Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Rationale: Usual treatment for opioid dependent adolescents/young adults is detoxification and counseling in outpatient or inpatient settings, and relapse rates are high. Extended use of buprenorphine may improve treatment outcome, but its cost-effectiveness is unknown. Objective: To estimate cost, quality-adjusted life years (QALYs), and cost per QALY in a clinical trial of buprenorphine versus standard treatment in 154 opioid dependent youth Methods: Participants aged 15-21 seeking outpatient treatment were randomized to 12-weeks of buprenorphine/naloxone (BUP) or a 14 day detoxification (DETOX), each with twice weekly drug counseling. Data were collected prospectively during the 3-month treatment and for 9 months afterward. Economic data collection included the cost of BUP administration and therapy; use of services related to adverse events; use of other non-study medical services; and quality of life assessment (EuroQOL). Prices were assembled from a DATCAP assessment of the clinical sites for treatment costs, lists of average wholesale price for prescription drugs, and insurance claims for medical services. We used OLS and GLM to evaluate cost and QALYs. We used inverse probability weights to address missing data. Confidence intervals for the cost-effectiveness ratio were calculated by use of Fieller's theorem method. Analyses were performed for the 3 months of active treatment as well as for the 1-year of total follow-up. Results: 3-month costs equaled $980 (SE=419) for DETOX and $2519 (503) for BUP with a significant $1538 incremental BUP cost (p = 0.02). 3-month QALYs were 0.163 (0.007) and 0.182 (0.005) respectively, with a significant 0.020 difference (p = 0.01). The $80,600 cost/QALY ratio had a confidence interval that ranged from $12,800 to $407,400. Continued follow-up for a full year diminished the strength of the signal for higher cost and QALYs (incremental BUP cost, $627 [2049], p=0.76; incremental QALYs, 0.055 [0.033], p=0.10). While the 1-year cost/QALY ratio shrank to $12,200, the 95% CI was undefined. Similar results were observed when we used OLS and GLM models to predict cost. Conclusions: While the point estimate from this 1 year trial indicates good value, variability around the estimates suggests that we cannot be confident of value. The initial strong signal of QALY gains dimmed during the 9 months in which active intervention was withdrawn. It will be important to defer conclusions about the social value of this intervention until we incorporate changes in criminal behavior and education into our analysis.