The Severity of Substance Abuse and the Effectiveness of Recovery Coaches in Child Welfare

Background and Purpose: A primary goal for child welfare is to improve family functioning and achieve family reunification. Yet this goal has proven particularly difficult for children with substance abusing caregivers - a group that experiences extraordinary low rates of reunification (U.S. General Accounting Office, 2003). To facilitate recovery and increase the likelihood of reunification, child welfare systems are experimenting with intensive case management models. In Illinois, recovery coaches are employed to work closely with substance abusing parents and treatment providers. Overall, the recovery coach model has achieved modest success. The purpose of the current study is to investigate how this particular model of intensive case management interacts with the severity of substance abuse. Similar to findings reported in other clinical trials (Fournier et al, 2010), we hypothesize that the intervention effects will be greater for caregivers that report higher levels and greater frequency of drug and alcohol consumption.

Methods: The design for the current study is experimental. The sample includes 462 female caregivers assigned to the experimental group (recovery coach) and 179 female caregivers assigned to the control group. We analyze a variety of survey and administrative data, including initial screening, caseworker assessments of treatment progress, official reports of maltreatment, and official reports of reunification. We calculate a composite substance abuse severity score for caregivers based on amount (3 levels) and frequency (6 levels) of consumption. We compute the composite score similar to the weighting procedure of the Addition Severity Index. The formula is as follows: composite score=1/2 (amount rating/3+frequency rating/6). Data analyses include descriptive statistics, analysis of three way contingency tables, and logistic regression.

Results: The descriptive analyses indicate that 15.4% of caregivers are low consumption, 42.4% medium consumption, and 42.1% high consumption. We use three way contingency tables to investigate the interactions between random group assignment and substance abuse severity. Caregivers identified their primary substance of choice. We ran separate analyses based on the following primary substance groups: alcohol (12.8%), marijuana (10.5%), cocaine (49.0%) and heroin (27.8%). We focus specifically on the likelihood of achieving family reunification. A significant interaction does emerge, but only for the caregivers using heroin. Specifically, the recovery coach model is most effective (with regard to reunification) for the female caregivers identified as high consumption (27% of the experimental group achieved reunification as compared with only 10% of the control group). This effect remains in logistic regression models even after controlling for other important covariates (Exp(B) = 7.40, p<.05).

Conclusions and Implications: Demonstration programs targeting substance abusing families are reported to effectively improve reunification rates. Yet, current economic constraints are forcing systems to better target clients. The findings from the current study indicate that the severity of substance abuse is an important factor for differentiating caregivers in child welfare. The benefits of intensive case management models (such as recovery coaches) appear to be non-existent for low and even moderate substance abuse. Targeting the most severe substance abusers in child welfare is likely the best strategy from both an effectiveness and efficiency perspective.