 7.4), Black (OR 3.1), or Hispanic (OR 10.5) caregivers. Caregivers with preschool children were 4 times more likely to receive MH services than caregivers with an adolescent. Having only a SU problem decreased chances of receiving services (OR .36). Caregivers with a child in-home were less likely than those with an out-of-home placed child (OR .43) to receive SU services and having only a MH problem decreased chance (OR .09) of receiving SU services. Findings indicate that SU problems receive more attention by CWS despite need for MH services. |
| Medicaid and Mental Health Care for Children in the Child Welfare System | Raghavan, R. Leibowitz, A. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker Interview Caregiver Interview Agency key informant (CCCW) 1999 Area Resource File (ARF) Baseline Wave 2 | Medicaid; Mental Health; Ambulatory care; Inpatient care; Services need; Services Use; Behavioral Problems; Medicaid managed care; Medicare carve-outs |
Children 2 years and older with identified county of residence (n= 3,460) | Examined effects of various Medicaid policies on the delivery of outpatient and inpatient mental health services. Bivariate analysis (chi square, logistic regression) Multivariate Logistic Regression. | Behavioral problems in clinical range: 36% Use of Ambulatory MH Services: 17.6% Inpatient MH services: 1.9% Ambulatory MH Services predictors (multivariate results): older age (OR: 1.1); out-of-home (OR: 1.9); uninsured (OR:0.4); caregiver college educated (OR: 1.8); behavioral problems in clinical range (OR: 3.8); caseworker identified need for services (OR: 2.5); child resident of counties with greater numbers of child psychiatrists (OR: 2.1). Medicaid variables not significantly associated with use of ambulatory MH services. Inpatient MH services predictors (multivariate analysis): older age (OR: 1.3); Native American and mixed race (OR: 0.2); caregiver college educated (OR: 0.2); behavioral problems in clinical range (OR: 4.6); caseworker identified need for services (OR: 7.7)child resident in county that carved out MH services (OR: 0.4) |
| Identifying Young Maltreated Children with Developmental Delay | Rosenberg , S.A. Smith, E.G. Levinson, A. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caregiver interview Caseworker interview Child assessment Baseline | Developmental delay, early intervention services | Children less than 3 years old with a substantiated or indicated investigation (n=1,138). | Examined rated of development delay. Univariates | Developmental delay based on child assessment (defined as one measure 1.5 SD under the mean or 2 measures 1.0 SD under the mean): 46% of children. Developmental delay based on caseworker report: 23%. |
| Placement Stability and Early Behavioral Outcomes among Children in Out-of-Home Care | Rubin, D.M O’Reilly Hafner, L. Luan, X. Localio, R. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker interview Caregiver interview Baseline Wave 3 Wave 4 |
Placement stability; permanency outcomes; behavioral problems | Children residing at home at investigation, subsequently placed in out-of-home placement that continued past wave 1 (n=1,099) | Examined the relationship between placement stability and child well-being. Bivariates | Placement stability: 33.5% reunified; 33.9% early stability in foster care; 12.2% late stability in foster care; 20.5% unstable in foster care. Mean number of placements by wave 4: 3.2, range: 1-18.Multiples placements associated with child older age. Children who began the study with a CBCL score in the normal range were less likely to stay in the normal range if they had unstable placement by wave 3 and 4. |
| The Impact of Placement Stability on Behavioral Well-being for Children in Foster Care | Rubin, D. M. O'Reilly, A. L. R. Luan, X. Q. Localio, A. R. | Pediatrics Vol. 119, pp. 336-344 2007 | Caregiver Interview Caseworker Interview Baseline Wave 3 | Child behavior; Child Behavior Checklist; Cohort studies; Outcome assessment; Foster care |
Children living at home initially who were placed in foster care or 18 > months (n = 729) | Examined whether placement stability influences child outcomes irrespective of a child’s attributes and problems. Analysis: Propensity analysis used to categorize children’s risk of placement instability which was used in a logistic regression to examine association with behavior well-being. |
At the 18-month follow-up (Wave 3), 52% of children achieved early placement stability (placement within 45 days lasting through 18 month interview), 19% achieved late stability (placement after 45 days that lasted through 18 month interview), and 28% remained unstable (no long-lasting placement maintained for > 9 months). Early stabilizers were more likely to be young and have no previous CPS history; borderline associations were found between early stability and normal baseline behavior and birth parents without serious mental/behavioral problems. Predictors of children’s behavioral outcome at 18-months included baseline behavioral problems and placement stability. Other predictors of better outcomes included younger age, no previous CPS history, and parents with no drug or alcohol problems. Risk of placement instability was estimated for each child and children were divided into low-, medium-, and high-risk groups; risk groups were predicative of placement instability and behavioral problems. Children with unstable placements had twice the odds of having behavior problems as early stabilizers at every level of instability risk. Findings indicate that independent of baseline behavior problems, placement stability is related to out-of-home children’s subsequent behavior problems with a 36% to 63% elevated risk. |
| Initial Construction of an Actuarial Risk Assessment Measure using the National Survey of Child and Adolescent Well-Being | Shlonsky, A. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker interview Baseline Wave 2 Wave 3 Wave 4 |
Risk assessment tool; rereports; resubstantiation. | In-home children at baseline.(n=2,401) | Developed a risk assessment tool for use by caseworkers for predicting maltreatment recurrence. Bivariates. Logistic regression. Survival analysis. |
Re-reports: Wave 2: 15.0; Wave 3: 22.2; Wave 4: 26.8 Predictors of rereports and resubstantiation: prior history of child welfare reports, prior investigations of maltreatment; prior history of child maltreatment; child special health care needs and behavioral problems; active alcohol use by parents; parents unrealistic expectations of the child; parent’s history of abuse and neglect; high stress in the family. Predictors of re-reports: prior CPS history (OR: 2.2); presence of another supportive caregiver (OR: 0.61); high stress in the family (OR: 1.8); very young children (OR: 1.7) Predictors of resubstantiation: child age younger than 4 (OR: 1.5); high stress in the family (OR: 1.7); caregiver with poor parenting skils (OR: 2.2) |
| Substance Use in Maltreated Youth: Findings from the National Survey of Child and Adolescent Well-Being | Wall, A. E. Kohl, P. L. | Child Maltreatment Vol 12, pp. 20-30 2007 | Child Interview Caregiver Interview Caseworker Interview Baseline | Adolescent substance use; Maltreatment; Risk; Protection |
Children 11-15 years of age (n = 1,179) | Examined association between different levels of substance use and demographics, placement type, youth and family characteristics. Analysis: Chi square used to examine associations of demographics, placement type, and youth and family characteristics with youth substance use. Logistic regression used to examine factors associated with increased odds of moderate/high substance use. | Reported levels of substance use were: 71% none, 20% low, 3% moderate, and 6% high. Association between substance use and both age (older greater use) and race/ethnicity (greater proportion of no use reported by African American than White youth). Substance use levels varied by level of conduct problems (greater proportion of youth with conduct problems reported high use than youth without conduct problems) and by caregiver relatedness (higher levels of substance use reported in youth reporting lower caregiver relatedness). Odds of moderate/high use were higher for (1) youth whose initial report was for physical abuse as compared to those reported for sexual abuse, failure to provide, or failure to supervise; (2) youth with conduct problems as compared to those without conduct problems; (3) low monitored youth as compared to higher monitored youth. No relationships detected between substance abuse and academic achievement or engagement or between youth and caregiver substance use or placement type. |
| Predictors of Reunification | Wildfire, J. Barth, R.P. Green, R.L. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker interview Caregiver interview Child interview Baseline Wave 2 Wave 3 |
Reunification; out-of-home care | Children who entered in out-of-home care between baseline and wave 2 (n= 1,568) | Examined characteristics of children who are reunified with their biological families, and activities of child welfare agencies and permanent caregivers that influence the rate of reunification. Bivariates Survival analysis (multivariate analysis based on Cox Proportional Hazard Models) | Reunification Rate: 30% Bivariate predictors of reunification: parent completed case plan goals (Reunified: 55%, not reunified: 27%); higher services dose (Reunified: mean 44 sessions, not reunified: mean 25 sessions); child self-reported delinquency (Reunified: mean 44, not reunified: mean 12). Multivariate predictors of reunification: Children 0-6 months: Black (RRR: 0.42, ref. White); Neglect (RRR: 0.2, ref. physical abuse); parenting support (RRR: 6.7); compliance with case plan (RRR: 5.4%) Children 7-35 months: Male (RRR: 1.78); compliance with case plan (RRR: 4.5). Children 3-5 years: Sexual abuse (RRR: 4.2, ref. physical abuse). Children 6-10 years: Male (RRR: 2.2); behavioral or emotional problems (RRR: 0.3); active substance use by permanent caregiver (RRR: 0.5); non-kin placement (RRR: 4.5, ref. kin care) Children 11 and older: Black (RRR: 0.14, ref. White); Other race (RRR: 0.13, ref White); placed in own neighborhood (RRR: 2.0) |
| Addressing the educational needs of children in Child Welfare Services | Webb, M.B. Harden, B.J. Baxter, R. Dowd, K. Shin, S.H. | Child Protection: Using Research to Improve Policy and Practice (Ron Haskins, Fred Wulczyn, and Mary Bruce Webb, Eds.) Washington , D.C. : Brookings Institution Press 2007 | Caseworker interview Caregiver interview Child interview Baseline Wave 3 | Special education services; need for services; referrals | Children of elementary school age receiving child welfare services (n=1,760) | Examined cognitive and social emotional functioning, receipt of special education services, and caseworker’s identification and referral of children to services. Univariates Bivariates (chi-square) |
Need for services defined as – 2 SD under the mean. Cognitive need: 7.3% Behavioral problems: 16.4% Cognitive and behavioral problems: 6.6% Any problems: 30.3% Need identified by caregiver: 31.3% Bivariate analysis: need associated with being male, older age, repeated grade, leaving in rural areas, and failure to provide. Of children identified by assessment as in need, 27.4% were already receiving services, 16.1% were referred to special education by caseworkers, and 14.9% received services. 57.7% of those in need did not receive services after close of an investigation. As reported by caregivers, 54.4% received special education. |
| Substance abuse treatment and the recurrence of maltreatment among caregivers with children living at home: A propensity score analysis | Barth, R.P. Gibbons, C. Guo, S.Y. | Journal of Substance Abuse Treatment Vol. 30, pp. 93-104 2006 | Caregiver Interview Caseworker Interview Baseline Wave 3 | Child welfare, reabuse, substance abuse treatment, propensity score | Caregivers with substance abuse problems (before PSM: n= 1,101, after PSM n=448) | Explored receipt of substance abuse services and association with re-reports for child maltreatment in the following 18 months. Bivariates (chi square), logistic regression, .propensity score mathching, survival analysis | By wave 3, 19.3% of those receiving services had a child abuse re-report, compared to 8.6% of untreated caregivers of the PSM group (Hazard ratio: 1.95) p < .01).Control variables positively associated with rereports: caregiver mental health problems, poverty. |
| Placement Into Foster Care and the Interplay of Urbanicity, Child Behavior Problems, and Poverty | Barth, R. P. Wildfire, J. Green, R. L. | American Journal of Orthopsychiatry Vol. 76, pp. 358-366 2006 | Caregiver Interview Caseworker Interview Baseline | Foster care; Poverty; Child behavior problems |
All children in CPS cohort (n = 5,504); mental health analyses restricted to those > 2 years (n = 3,798) | Examined the extent to which CWS placement decisions for children are mediated by poverty, the need for mental health services, and type of setting (urban/nonurban). Analysis: bivariate relationships were examined and confirmed using logistic regression. | Overall 24% of the sample had trouble paying for basic necessities; 49% of the sample of caregivers of children entering out-of-home care had trouble paying for basic needs as compared to 21% of those whose children remained at home (34% of those in home received CWS services were classified as having trouble paying for basic needs). Families of children 0-2 in out of home care were the poorest (62%). Very poor children in urban settings more likely to enter placement than non-poor urban children; no differences as a function of poverty were found for children in non-urban settings. Logistic regression indicated that in urban settings, difficulty paying for basic needs significantly contributes to placement as is child’s age (3-5 years least likely) and parent mental illness (MI), substance abuse(SU) or domestic violence (DV). In non-urban settings, child behavior problems and parent MI/SU/DV were related to placement. Results show that there are a sizable proportion of cases in which children are receiving CW services not because of child abuse by unfit parents. |
| Accuracy of Caregiver Identification of Developmental Delays Among Young Children Involved with Child Welfare | Berkoff, M. C. Leslie, L. K. Stahmer, A. C. | Journal of Developmental and Behavioral Pediatrics Vol. 27, pp. 310-318 2006 | Caregiver Interview Caseworker Interview Child Interview Baseline | Child welfare; Developmental delays; Sensitivity and specificity |
Caregivers of children 0 to 5 years at baseline who had never been tested for develop-mental disabilities while in the care of caregiver who was interviewed (n = 2,189) | Examined the accuracy of identification of developmental delay among caregivers of children involved with CW and whether foster caregivers were better able to identify such delays than other caregivers. Analysis: Chi-square tests used to assess relationship among variables. Logistic regression was used to examine multivariate relationships between independent variables and caregiver identification of developmental delay. |
Differences between caregiver types (in-home, kinship, foster care, and other) were found for child age (foster parents had younger children), insurance types (foster parents had the highest rates of Medicaid), rates of neglect (foster and kinship caregivers had highest rates of kids who experienced neglect), and delays in adaptive functioning (children with adaptive delays most likely to be in foster care). Caregiver type was associated with age, education, and neighborhood perception of drug use. 25% of caregivers identified a developmental delay; there was an association between caregiver type and identification with foster parents most likely to identify developmental delays and in-home caregivers least likely. Sensitivity of caregiver identification of a delay in any domain was 35%; foster parents and other caregivers were more accurate than in-home parents. Sensitivity of identification of language and cognitive delays were 36% and 35%; foster parents were more accurate than in-home for identifying language delays. Sensitivity of identification of adaptive delays was greater (53%); other caregivers were more accurate than kinship or in-home caregivers. Specificity of identification of any delay was 84%, with no differences by caregiver type. Specificity of identification of language and cognitive delays was 78%; in-home and kinship caregivers were more accurate in identifying no delay than foster caregivers for both domains. Specificity of adaptive behavior delays was 81%; in-home were more accurate than were foster parents. After adjusting for child’s age, race, and insurance type as well as caregiver health status and perception of community, foster caregivers had 6.7 times the odds of correctly identifying a child with any delay as compared to in-home caregivers. |
| Persistence of Intimate Partner Violence Among Families Referred to Child Welfare | Connelly, C. D. Hazen, A. L. Coben, J. H. Kelleher, K. J. Barth, R. P. Landsverk, J. A. | Journal of Interpersonal Violence Vol. 21; pp. 774-797 2006 | Caregiver Interview Baseline Wave 3 | Intimate partner violence; Longitudinal; Child welfare; Predictors; Race; Ethnicity |
Caregivers whose children were living at home at baseline and wave 2 in which a history of physical IPV was reported at baseline in the previous 12 months. (n = 861). | Examined the longitudinal course of intimate partner violence (IPV) reported by maternal caregivers, the caregiver and environmental characteristics that differentiate cases in which IBP diminished or continued based on baseline violence severity. Analysis: polychotomous logistic regression used to examine the association between severe and minor IPV with sociodemographic variables and caregiver functioning. | At baseline, 40.8% of the caregivers reported victimization by minor physical assault and 59.2% reported severe physical assault; at Wave 3 23.2% reported severe physical violence, 16.3% reported minor physical violence, 60% reported no physical violence in the previous 12 months. IPV abates with time for most women: among those who reported severe IPV at baseline, 27.3% reported severe violence, 11% reported minor violence, and 62% reported no violence in the 12 months prior to Wave 3. Among those who reported minor violence at baseline, 15% reported severe violence, 18% reported minor violence, and 67% reported no violence in the 12 months prior to Wave 3. Variables associated with severe violence at Wave 3 include race/ethnicity (AA women 2 times greater than White women) and marital status (married women had higher odds than separated, divorced, or widowed women). Variables associated with minor violence at Wave 3 were race/ethnicity (Hispanic women and “other” women had lower odds than White women), severe violence at baseline (these women had lower odds at Wave 3), and living with a partner at baseline only. The correlates of IPV differed by race/ethnicity. White women who reported severe IPV at baseline had 3 times the odds for reporting severe IPV at Wave 3 than did those reporting minor IPV. African American women living with a partner at one or both time points had greater odds for experiencing minor IPV at Wave 3 and those who were older had decreased odds for IPV at Wave 3. In Hispanic women, older age was associated with increased risk for severe IPV. No relationship between violence patterns and depression, substance use, education, number of children, social support or community environment. Limitations include lack of other forms of IPV, and possible underreporting of substance use. |
| Between Two Systems: Children in TANF Child-only Cases with Relative Caregivers | Gibbs, D. Kasten, J. Bir, A. Duncan , D. Hoover, S. | Children and Youth Services Review Vol. 28, pp. 435-446 2006 | Child Interview Caregiver Interview | TANF program; Relative caregivers; Kinship; Foster care; Well-being; Case studies |
Children in out of home care: with relative caregivers receiving TANF child-only (TANF n = 54), and no TANF (KINCARE n = 456), and in foster care (FOSTER n = 565) | Examines whether children in TANF child-only with relative caregivers differ from others in out of home care and the extent to which children receive adequate support. In addition, case study data from 5 states were used to describe and explain practices and policies for children in TANF relative care caregiver child-only. Analysis: Chi square to examine differences in groups. | Children in relative child care with TANF were more likely to have had a hearing test than FOSTER, less likely than FOSTER and KINCARE to have been admitted to a hospital, less likely than FOSTER to have been diagnosed with a learning problem, less likely than FOSTER and KINCARE to have received special ed. They had significantly higher ratings on social skills in preschool than KINCARE and FOSTER, but more behavioral problems on ratings than KINCARE on the CBCL in school-aged children and self-report of behavior problems in adolescence. Case study indicated that children in TANF only cases with relative caregivers have extensive service needs and a lack of case management—many of these are outside the child welfare system. |
| Propensity score matching strategies for evaluating substance abuse services for child welfare clients. | Guo, S. Y., Barth, R. P. Gibbons, C. | Children and Youth Services Review, Vol. 28 , pp. 357-383. 2006 | Caseworker interview Caregiver interview Baseline Wave 3 | Causal effects; nonexperimental evaluation; propensity score matching; substance abuse treatment; child welfare services | Children who lived at home whose primary caregiver was female (n= 2,758) | Examined use of propensity score matching strategies to study substance abuse treatment effect on children’s re-reports of maltreatment | 10.8% of caregivers received substance abuse treatment (SAT). Original sample: 28.6% of children with caregivers receiving SAT had a re-report. PSM sample: 29.6% of children with caregivers receiving SAT had a re-report. PSM sample: children with caregivers receiving SAT had an increase in mean CBCL externalizing score of 0.15, children of caregivers not receiving SAT decreased externalizing score by 1.8 (mean difference 1.97). |
| Female Caregivers’ Experiences with Intimate Partner Violence and Behavior Problems in Children Investigated as Victims of Maltreatment | Hazen, A. L. Connelly, C. D. Kelleher, K. J. Barth, R. P. Landsverk, J. A. | Pediatrics Vol. 117, No. 1, pp. 99-109 2006 | Caregiver Interview Caseworker Interview | Intimate partner violence; Child behavior problems; Child maltreatment; Child protective services | Female caregivers of children in core sample between 4-14 years of age (n = 2020) | Examines the relationship between caregiver’s experiences with intimate partner violence and reports of child behavior problems, accounting for other risk factors and potential moderators (depression and parenting practices). Analysis: Multiple linear regression used. | Severe intimate partner violence was associated with both externalizing and internalizing behavior problems when other risk factors were controlled (race, gender, child health, urbanicity, caregiver education, poverty, caregiver arrest history, caregiver substance use, community environment, child abuse, and prior reports of maltreatment). Use of corporal punishment and psychological aggression were significant moderators, but maternal depression did not moderate the relation between intimate partner violence and behavior problems. |
| Children in Out-of-Home Care: Entry Into Intensive or Restrictive Mental Health and Residential Care Placements | James, S. Leslie, L. K. Hurlburt, M. S. Slymen, D. J. Landsverk, J. Davis , I. Mathiesen, S. G. Zhang, J. J. | Journal of Emotional and Behavioral Disorders Vol. 14, pp. 196-208. 2006 | Child Interview Caregiver Interview Caseworker Interview Baseline Wave 2 Wave 3 Wave 4 |
Foster care; Residential treatment facilities; Emotional and behavioral problems | Children > 2 years who had any out-of-home care in waves 1-4 (n =981). | Examined timing, type, and number of different intensive/ restrictive settings children entered in their first episode of out-of-home care; the likelihood of entry into these settings, and the role of clinical and non-clinical factors in reducing or enhancing the odds of entry into the settings. Analysis: Logistic regression was performed with entry into intensive/restrictive settings as the outcome variable and sociodemographic, clinical, placement history, service use, and resource variables as predictors. | 280 of the 981 (26%) experienced entry into an intensive/ restrictive setting during their first out-of-home-care placement with a total of 635 episodes in such settings (47% had 1 episode; 31% had 2 episodes, 10% had 3 episodes, and 3% had 4 or more). The distribution was: 5% therapeutic foster care only; 31% group homes; 36% residential treatment centers; 4% inpatient psychiatric care; and 24% mixed. First placement was 42% residential treatment, 39% group homes, 10% treatment foster care, 10% inpatient psychiatric care. 48% were placed in a restrictive setting as their first placement, and 38% experienced restrictive setting as their second placement. There were 4 significant predictors of entry into intensive settings: gender (boys 2.4 times as likely), age (with each year, children were 32% more likely), behavior problems (presence of problems made it 5 times as likely), and number of placements (fewer placements OR = .31). Counting all placements, children with an additional placement were 1.7 times more likely to have entered into a restrictive/intensive setting. Differences between children who entered intensive/restrictive settings immediately v. later on revealed that they only differed on two variables: those who entered intensive setting as their first placement more likely to have experienced supervisory neglect and had 2 fewer placements during total study period. Findings indicate that children are not placed into intensive/restrictive settings as a last resort. Rather, CW may be considering children’s emotional/behavioral needs early on. Early placement may also reflect difficulty identifying available foster caregivers, particularly for older children. |
| Factors Associated with Chronic Conditions among Children in Foster Care | Jee, S. H. Barth, R. P. Szilagyi, M. A. Szilagyi, P. G. Aida, M. Davis , M. M. | Journal of Health Care of the Poor and Underserved Vol. 17, pp. 328-341 2006 | Caregiver Interview | Foster care; Chronic conditions; Children; Well-being |
Children in foster care for 1 year at baseline (n = 727) | Assesses factors associated with caregiver-identified chronic illnesses in children in foster care for 1 year. Analyses: bivariate and multivariate logistic regression models used to examine associations between prevalence of chronic disease and independent variables (demographics, insurance coverage, child remaining in foster care after one year). | Prevalence of chronic conditions was 30%--20% reported with one, 3.8% reported with two, and 3.1% reported with 3 or more. The most common condition was asthma (33%), followed by other respiratory problems (12%). Less prevalent were severe allergies, repeated ear infection, and skin diseases such as eczema. Children < 2 years and those who were in a family of no more than 3 were most likely to have a chronic condition and those who were Hispanic were least likely to have a chronic condition. Mental health problems were reported in 52% (using a binary index) and 60% (using a second index). Results from the multivariate model indicated that with demographics controlled, child’s age (< 2 years) and household composition of 3 or less remained significant. In addition, Hispanic caregivers were less likely to identify themselves as caring for a child with a chronic disease. |
| Alcohol, Drug, and Mental Health Specialty Treatment Services and Race/Ethnicity: A National Study of Children and Families Involved with Child Welfare | Libby, A. M. Orton, H. D. Barth, R. P. Webb, M. B. Burns, B. J. Wood, P. Spicer, P. | American Journal of Public Health Vol. 96, pp. 628-631 2006 | Caseworker Interview Baseline Wave 2 | Race/ethnicity; Substance treatment; Mental health treatment; Child welfare |
Caregivers who responded during both baseline and wave 2 (n = 3,340). | Compared need for and receipt of specialty alcohol drug, and mental health treatment in American Indian (AI), White, Black, and Hispanic caregivers and examined predictors of specialty service receipt. Analysis: Logistic regression used to estimate relationships between baseline characteristics and service receipt at 18-months post-baseline. | The most prevalent risk factor in AI caregivers was impaired parenting skills as it was in White and Black caregivers. 22% of all caregivers had an alcohol, drug, or mental health problem (ADM); it was similar in White and Black caregivers, but lower in Hispanic caregivers. 15% of the AI caregivers with ADM problems received a formal assessment; 25% were referred for services; and 12% received any specialty service. Logistic regression indicated that AI caregivers were less likely to receive services than Hispanic caregivers but not White or Black caregivers. Caregivers of young children received more services; caregivers with in-home children were less likely to receive ADM services. Co-occurring problems doubled a caregiver’s chance of receiving services compared with having only a substance use problem. |
| Profile of Children Investigated for Sexual Abuse: Association With Psychopathology and Services | McCrae, J. S. Chapman, M. V. Christ, S. L. . | American Journal of Orthopsychiatry Vol. 76, pp. 468-481 2006 | Child Interview Caregiver Interview Caseworker Interview Baseline | Sexual abuse; Latent profile analysis; Mental health; Child welfare services |
Children 3-14 who were investigated for sexual abuse as most serious abuse (n = 553) | Identified subgroups of children investigated for sexual abuse (SA) who exhibited or were at risk for mental health problems due to maltreatment and family experiences. Analysis: Chi square, t-tests, latent class analysis (LCA) using factor mixture modeling and regression analysis. Children were grouped by age (3-7, 8-11, 12-14 years). | Age group analysis indicated that the groups did not differ in the severity of SA but duration of abuse differed, with youngest children investigated for SA of shorter duration. Groups differed on perpetrator: 3-7 and 8-11 age groups more frequently investigated for other related abuse as compared to oldest group. Rates of PTS were higher in 8-11 age group; greater percentage of 12-14 year olds than 3-7 year olds had at least one behavior problem symptom. Factor mixture modeling analysis indicated that about 50% of children in each age group were placed in classes that were investigated for less severe abuse occurring over a short duration or only once and other family problems were only weakly endorsed. Two classes in each age group were distinguished by more severe abuse but no consistent pattern was found with family problems, other maltreatment or perpetrator; these classes accounted for one-third of children in each age group. Other classes included those that were distinguished by differences in perpetrators (older children) and by differences in mental illness (MI), domestic violence (DV), and substance use. Two classes were mostly boys (one in 3-7 and one in 12-14 ages), characterized by moderate severity, shorter duration, and caregiver mental illness; two classes of 12-14 year olds were mostly Black girls. Regression analyses with 3-7 age group indicated that children in the DV/MI and mostly boys groups had significantly more behavior problems and children with substantiated abuse had lower internalizing scores than those whose abuse was not substantiated. The chronic-relative class in the 8-11 year group was marked by the highest externalizing and PTS symptoms and internalizing and depressive symptoms were highest in the severe class. In the 12-14 age group, the mostly boys class had the highest externalizing behavior problems and depressive symptoms; they also had high internalizing and depression scores. Normative and severe chronic classes had the least symptoms overall. Out-of-home care had higher internalizing scores; substantiated cases had higher externalizing scores; White and Black children and boys had fewer depressive and PTS symptoms. Irrespective of children’s age, substantiation status was related to mental health service referral. Reports of mental health problems were also linked to service referral/receipt in 3-7 and 8-11 year groups; whereas presence of one or more borderline/clinical symptom was linked to service referral/receipt in 12-14 year olds. Limitations include missing data, use of different raters to assess psychological symptoms, and the cross sectional nature of data. Study shows that children with sexual abuse vary greatly in their outcomes and that characteristics of the abuse alone do not always predict service needs. |
| Effects of Medicaid managed care policies on mental health services use among a national probability sample of children in the child welfare system | Raghavan, R. Leibowitz, A. A. Andersen, R. M. Zima, B. T. Schuster, M. A. Landsverk, J. | Children and Youth Services Review Vol. 28, pp 1482-1496 2006 | Caregiver Interview Caseworker Interview Caring for Children in Child Welfare study 1999 Area Resource File (ARF) Baseline Wave 2 Three data sets merged: (1) NSCAW: baseline and wave 2, (2) data from the parallel Caring for Children in Child Welfare study, and (3) 1999 Area Resource File (ARF) |
Carve-out; Medicaid; Child welfare; Mental health |
3450 children aged 2 to 16 with data on ambulatory and inpatient mental health services | Analyzed the effects of three county-level Medicaid managed care variables on use of ambulatory and inpatient mental health services by children involved with child welfare services. The Medicaid variables were: 1) enrollment in Medicaid managed care, (2) use of a carve-out (which places mental health care in a separate managed care insurance contract), and (3) type of mental health provider reimbursement. Analyses included simple logistic regression for continuous variables, χ2 analyses of homogeneity for categorical variables, and multivariate logistic regression. | 16% of children in the sample obtained ambulatory mental health services and 2% inpatient care. Children who were older, had clinical levels of behavior problems, and had been placed out-of-home with college-educated caregivers had higher odds of using ambulatory mental health services. Uninsured children had only half the odds of using ambulatory care than insured children. Older age, higher levels of caregiver education, and a clinical level of externalizing behavior problems were significantly associated with use of inpatient care. African American children had significantly lower odds of obtaining inpatient care, controlling for other covariates. Managed care policy variables were not significantly associated with ambulatory mental health services. Children in counties with greater numbers of child psychiatrists had twice the odds of obtaining ambulatory services, while those with greater numbers of pediatricians had slightly lower odds of obtaining ambulatory care. Children living in counties that carved out mental health services had half the odds of inpatient mental health care. |
| Modeling behavioral problems of children in the child welfare system: Caregiver, youth, and teacher perceptions | Rosenthal, J. A. Curiel, H. F . | Children and Youth Services Review Vol. 28, pp. 1391-1408 2006 | Child Interview Caregiver Interview Teacher Interview CPS sample LTFC sample Baseline Wave 3 Wave 4 |
Mental health services; Kinship care |
4138 children aged 4 and older. NSCAW’s child welfare services sample and long-term care foster care sample were combined | Examined predictors of child behavior problems, including living situation, education level, ethnicity, and gender. A three level hierarchical linear model was used in which repeated measurements over time (Level 1) were nested within child (Level 2), which was nested within primary sampling unit (Level 3 – these were primarily county agencies). Behavior problems scores were used from three sources: caregivers, teachers, and youths themselves (age 11 and older). | In both weighted and unweighted analysis, non-kinship foster caregivers reported higher child behavior problem scores than biological caregivers, kinship caregivers and other caregivers. Teachers, on the other hand, reported behavior problems of children in kinship foster homes as higher than those of children in non-kinship foster homes. In weighted analysis, youth in adoptive homes self-reported fewer behavior problems than youth living with birth parents. Teachers reported few behavior problems on average for girls than boys. Teachers also rated increased behavior problems when caregivers were at a lower educational level, and when caregivers were African American. Girls’ self-reports of behavioral problems were significantly higher than boys’ self-reports. |
| A Longitudinal Study of the Effects of Early Abuse on Later Victimization Among High-Risk Adolescents | Tyler , K.A. Johnson, K.A. | Violence and Victims Vol. 21, pp. 287-306 2006 | Child Interview Caregiver Interview Caseworker Interview Baseline Wave 2 Wave 3 Wave 4 |
Child maltreatment; Running away; Victimization; Delinquency; Adolescents | All children in CPS cohort who were 11 to 15 years at baseline (n = 730) | Examined the effects of early abuse and poor parenting on victimization via running away, delinquency, and early sexual onset. Analysis: Path analyses using LISREL were used to examine the relationships. | Sexual and physical abuse and lower levels of parental monitoring and closeness were associated with running away at baseline. Running away at baseline was associated with running away, delinquency and early sexual activity at wave 3, which all predicted victimization at wave 4. The effect of sexual abuse and parental closeness on running away at wave 3 was mediated by running away at wave 1. Significant gender, race and age interactions were found. Males who were physically abused and males who were delinquent were at greater risk for victimization than females. Non-Whites who were delinquent were at greater risk for victimization than Whites. Older males were at greater risk for victimization than younger males; older children who had experienced sexual abuse were at greater risk than younger children who had experienced sexual abuse. Older youth who had run away had a greater risk of victimization than younger youth who had run away. |
| Developmental status and service use among children in the child welfare system - A national survey | Zimmer, M. H. Panko, L. M. . | Archives Of Pediatrics & Adolescent Medicine Vol. 160, pp. 183-188 2006 | Caseworker Interview Child Interview Caregiver Interview Baseline | Foster care; Child neglect; Home environment; Disciplinary problems; School performance; Young children; Child maltreatment; Psycho-pathology |
4324 children aged 0 to 10 years | Estimated the prevalence of developmental delay and developmental service use among children in the child welfare system, and identified factors that influence developmental delay and use of these services. Frequency percentages and results of logistic regression analyses were presented. |
24% of children were developmentally delayed on at least one measure. Children aged 0 to 2 and 3 to 5 had higher rates of developmental delay (33% and 36%, respectively) than school-aged children (13%). Only 38% of children with developmental delay were using developmental services. Children aged 0 to 2 years were less likely to receive developmental services than preschool-aged children or school-aged children. |
| Intimate Partner Violence as a Risk Factor for Children’s Use of The Emergency Room and Injuries | Casanueva, C. Foshee, V. A. Barth, R.P. | Children and Youth Services Review Vol. 27, pp. 1223-1242 2005 | Caregiver Interview Caseworker Interview Baseline | Intimate partner violence; Domestic violence; Children’s health; Emergency room; Injuries; Maternal mental health |
Core sample of children (n = 5,504) who entered the system ages 0-14 years | Examines the associations between intimate partner violence (IPV), maternal factors (e.g., depression, alcohol and substance use), lack of supervision and children’s injuries/use of the E.R. Analysis: Log-linear regression and logistic regression used to examine the association between IPV and ER/injuries maternal variables. Multivariate analysis used to examine effect of mediating variables. | Current severe IPV was associated with maternal depression and with children’s use of E.R; depression mediated the link between IPV and ER use. Maternal alcohol abuse, drug abuse, and lack of supervision did not mediate the association between IPV and children’s use of ER. Maternal alcohol abuse and drug abuse were each related to lack of supervision and mediated the relationship between IPV and lack of supervision. Maternal depression and lack of supervision were each associated with children’s injuries. Limitations include: inability to explore the direct connection between male batterers’ behaviors and children’s use of the ER, by caseworker data that may have been obtained much after investigation, and by the cross-sectional nature of the data. |
| Police Involvement in Child Protective Services Investigations: Literature Review and Secondary Data Analysis | Cross, T. P. Finkelhor, D. Ormrod, R. | Child Maltreatment Vol. 10, No. 3, pp. 224-244 2005 | Caregiver Interview Caseworker Interview Child Interview | Child abuse; Child maltreatment; Child protective services; Child protection; Police; Law enforcement; Multidisciplinary team |
Core sample limited to cases in which most serious allegations were physical abuse, sexual abuse or neglect (n = 3,842). | Compares CPS cases with and without police involvement (investigations, placement decisions, and safety plans) and examines the relationship of police and CPS co-involvement to the outcomes of maltreatment investigations. Analysis: logistic regression used to examine the relationship of police involvement (allegation credibility, provision of services), to outcome when controlling for sample characteristics. | 45% of sexual abuse cases; 28% of physical abuse cases, and 18% of neglect cases included police involvement in the investigation. Less police involvement in safety or placement planning (18% for sexual abuse). Police involvement in investigation was associated with increased likelihood that abuse allegation was found credible and that services would be provided; police involvement in planning and use of multidisciplinary team was related to provision of services. These relationships were independent of control variables (e.g., caregiver alcohol and drug abuse, severity of abuse, domestic violence), Police involvement seemed to promote CPS effectiveness. Families received more interventions with the police involvement in the case possibly because police involvement provides more thorough investigations and greater evidence. |
| Cognitive and Academic Functioning in Maltreated Children | Crozier, J.C. Barth, R.P. | Children & Schools Vol. 27, pp. 197-206 2005 | Child Assessment Caseworker Interview Caregiver Interview | Cognitive functioning; Academic achievement; Child maltreatment | School aged children (6-15 years) from core sample who completed cognitive assessments (n = 2,488) | Examines cognitive functioning and academic achievement in maltreated children. Analysis: Chi square and regression analyses used to examine differences between children’s performance as a function of age, gender, maltreatment type, race/ethnicity, and risk factors. | Children did not perform as well as their nationwide peers on tests of cognitive and academic achievement. Neither gender, age, nor maltreatment type were related to below average performance. Race/ethnicity were related to the performance on the cognitive, reading, and math measures, with African American and Hispanic students scoring lower than white children. Risk factors (poverty, prior CWS, caregiver mental health, teacher reported behavior problem) were related to one or more child measure and the cumulative risk was highly related to reading and mat scores. |
| Safety of Children Involved With Child Welfare Services | Gibbons, C.B. Chapman, M.V. Barth, R.P. Kohl, P.L. McCrae, J.S. Breen, R.L. Carlton , C. | In K. Kendall Tackett & S. Giacomoni (Eds.). Victimization of Children and Youth: Patterns of Abuse, Response Strategies; Kingston, NJ: Civic Research Institute Chapter 21, pp 21-1 - 21-26 2005 | Caseworker Interview Caregiver Interview Child Interview | Recurrent maltreatment; Safety; Child welfare services |
Core sample in baseline with closed investigations (n = 5, 504) | Examines child safety (i.e., prior/ repeated CWS involvement, children reports of violence, injuries) as a function of administrative data and caregiver reports in relation to demographic factors and placement type. Analysis: Crosstabs, logistic regression, t-tests. | Administrative data indicate that prior CWS involvement more likely in children in out of home care and in in-home cases with on-going services; older children had more prior reports; differences also a function of maltreatment type. Child reports indicated significantly higher rates of violence than in the general population of children. Severe violence reported in 20% of children who entered CWS, which was significantly less than in those placed in foster care. Furthermore, many children reported witnessing violence soon after a CWS investigation, and rates were highest in youngest children. Children in-home were more likely to report injuries requiring medical attention than those out of home and than the general population. |
| Emergency Department Utilization by Children in Foster Care | Jee, S. H. Antonucci, T. C. Aida, M. Szilagyi, M. A. Szilagyi, P. G. | Ambulatory Pediatrics Vol. 5, No. 2, pp. 102-105, March-April 2005 | Non-Permanent Caregiver Interview | Emergency department; Foster care; Health care utilization | Children in long term foster care ages 1-14 who were not reunified with their biological families and whose foster parent answered ED/ UCC (n = 559) | Examines emergency department (ED) utilization of children in foster care and which children in foster care would be most likely to use services. Analysis: Logistic regression to examine bivariate and multivariate relationships between emergency use/urgent care use and independent variables. | 31% percent of foster children had visited the ED or UCC in the past 12 months. Demographic factors associated with ED/UCC usage included: race (Hispanic less use), age (0-2 used more), chronic disease (used more), caregiver age (>54 used less). Multivariate logistic regression analysis revealed that out of all children in foster care, children who had a chronic condition, children of younger age, and children with younger foster caregivers were significantly more likely to have used the ED. |
| The Changing Lens of Research on Family Violence and Child Welfare (Introduction) | Kelleher, K. J. Barth, R. P. Edleson, J. L. | Children and Youth Services Review Vol. 27, pp. 1163-1166 2005 | None – Introduction to Special Issue | Family violence; Child welfare; Domestic violence; Child maltreatment |
None—not research paper | Issue includes 3 data-based papers using NSCAW data that focus on services and safety, especially with domestic violence in the CW caseload. Analysis: see specific papers. | “Domestic Violence and Pathways into Child Welfare Services", "Child Welfare as a Gateway to Domestic Violence Services”, “Intimate Partner as a Risk Factor for Children’s Use of the Emergency Room and Injuries”, each examine how domestic violence in the presence of maltreatment leads to different placements and services. See individual papers for results. One other paper describes a study that has brought the group of collaborators represented in this volume together (the Family Violence Services Study (FVSS) that will re-interview CWS and domestic violence agencies involved with NSCAW and will link the contextual information with individual NSCAW data. |
| Child Welfare as a Gateway to Domestic Violence Services | Kohl, P. L. Barth, R. P. Hazen, A. L. Landsverk, J. A. | Children and Youth Services Review Vol. 27, pp. 1204-1221 2005 | Caseworker Interview Caregiver Interview | Child welfare services; Domestic violence | Female permanent caregivers of children from core sample remaining in the home following an investigation for maltreatment (n = 3,135) | Examines the identification of Domestic Violence (DV) by CW workers during maltreatment investigations and how DV identification contributes to receipt of DV services. Analysis: Sensitivity and specificity for reports of DV using caregiver report as criterion. Logistic regression to 1) examine factors associated with caseworker under identification of DV (e.g., age, race, CW service status, maltreatment type, partner in home, substance abuse, prior receipt of CWS, caregiver’s maltreatment history and cumulative risk score and 2) examine predictors of service referral and receipt. | Caregivers reported that 31% of them had > 1 incident of DV in past year and 45% reported lifetime DV. Sensitivity was 8%; specificity was 65%. Underidentification occurred in 22% of the cases and were slightly more likely in the cases receiving services. Predictors of caseworker under identification included caregiver substance abuse, the family having prior experience with CWS, or the caregiver’s childhood history of abuse or neglect. The case workers identification of DV increased when the secondary caregiver had a substance abuse problem, when the most serous maltreatment type was “other” as compared to physical abuse, and when the cumulative risk was high. Recognition of DV by the case worker, coupled with an open CWS case were highly related to receipt of DV services. Lack of transportation and child care difficulties were cited as barriers to receipt of services, but not the availability of services themselves. |
| Domestic violence and pathways into child welfare services: Findings from the National Survey of child and adolescent Well-Being | Kohl, P. L. Edleson, J. L. English, D. J. Barth, R. P. | Children and Youth Services Review Vol. 27, pp. 1167-1182 2005 | Caseworker Interview | Domestic violence; Child welfare services; Out of home placement |
Children from core sample who were in home with and without CWS, foster care, kin care and caseworker information on DV (n = 3,931). | Explores the role of domestic violence in investigations of child maltreatment including relationship of DV to maltreatment classification, substantiation, decisions made by caseworker, other risk factors, and placement into out of home care. Analysis: Chi square analyses to examine bivariate associations between DV and factors. Multinomial regression to examine factors related to placement. | 14% of families had active DV and 19% had a history of DV. Foster care higher in families with active and history of DV and kin care was higher in families with active DV. Emotional maltreatment was the most serious type of abuse in DV families Substantiation of maltreatment was more common in those with active DV. DV was used in decision making process when children remained in the home with CWS or placed in kin care. History of DV was associated with highest rate of prior reports. DV was associated with substance abuse, mental health problems, arrest history, childhood history of maltreatment; financial difficulties, and cumulative risks. Logistic regression conducted with home v. out of home placement showed that DV alone did result in higher out of home placement; only a history of DV was related to placement (less likely than those without DV to be removed). Other variables related to placement were: age of child (3-5 year olds less likely to be placed out of home compared to 11+); children considered at most severe risk more likely; children of substance abusing primary caregivers more likely. Children in families with the highest level of cumulative risk were 10 times more likely to be placed into foster care than children in families assessed with low risk levels. |
| Relationship Between Entry Into Child Welfare and Mental Health Service Use | Leslie, L. K. Hurlburt, M. S. James, S. Landsverk, J. Slymen, D. J. Zhang, J. J. | Psychiatric Services Vol. 56, No. 8, pp. 981-987 2005 | Caseworker Interview Caregiver Interview | Mental health service use; Child Welfare involvement |
Children from core sample who were at least 2 years of age at baseline (n = 3, 592) | Examines the initiation of mental health service use as a function of level of child welfare involvement (i.e. in-home care and no CW services, in-home care and additional CW services and foster care) and socio-demographic and need factors. Analysis: Cox proportional hazards model used to estimate factors associated with mental health use. | Children who remained in their home without child welfare services were least likely to gain access to mental health services (1/3 less likely for those who didn’t receive any further child welfare services and 50% less likely for those who received additional child welfare services)and children in out-of-home care were most likely to gain access to such services. (onset of service use is related to level of involvement in child welfare). Other factors that predicted onset of mental health services included age, race, need, physical abuse, physical neglect, and insurance. |
| Child Welfare Reform in the United States: Findings from a Local Agency Survey | Mitchell, L. B. Barth, R. P. Green, R. Wall, A. Biemer, P. Berrick, J. D. Webb, M. B. | Child Welfare Vol. 84, pp. 5-24 2005 | Local Agency Interview | Legislation; Adoption; Safety; Child welfare reform policy implementation |
92 Child Welfare Agencies | Examines administrators’ report on the implementation of 1990’s child welfare reform legislation and its impact on service delivery. Analysis: Chi square and t-test analyses used to make comparisons of implementation of policies between administration locale, urbanicity, county size, poverty. | Respondents reported that the Adoption and Safe Families Act brought greater emphasis on child safety, shortened time frames for decision-making, and increased concurrent planning. Other legislation (TANF, MEPA-IEP) has had fewer effects. Implementation differences found as a function of contextual features. State-administered CW agencies made greater progress in implementing reforms than county-administered agencies; urban counties more likely to implement aspects of MEP-IEP but were less likely to have implemented concurrent planning. |
| Psychotropic Medication Use in a National Probability Sample of Children in the Child Welfare System | Raghavan, R. Zima, B. T. Andersen, R. M. Leibowitz, A. A. Schuster, M. A. Landsverk, J. | Journal of Child and Adolescent Psychopharmacology Vol. 15, pp. 97-106 2005 | Caseworker Interview Caregiver Interview | Child welfare; Psychotropic medication; Medical insurance; Medicaid; Foster care |
Core sample > 2 years who were seen at baseline and 1-year (n = 3,114). | Estimates the prevalence of psychotropic medication use, and describe relationships of child characteristics, mental health need, and facilitating factors (insurance and placement) with medication use among children in the CWS. Analysis: Chi square | 13.5% of children were taking psychotropic medications at the 1-year follow-up. Older age (12-16 years), male gender, whites, history of physical abuse, public insurance, and CBCL scores in the clinical range were associated with higher medication use. African-American and Latino, and a neglect history, were associated with lower medication use. Children taking psychotropic medication who were sexually abused saw specialty providers more often than non-specialty. Children with elevated CBCL internalizing scores were more likely to have specialty provider visits; children with elevated CBCL externalizing scores were more likely to have co-managed visits. |
| Overview of the children and domestic violence services study. | Smith, K. C., Kelleher, K. J., Barth, R. P., Coben, J. H., Hazen, A. L., Connelly, C. D., Rolls, J.A. | Children and Youth Services Review, Vol. 27 , pp. 1243-1258. 2005 | NA | Domestic violence; child maltreatment; child welfare |
NA | NA | Describes history of NSCAW and a linkage supplementary study, the Children and Domestic Violence Services (CADVS), for which the original NSCAW PSU were re-contacted and agencies provided information on child welfare services for children and domestic violence services for woman. |
| Developmental and Behavioral Needs and Service Use for Young Children in Child Welfare | Stahmer, A. C. Leslie, L. K. Hurlburt, M. Barth, R. P. Webb, M. B. Landsverk, J. Zhang, J. | Pediatrics Vol. 116, No. 4, pp. 891-900 2005 | Caregiver Interview Child Interview | Child welfare; Foster care; Developmental services; Developmental need; Child abuse |
Children < 6 years of age in Core Sample (n = 2,813) | Examines variation in developmental and behavioral need and service use based on race, age and level of involvement with child welfare. Analysis: Chi square analyses, correlation and logistic regression. | Children remaining at home were more likely to be older (3-5 years) and white and were less likely to have been referred because of supervisory neglect or abandonment than those removed from home. Children at home with active case were less likely to be Hispanic Children at home without active CW case were more likely to have been referred for sexual abuse. Both toddlers (41.8%) and preschoolers (68.1%) had high developmental and behavioral needs; few children received services for these issues (22.7% overall). Children with the greatest need as defined by number of areas of developmental and behavioral risk most likely to get services. Predictors of service use independent of need included: age (older than 3 more likely to receive services), level of CW involvement (children living at home less likely), maltreatment history (abandonment more likely), race (African American less likely than white). Methods need to be developed to address disparities in access to intervention. |
| Aggressive and Delinquent Behavior of Maltreated Adolescents: Risk Factors and Gender Differences | Wall, A.E. Barth, R.P. NSCAW Research Group, | Stress, Trauma, and Crisis Vol. 8, pp. 1-24 2005 | Child Interview Caseworker Interview Caregiver Interview | Aggression; Delinquency; Maltreatment; Adolescents; Gender | Adolescents (11-15 years) from the Core sample with non missing data (n =739) | Examines level of aggressive and delinquent behavior; and parent and child factors associated with aggression and delinquency. Analysis: regression to examine factors associated with aggression and delinquency. Descriptive statistics and bivariate analyses to examine gender differences. | Males reported greater caregiver relatedness and lower parental monitoring than females as well as more ADHD and less sexual abuse. Gender interactions were not included in the model because interactions between aggressive/delinquent behavior and variables were similar in males and females. Aggression and delinquency were predicted by age (11 < 15 yr olds), below average social skills, a low sense of caregiver relatedness, and being female. Aggressive and delinquent behaviors did not vary by type of maltreatment. |
| Mental Health Need and Access to Mental Health Services by Youths Involved with Child Welfare: A National Survey | Burns, B. J. Phillips, S. D. Wagner, H. R. Barth, R. P. Kolko, D. J. Campbell, Y. Landsverk, J. | Journal of the American Academy of Child and Adolescent Psychology Vol. 43, No. 8, pp. 960-970, August 2004 | Child Interview Caregiver Interview Caseworker Interview Teacher Report | Mental health services; Child welfare; Foster care |
Children from core sample whose investigations were closed and who were 2 – 14 years of age at baseline (n = 3, 803) | Assesses the relationship between the need for and use of mental health services as well as characteristics of need (as measured by the CBCL), correlates of use of mental health services, and rates and types of mental health services. Analysis: Chi square and t-tests with control for multiple comparisons; logistic regression for analyses of service use. Separate models run by age groups. | 47.9% of the youths with completed CW investigations had clinically significant emotional or behavioral problems, but only 25% received any specialty mental health care during the 12 months prior to the survey. Children in the clinical range were more likely to be adolescents and less likely to be preschoolers; more likely to be placed in nonrelative foster care or group homes; less likely to be reported for neglect; and more likely to have parents with reports of impaired parenting. Children receiving services had the same characteristics as those in the high clinical range, including higher CBCL scores, and also were more likely to have a parent identified with severe mental illness and were more often exposed to a greater number of parental risk factors. Clinical need was related to receipt of mental health care across all age groups (odds ratio = 2.7-3.5). For young children (2-5 years), sexual abuse (versus neglect) increased access to mental health services. For children 6-10 yrs, African American race and living at home significantly reduced the likelihood of services. Adolescents living at home were also less likely to receive services, whereas having a parent with severe mental illness increased (odds ratio = 2.4) the likelihood. Psychiatric hospitalization was the least common service and outpatient services were the most common service. A variety of policy were discussed. |
| Children's Voices: The Perceptions of Children in Foster Care | Chapman, M. V. Wall, A. Barth, R. P. | American Journal of Orthopsychiatry Vol. 74, pp. 293-304 2004 | Child Interview | Foster care; Children’s perceptions of placements | Children > 6 years in foster care for 1 year at baseline (n = 320). Includes children in kin care (KC), foster care (FC), and group care (GC) | Examines children’s views on their foster care experiences including their placement history, family visits, hopes for the future, the child’s relationship with current caregivers. Analysis: percentages and means for dependent variables and logistic regressions used to examine relationship between placement type and race with dimensions of the placement experience and relationship with current caregiver. | Most children reported seeing biological mother and father < 2 times per month, desiring more contact with bio mother and father; and having less contact with sibs. Most children felt happy after family visits. Most children experienced a change in neighborhood/ school. Children generally felt positively toward the providers yet maintained hope for reunification with their biological family. Children in GC were less likely to like with whom they were living and to want their current placement to become permanent. Younger children were more likely to want their current placement to become permanent. No relationships were found with race. Most children felt high levels of relatedness to their caregivers; no differences by race or placement in overall scores. However, there were placement differences in specific activities. Children in KC felt closer to their caregiver than those in GC and felt their caregiver cared for them more than children in the other 2 groups. The sample was similar to the AD Health sample. Policy implications offered. |
| Human Subjects Issues in the National Survey of Child and Adolescent Well-Being | Dowd, K.L. | In S.B. Cohen, and J.M. Lepkowski (Eds.), Proceedings of the 8th Conference on Health Research Methods. NCHS Hyattsville, MD DHHS Publication Number (PHS )04-1013 2004 | None – Background Paper | Child welfare system; Human Subjects Protections; Child Protection Services |
None- background paper | Describes the human subject protection and research ethics experiences encountered in conducting the study in order to provide lessons learned to others involved in research with vulnerable populations. | Issues covered include: Process of obtaining IRB approval; agency data for sampling; timing of data collection from caseworkers, risk/benefit ratio and procedures for gaining participant cooperation; consent/assent forms, reminders of risk throughout the interview, mandatory reporting of child maltreatment and suicidal intent; data confidentiality and release of data for secondary analysis. Results: sampling from all agencies was completed as designed; acceptable response rates were achieved in the baseline and follow-up visits; 215 incident reports filed for child maltreatment and suicidal intent; no breaches of confidentiality; data were released to research community and 25 licensing agreements were made. NSCAW team provided advice to other teams for assistance and consent forms were shared. |
| Intimate Partner Violence among Female Caregivers of Children Reported for Child Maltreatment | Hazen, A. L. Connelly, C. D. Kelleher, K. Landsverk, J. Barth, R. . | Child Abuse & Neglect Vol. 28, pp. 301-319 2004 | Caseworker Interview Caregiver Interview | Intimate partner violence; Child maltreatment; Child protective services |
Core sample not in out of home placement and on whom a baseline interview with a female caregiver was obtained in which IPV information was available (n = 3,612). | Determines the prevalence and correlates of intimate partner physical violence among female caregivers of children reported to CPS. Analysis: Chi square and F-statistics used to test associations between IPV and demographic, caregiver, and CPS variables. Logistic regression used to examine relationship of severe and less severe IPV with selected variables. Poisson regression used to examine relations between frequency of IPV with other variables. | 45% of female caregivers experienced one or more episode of IPV in their lifetime, with 33% reporting at least one incident of severe violence. Rates in previous year were 29%, with 17% reporting at least one incident of severe violence. Most who reported severe violence also reported less severe violence (94%). Caregiver depression, alcohol and drug dependence, prior reports of maltreatment were associated with increased odds for severe physical violence as was younger age. Less severe physical violence was associated with presence of a male intimate partner in home, caregiver depression, and alcohol dependence (lower risk). Similar associations were found for frequency of IPV. Rates of IPV did not vary by demographic variables. The study shows the need for effective screening and identification of intimate partner violence in families with child maltreatment experiences. |
| Contextual Predictors of Mental Health Service Use Among Children Open to Child Welfare | Hurlburt, M. S. Leslie, L. K. Landsverk, J. Barth, R. P. Burns, B. J. Gibbons, R. D. Slymen, D. J. Zhang, J. J. | Archives of General Psychiatry Vol. 61, pp. 1217-1224 2004 | Caseworker Interview Caregiver Interview Local Agency Interview | Children’s use of mental health services; Service coordination; Predictors of service use |
Children ages 2-14 from core sample years who were removed from their homes or were living at home if their case was opened (n = 2,823). | Explores contextual predictors of specialty mental health service use among children in the child welfare system by examining degree of coordination between local child welfare and mental health agencies and supply of specialty mental health providers. Analysis: multivariate logistic regression analyses. | Only 28.3% of all children received services although 42.4% demonstrated need. Younger children and children remaining at home had lowest rates of service use. Increased coordination between local child welfare and mental health agencies was associated with stronger relationships; mental health need and service use and decreased differences in rates of service use between white and African American children. There was no association between the supply of mental health providers and service use. |
| Outpatient Mental Health Services for Children in Foster Care: a National Perspective | Leslie, L. K. Hurlburt, M. S. Landsverk, J. Barth, R. Slymen, D. J. | Child Abuse & Neglect Vol. 28, pp. 697-712 2004 | Child Interview Caseworker Interview Caregiver Interview | Mental health; Child abuse; Foster care; Medicaid | Children in foster care for one year who were 2-15 years old at baseline (n = 462) | The objective was to determine factors influencing the use of outpatient mental health services for children in foster care. Analysis: Chi square, correlations, logistic regression to examine relationship between use of mental health services and need for services (as measured by CBCL) and background factors. | Almost half the children had a high rate of mental health need, and over half of the received at least one outpatient mental health service. About 25% of those with high CBCL scores had not accessed mental heath services. Need, older age, and history of sexual abuse history all positively predicted the use of these services. African American children and physically neglected children were less likely to access services and children in group care were more likely to access services. A significant interaction between CBCL scores and race (AA) indicated that AA children in comparison to white children were less likely to access services when CBCL scores were lower. |
| Parental Arrest and Children Involved With Child Welfare Services Agencies | Phillips, S. D. Burns, B. J. Wagner, H. R. Barth, R. P. | American Journal of Orthopsychiatry Vol. 74, No. 2, pp. 174-184 2004 | Caseworker Interview Caregiver Interview | Child welfare services; Parental arrest; Mental health care; Out-of-home care; Domestic violence; Substance abuse |
Core Sample on whom there were parental arrest data (n = 5,322) | Examines rates of parental arrest among children who were the subjects of maltreatment reports. Examine differences between children whose parents were arrested and others with regard to demographics, parent risk factors, CWS case characteristics, and race/ ethnicity Analysis: Chi square to compare cases with arrested parents and others. Logistic regression to estimate effect of parental arrest on out of home placement controlling for child and parent factors. | 12.5% children who were subjects of maltreatment reports had parents who were recently arrested, and they were disproportionately African American, younger, more likely to have had previous contact with CWS, but less likely to have emotional/ behavioral problems. They were more likely to have had reports for some form of neglect. Their parents generally had higher risk factors. Children whose parents were arrested were significantly more likely to be in out-of-home care, usually relative care. Parental arrest contributed to out of home placement even after accounting for the effect of parent risk factors and child behavior problems. However, odds of being in non-relative care were increased by multiple parent risk factors, maltreatment, and child behavior problems but parent arrest did not make an independent contribution. Although 20% of the African American children had an arrested parent, African American parents who were arrested had lower rates of each parent risk factor than other arrested parents. Policy implications indicated. |
| Parental Views of In-Home Services: What Predicts Satisfaction with Child Welfare Workers? | Chapman, M. V. Gibbons, C. B. Barth, R. P. McCrae, J. S. | Child Welfare Vol. 82, No. 5, pp. 571-596 2003 | Caregiver Interview Caseworker Interview | Caregivers receiving child welfare services; Satisfaction with child welfare workers | Caregivers of children from core sample remaining in the home who received CWS and talked with a CW worker (n = 1,872) | Examines caregivers’ satisfaction with child welfare workers and caregiver and worker characteristics that predict satisfaction. Analysis: descriptive statistics and multiple regression analyses. | Most caregivers had spoken with a caseworker in the past month (66%); only 13% had spoken to a worker > 4 months ago. The mean score for relationship quality indicates a moderate degree of helpfulness. Service characteristics negatively impacting the relationship include >2 child welfare workers, longer length of time between visits, not being offered the kind of help needed, too few services being offered, and caseworker report of caregiver cooperation. |
| Comprehensive Assessments for Children Entering Foster Care: A National Perspective | Leslie, L. K. Hurlburt, M. S. Landsverk, J. Rolls, J. A. Wood, P. A. Kelleher, K. J. | Pediatrics Vol. 112, No. 1, pp. 134-142 2003 | Caring for Children in Child Welfare Entry Screening and Services Interview of CW Agency Personnel Abstracted policy and procedure information | Foster care; screening; Health care; Mental health; Developmental delay; Child health |
91 of 92 NSCAW PSU’s | Investigates the status of policies regarding assessments of children entering out-of-home care, as to the inclusiveness and comprehensiveness of physical, mental, and developmental assessments. Examine the relationship between inclusiveness and the estimated percentage of children assessed, primary assessment location, and principal assessment provider type. Analysis: Chi Square and correlations. | 86% of PSUs had a policy covering all children for physical health assessments; 48% of PSUs had policies covering all for mental health and 58% for developmental assessments. Only 42.6% of PSUs provided comprehensive physical, mental health, and developmental examinations inclusive of all children entering out-of-home care. Estimated proportion receiving assessments was higher in PSU’s with inclusive policies. Health assessments usually occurred at community locations and were nearly universally given by primary care providers. Local mental health agencies were primary locations for mental health assessments with a substantial portion at central child welfare location (the latter more often were for PSU’s with inclusive policies). Health facilities were usually used for developmental assessments (most often for those with inclusive policies). Primary care providers were the usual source of mental health and developmental assessments in PSU’s with inclusive policies, whereas specialists were used for other types of PSU’s. Only 25% required specific tools for identification. |
| Foster care and Medicaid managed care. | Leslie, L. K., Kelleher, K. J., Burns, B. J., Landsverk, J., & Rolls, J. A.. | Child Welfare Vol. 82, pp. 367-392. 2003 | NA | Out of home are; health care; mental health; vulnerable children; substance abuse; California; access; needs; experience; services |
NA | NA | Describes NSCAW and CCCW potential use to study foster care and Medicaid managed care. |
| Children in Foster Homes: How Are They Faring? | Vandivere, A. Chalk, R. Moore, K. A. | Child Trends Research Brief Publication No. 2003-23; www.childtrends.org 2003 | Child Assessment Child Interview Caregiver Interview | Foster care; Child well-being; Policy implications |
Children living in foster homes during baseline (n = 1,279); 669 children in foster care and 60,000 not in FC in the NSAF | Examines the child well-being of children in foster care and addresses the implications for policy and practice. Analysis: descriptive statistics (i.e., percentages) for various indicators of health, development, education, and behavioral problems. | Children in foster care more likely to have problems than non-foster children but also have strengths. Foster care may improve child outcomes if the placement facilitates greater access to services. Two suggestions made for obtaining data about children in foster care: the addition of measures of child well-being to state and federal child welfare administrative data systems and the inclusion of foster children in nationally representative surveys of children and youth. Other suggestions were made for strategies that could help foster families support foster children’s development. |
| Methodological Lessons from the National Survey of Child and Adolescent Well-being: The First Three Years of the USA’s First National Probability Study of Children and Families Investigated for Abuse and Neglect | NSCAW Research Group | Children and Youth Services Review Vol. 24, No. 67, pp. 513-541 2002 | None – Background Paper | Child welfare system; Child maltreatment; Longitudinal survey design; Methodological lessons |
None | Focuses on the process of implementing the study using information from the planning stages and early months of field work, including sampling, instrumentation, agency and family recruitment, data collection, human subjects’ protection, and analysis plans. | Among the lessons learned during early phase of study were: Agency recruitment - include CWW supervisors at recruitment meetings to facilitate relationships; have senior data staff at recruitment meetings to work with sampling team; ask key senior staff to sign an agreement letter to lessen agency problems due to staff turnover; develop multifunction teams for working with CW agencies. Recruitment of families – avoid including sample members who require active consent prior to initial contact but if required provide additional resources to the agencies and if required for policy reasons ask that agencies review the policy; identify agency consent requirements early on; justify request for direct contacting of participants. Data collection procedures - use ACASI for some questions for children and to use CARI as a means for monitoring quality of interviews. After the first 5 months, response rates ranged from 40% to 85% for stratum. Incentives were raised to try to increase rates. Data collection more time consuming resulting in decision to extend sampling and baseline data collection by 3 months. Costs greater than anticipated resulting in decision to eliminate some data collection time points and extending time between surveys. |
| Design and Methodological Issues in a National Longitudinal Survey of Children in the Child Welfare System | Dowd, K. L. Biemer, P.P. Weeks, M. F. | In M.L. Cynamon, and R.A. Kulka (Eds.), Proceedings of the Seventh Conference on Health Research Methods, Williamsburg, VA Vol. 24, No. 67, pp. 513-541 1999 | None – Background Paper | Child welfare system; Longitudinal survey design and methodology; Child Protection Services | None—description of how the sample will be obtained. | Describes the NSCAW and the issues that were faced in designing the survey, Analysis: none—not a data paper. | Results – none. Issues covered include: Instrumentation – need to have them answer key questions and be programmed for computer; 2 stage sampling – PSUs and within PSU sampling – needed to ensure estimates for 8 key states and the rest of the U.S (28 other states); Agency recruitment required significant preparation follow-up, time, and networking. Response Rates – incorporated best practices for surveys; Human Subjects Protection and OMB Reviews – Human Subjects Work Group who helped with design of study recommended balance between protection of study subjects and getting a sufficiently high response rate; Analysis Plans – focus on key study issues concerning experiences of children and families along with information about agencies and caseworkers. |
| Return to Previous page |