Evaluating Psychiatric Hospital Admission Decisions for Children in Foster Care: An Optimal Classification Tree Analysis
ABSTRACT This study explored clinical and nonclinical predictors of inpatient hospital admission decisions across a sample of children in foster care over 4 years (N = 13,245). Forty-eight percent of participants were female and the mean age was 13.4 (SD = 3.5 years). Optimal data analysis (Yarnold & Soltysik, 2005) was used to construct a nonlinear classification tree model for predicting admission decisions. As expected, clinical variables such as suicidality, psychoticism, and dangerousness predicted psychiatric admissions; however, several variables that are not direct indications of acute psychiatric distress, such as the presence of family problems and the location of the hospital screening, impacted decision making in a subsample of cases. Further analyses indicated that the model developed in Year 1 reliably and consistently predicted admission decisions (with 64%-68% overall accuracy) across the next 3 years. Policy, research, and clinical implications are discussed.
- SourceAvailable from: ecommons.luc.edu
[Show abstract] [Hide abstract]
- "Past research has examined the relationship between different demographic variables and child welfare placement decisions, including age (Knapp et al., 1987, Britner & Mossler, 2002; Danglish & Drew, 1989, Courtney, 1988; McMurty & Lie, 1992, Brooks, James, & Barth, 2002), gender (Budde et al., 2004; Britner & Mossler, 2002, Glisson, Bailey, & Post, 2000), and geographic location in which the child resides (Budde et al., 2004). has also assessed the influence of clinical variables on placement decisions (Courtney, 1998; Snowden et al., 2007; Child Welfare League of America, 2005, Glisson & Green, 2006). In general, the previous research studies have offered support, to varying degrees, that decision makers take into account a multitude of factors when making placement decisions. "
ABSTRACT: Building on prior research that utilizes clinical vignettes to analyze child welfare placement decisions (Briar, 1963; Britner & Mossler, 2002; Donnelly, 1980; Drury-Hudson, 1999; Gold, et al., 2001; Taylor, 2006), the present study employs an experimental format to assess the relative importance that social workers place on variables related to placement decisions, and to study any differences regarding decision making that may emerge amongst the child welfare professionals themselves. The study consists of a single hypothetical vignette that will be sent to social workers in the state of Illinois. Prior to administration of the questionnaire, multiple experts in the field of child welfare were consulted in order to ensure that the vignette is a realistic depiction of a placement scenario. The study was sent via postal mail to social workers who are both experts and novices within the field of child welfare. The respondents indicated if their preferred placement option would be community-based or residential and completed a portion of a standardized assessment tool based on their impressions of the child in the vignette. The respondents also answered basic demographic questions about themselves, such as the type of clients that they typically work with and the number of years they have been at their job. Finally, the respondents answered questions pertaining to specific experience in decision making in the child welfare domain. The entire study was designed to be completed by the respondents in about five minutes.
- [Show abstract] [Hide abstract]
ABSTRACT: The Child and Adolescent Needs and Strengths (CANS) is a functional assessment used in approximately 27 states to evaluate youth service outcomes. The CANS purports to measure both the youth's risk and protective factors, but its validity is largely un-researched. This study compares ratings of 304 delinquent youth on the CANS and ratings on a functional assessment whose validity is well established: The Child and Adolescent Functional Assessment Scale (CAFAS). Participants were selected from an Illinois juvenile justice program and had a mean age of 14 years, 7 months (SD = 1.30 years). Females constituted 33.3% of the sample and males 67.7%. It was hypothesized that correlations obtained between CANS items and CAFAS subscales purporting to measure the same general aspect of functioning would be positive, statistically significant, and of moderate to high magnitude. Results suggest the CANS is a valid measure of outcomes within this population. The implications of these results are discussed and recommendations for future research are made. (Contains 2 tables.) [This paper was presented as a poster at the American Psychological Association 2003 Convention in Toronto, Ontario.]