Identification of patients at high risk of noncompliance in substance dependence treatment

Noncompliance in substance dependence treatment seriously threatens its effectiveness. Pretreatment identification of those at the highest risk allows targeting of specific compliance enhancing interventions to those who may benefit most from it. In a cohort of 292 patients entering a 30-day treatment program for substance abuse treatment, several potential predictors of noncompliance were recorded before treatment. Compliance was registered when treatment was discontinued before the 30th day. Logistic regression was used to identify independent predictors of noncompliance and the area under the receiver-operating characteristic (ROC) curve was quantified to assess their joint predictive value. Mean age was 35 years, 80% were male. During follow-up 31% became noncompliant. Independent predictors of noncompliance were male gender, less than 10 years of education, severity of medical and drug problems, and behavioral intention to comply with treatment. The area under the ROC curve was 0.70 (95% CI: 0.64-0.76). In conclusion, noncompliance in a substance abuse treatment program can be predicted using five easy to assess patient characteristics. These results may be useful in preventing noncompliant behavior early in treatment by motivational counseling strategies.