Impact of Residential Versus Outpatient Substance Abuse Treatment on Child Welfare Outcomes: A Secondary Analysis of Nscaw II Data

Background/Purpose:

Between 40–80% of substantiated child maltreatment cases in the United States involve some degree of parental substance use.  Outcomes for children of parents with substance abuse issues are worse than for children of parents without those issues.  Treatment options for parents range from informal and low-cost outpatient services to expensive residential programs.  While some studies have demonstrated that substance abuse treatment can improve child welfare outcomes, results are largely inconclusive as to whether residential treatment is more effective than outpatient options. Using data from the National Survey of Child and Adolescent Well-Being (NSACW) II, this study evaluates the relationship between residential and outpatient treatment on caseworker assessments of case plan progress for parents involved in the child welfare system. 

Methods:

This study utilizes data from Waves I and II of the NSCAW II, a nationally representative sample of 5,872 children, ages 0 to 17.5, referred to child welfare agencies for maltreatment between February 2008 and April 2009.  The final sample was restricted to 203 cases where parents completed Wave I assessments of substance abuse severity and participated in either residential or outpatient treatment during Wave II. 

Using Wave I data, propensity scores were generated predicting the likelihood that a parent would participate in residential instead of outpatient treatment given a number of covariates including alcohol and drug abuse severity.  Propensity scores were multiplied with Wave II survey weights for each case to control for selection bias into treatment while also accounting for the NSCAW II sampling design.  Weighted logistic regression models were used to determine whether participation in residential versus outpatient treatment was related to caseworker assessments of case plan progress, controlling for a number of additional covariates. 

Results:

Results indicate that overall, residential substance abuse treatment is negatively correlated with caseworker assessments of case plan progress (OR = 0.0015, p = 0.002). Other significant relationships included child gender (female OR = 8.42; p = 0.013), caregiver education (OR = 1.69; p = 0.007), and worker assessed parental difficulties paying for basic necessities (OR 0.15; p = 0.05).  For parents with self-reported drug problems, a significant relationship was also found which indicates that parents with higher severity problems were more likely to be assessed as demonstrating case plan progress if they attended residential treatment (OR = 43.23, p = 0.008). 

Conclusions and Implications:

Residential treatment is significantly more expensive than outpatient options for treating substance abuse in terms of actual costs as well as the often-necessary leaves of absence that parents must take from employment, childcare, and other responsibilities in order to participate. Considering these costs and the finding that participation in residential treatment is not associated with a higher likelihood of caseworker assessments of case plan progress, case-planning efforts with parents should not presume that residential treatment is more beneficial, except in cases where parents report higher levels of drug abuse problems.  To ensure the best possible fit between parental substance abuse and treatment, child welfare assessments should include standardized measures of drug and alcohol abuse severity.