Brief interventions to address substance use among patients presenting to emergency departments in resource poor settings: A cost-effectiveness analysis

Background: There are limited data describing the cost-effectiveness of brief interventions for substance use in resource-poor settings. Using a patient and provider perspective, this study investigates the cost-effectiveness of a brief motivational interviewing (MI) intervention versus a combined intervention of MI and problem solving therapy (MI-PST) for reducing substance use among patients presenting to emergency departments, in comparison to a control group. Methods: Effectiveness data were extracted from Project STRIVE (Substance use and Trauma InterVention) conducted in South Africa. Patients were randomised to either receive 1 session of MI (n = 113) or MI in addition to four sessions of PST (n = 109) or no intervention [control (n = 110)]. Costs included the direct health care costs associated with the interventions. Patient costs included out of pocket payments incurred accessing the MI-PST intervention. Outcome measures were patients' scores on the Alcohol, Smoking and Substance Use Involvement Screening Test (ASSIST) and the Centre for Epidemiological Studies Depression Scale (CES-D). Results: Cost per patient was low in all three groups; US$16, US$33 and US$11, and for MI, MI-PST and control respectively. Outcomes were 0.92 (MI), 1.06 (MI-PST) and 0.88 (control) for ASSIST scores; and 0.74 (MI), 1.27 (MI-PST) and 0.53 (control) for CES-D scores. In comparison to the control group, the MI intervention costs an additional US$119 per unit reduction in ASSIST score, (US$20 for CES-D); MI-PST in comparison to MI costs US$131 or US$33 per unit reduction in ASSIST or CES-D scores respectively. The sensitivity analyses showed that increasing the number of patients who screened positive and thus received the intervention could improve the effectiveness and cost-effectiveness of the interventions. Conclusion: MI or MI-PST interventions delivered by lay counsellors have the potential to be cost-effective strategies for the reduction of substance use disorder and depressive symptoms among patients presenting at emergency departments in resource poor settings. Given the high economic, social and health care cost of substance use disorders in South Africa, these results suggest that these interventions should be carefully considered for future implementation.