Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: Mental health treatment often strives to optimize treatment success by assisting women to repair key relationships and build supportive network ties. Conversely, substance abuse treatment, particularly residential treatment, often seeks to cut clients' ties to past associates and encourages them to build new, non-drug using networks. This study examines changes in the social support networks of women with co-occurring disorders and assesses the relationship of social support and treatment outcome for a year post residential treatment entry. It is hypothesized that larger social networks and greater perceived material and emotional support at baseline and subsequent time points will be associated with positive outcome.

Method: The data was collected as part of the SAMHSA Women Co-Occurring Disorders and Violence Study, a cross-site study to test the impact of integrated trauma informed approaches on women with the triple troubles of psychiatric disorders, substance abuse and a history of trauma. This study includes data from a sample of 173 women in NYC. A modified version of the Social Network Social Support Interview (El Bassel et al, 1998) was administered at baseline, and at 3, 6, 9 and 12 months post baseline. Outcome measures including psychiatric symptom distress, trauma symptom distress and substance use severity were gathered at baseline, 6 and 12 months. Bivariate analysis was used to assess changes over time in network properties as well as the impact of social support.

Results: Support networks were initially small and did change during the year, with significant growth in network size from baseline (mean=1.9657) to three months (mean=2.400 t=-2.904, p=.004) and subsequent shrinking from 3 months to 6 months (t=3.640, p=.001). Treatment outcomes are not related to network size or perceived support at baseline or at subsequent time points as assessed via correlation coefficients (raw symptom scores). Further analysis examining patterns of symptoms worsening, remaining the same, and improving for all four measures showed worsening of drug use from baseline to 12 months associated with larger network size (Χ=28.483, df=12, p=.005).

Implications: Though practice wisdom lauds social support as an important ingredient for maintaining well-being and stability, the data presented here suggests social support is not a significant positive contributor to treatment success in residential settings for trauma survivors with dual diagnosis. Family and friends may provide support along with memories of past shared victimization and/or they may be perpetrators of the client's abuse. Practitioners should carefully assess a client's social support network at key intervals during her treatment career to assess the evolution of the support network and to probe more deeply on both the supports and strains the social support network creates for the client. The data suggests we need to be far more cautious in our expectations for network interventions with these vulnerable women.