Substance abuse treatment and psychiatric comorbidity: do benefits spill over? Analysis of data from a prospective trial among cocaine-dependent homeless persons

Abstract

Background: Comorbid psychiatric illness can undermine outcomes among homeless persons undergoing addiction treatment, and psychiatric specialty care is not always readily available. The prognosis for nonsubstance abuse psychiatric diagnoses among homeless persons receiving behaviorally-based addiction treatment, however, is little studied.

Results: Data from an addiction treatment trial for 95 cocaine-dependent homeless persons (1996-1998) were used to profile psychiatric diagnoses at baseline and 6 months, including mood-related disorders (e.g. depression) and anxiety-related disorders (e.g. post-traumatic stress disorder). Treatment interventions, including systematic reinforcement for goal attainment, were behavioral in orientation. There was a 32% reduction in the prevalence of comorbid non-addiction psychiatric disorder from baseline to 6 months, with similar reductions in the prevalence of mood (-32%) and anxiety-related disorders (-20%) (p = 0.12).

Conclusion: Among cocaine-dependent homeless persons with psychiatric comorbidity undergoing behavioral addiction treatment, a reduction in comorbid psychiatric disorder prevalence was observed over 6 months. Not all participants improved, suggesting that even evidence-based addiction treatment will prove insufficient for a meaningful proportion of the dually diagnosed homeless population.

Figure 1

Percentage of Participants with Mood-Related Disorders (MRDs) and Anxiety-Related Disorders (ARDs) at Baseline and at 6-month Follow-up. Ninety-five homeless, cocaine-dependent treatment trial participants were diagnostically assessed at baseline and at 6 months' follow-up for the presence of Mood-Related Disorders (MRDs) or Anxiety-Related Disorders (ARDs), using a clinical interview conducted by a trained psychologist. MRDs included disorders such as depression and bipolar disorder. ARDs included disorders such as post-traumatic stress disorder and generalized anxiety disorder (for details, see Methods). The figure depicts the prevalence of each disorder type (MRDs and ARDs) among this sample at baseline and 6 months. The 32% absolute decline in prevalence of non-addiction psychiatric disorders from baseline to 6 months corresponds to -26% change in a statistical model adjusting for disorder type, and accounting for multiple observations per person. (χ² [df 1] = 30.49, p < 0.001). A test of whether the decline in MRD prevalence from 0 to 6 months differed from the decline for ARDs was nonsignificant, (χ² [df 1] = 2.48, p = 0.12).