Women with co-occurring disorders and histories of abuse: moderators of treatment effect on services use and costs Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 22, 2019
  • Gilbert, Allison R.
    • Affiliation: Gillings School of Global Public Health, Department of Health Policy and Management
  • Objective: Women with co-occurring mental health and substance abuse disorders and histories of abuse are heterogeneous in symptom severity and use of support services, which may affect their treatment response. This dissertation estimated differential effects of an integrated counseling intervention (IC) across sub-groups of women in this population on their outcome services use and costs. Data Sources/Study Setting: Data from a national study conducted from 1998-2003. 2,729 eligible women were recruited into IC or usual care study groups at nine study sites. Study Design: Interviews were conducted with participants at baseline, three, six, nine and 12 months regarding their service use. Data Collection/Extraction Methods: Cluster analyses identified sub-groups of participants according to symptom profiles at baseline and separately according to service use profiles at baseline. Regression analyses estimated the effect of IC, by sub-group, on participants' outcome use of outpatient counseling, residential substance abuse treatment, medical and overall costs. Latent class analyses were also conducted as a comparative approach to modeling sub-group effects of IC. Principal Findings: Among women with moderate-to-severe PTSD at baseline, the IC group had fewer counseling visits and a lower probability of having any medical costs at follow-up than the usual care group. Among women who used counseling intensively at baseline, the IC group had fewer days of residential treatment at 12 months than the usual care group. Among women with high drug addiction and PTSD at baseline and women with high alcohol severity, those in IC had lower medical costs than women in usual care. Latent class models produced generally consistent effects for residential treatment and medical costs. Conclusions: IC was relatively inefficient for women whose predominant symptom at baseline was moderate-to-severe PTSD. Efforts should be made to improve treatment and outcomes for these women. IC worked well for women who at baseline used counseling intensively, had high drug addiction and PTSD, or had high alcohol severity. Practitioners can identify these women among their patients and direct them toward IC if they are not already engaged. Optimizing the effect of IC for women in this population can improve patient outcomes and conserve public resources.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Domino, Marisa
  • Open access

This work has no parents.