RESUMO
BACKGROUND: Utilization of extended release naltrexone (XRN) for alcohol use disorders (AUDs) in the U.S. Veterans Health Administration (VHA) has been limited, perhaps due to high cost, lack of established superiority over less expensive alternatives including oral naltrexone, and related formulary restrictions. Despite these barriers, pockets of higher utilization exist in VHA, allowing for the quasi-experimental examination of the effects of XRN on 1-year mortality and number of subsequent detoxification episodes among patients with high rates of psychiatric comorbidities and previous psychosocial and pharmacological addiction treatment. METHODS: Using propensity score-weighted mixed-effects logistic regression, 1-year mortality was compared between patients with AUDs who received XRN in fiscal year 2010 (n = 387) and a random sample of patients with AUDs who did not receive XRN (n = 3,759). Among the subgroup of patients who had at least 1 detoxification episode in the previous year, 1-year mortality and number of subsequent detoxification episodes were compared between those who did and did not receive XRN. RESULTS: Overall, 1-year mortality for the patients receiving XRN was significantly lower than for the comparison group who did not receive XRN (odds ratio [OR] = 0.30; p < 0.001). Among patients with a detoxification episode in the previous year, those receiving XRN had, on average, 0.80 fewer subsequent detoxification episodes (p < 0.001) and significantly lower mortality (OR = 0.78, p < 0.001) in the postindex year. CONCLUSIONS: Among patients with AUDs, those receiving XRN had lower 1-year mortality and fewer detoxifications compared to similar patients not receiving XRN. These results, although observational, support the use of XRN, especially among patients with high rates of psychiatric comorbidities and previous addiction treatment who are still struggling with AUDs and/or facing a period of vulnerability to relapse.