Predictors of Peer Social Support in a Dyad-Based Peer Intervention for Substance Misuse

Background and Purpose: The Peer-Enhanced Community Reinforcement Approach (P-CRA; Smith et al., 2016) is a novel intervention for emerging adults suffering from substance use disorders.  It integrates the peer of an emerging adult target client in a helping role where they receive up to four sessions of counseling on how to increase their alcohol and drug-specific social support of the target client (i.e., peer supportive behaviors (PSB)).  Target clients received 12 weeks of P-CRA.  Selecting appropriate peers for this intervention is a matter of debate, as target clients may only have substance using peers in their existing social networks. Furthermore, we know little about whether the intervention actually impacted peer behaviors as a result of the intervention.  Therefore, the purpose of this study was to explore whether changes in PSB and peer substance use were related to intervention exposure and peer baseline substance use. 

Methods: We analyzed data from a pre-experimental study of P-CRA (n=69, 35 peer dyads) conducted in a community not-for-profit substance use disorder treatment agency. Data were collected on both peers and target clients (39.1% female, 53.6% African American, Mean Age =21.8) at baseline (i.e., pre-treatment), three months (i.e. post-treatment), and six months. A 9-item scale (α=.74) measured PSB. Substance use was measured by the 8-item Substance Frequency Scale (SFS), a reliable and valid measure (α=.79).  Finally, peer intervention exposure in number of minutes was measured from administrative records and coding. Peer supportive behaviors at the six-month follow-up were regressed on intervention exposure, peer substance use frequency, and peer supportive behaviors at baseline. Additional analyses regressed peer substance use frequency at follow-up on intervention exposure and peer substance use frequency at baseline.

Results: Peer substance use frequency at baseline was not a significant predictor of PSB at three (b=-.002, t(15)=-.876, p=.394) or six months (b=.002, t(15)=.981, p=.339). In addition, intervention exposure did not significantly predict PSB at three (b=.000, t(15)=.365, p=.720) or six months (b=.000, t(15)=-.554, p=.586) either. However, peer intervention exposure predicted three month peer substance use frequency (b=-.119, t(18)=-2.664, p<.02) even after controlling for peer substance use frequency at baseline. 

Conclusions and Implications:  Although this study was very small, initial indications are that P-CRA did not have a large impact on peer supportive behaviors.  However, peer session attendance at P-CRA influenced peer substance use at follow up.  Potential benefits of the P-CRA model to target clients may operate through peer reductions in substance use and not increases in peer supportive behaviors.  Larger randomized trials are needed to test hypotheses on mechanisms of change operating in P-CRA, as prior research found both peers and target clients made significant reductions in their substance use over time.