Interaction between pre-treatment drug use and heterogeneity of psychiatric diagnosis predicts outcomes in outpatients with co-occurring disorders.

We examined whether the interaction of baseline stimulant use, assessed by urine drug tests, and type of serious mental illness (SMI) diagnosis predicted stimulant use in a trial of contingency management (CM). The interaction between baseline stimulant use and SMI diagnoses was significant in the overall sample (p=0.002) when controlling for the main effects of treatment condition, baseline stimulant use, and SMI diagnosis. Similar results were also found within the CM sample. Individuals with bipolar disorder were more or less likely, depending on their baseline stimulant-drug test results, to use stimulants during treatment compared to those with other SMI diagnoses.