Objective: To examine changes in opioid use and healthcare costs among commercially insured patients with diabetic peripheral neuropathic pain (DPNP) who initiated duloxetine versus other standard of care (SOC) medications (tricyclic antidepressants, venlafaxine, gabapentin, pregabalin).
Research design and methods: Using an administrative claims database, patients with DPNP aged 18-64 who initiated duloxetine or SOC between March 1, 2005 and December 31, 2005 were identified. Initiation was defined as a 90-day clean period without the available study medication. Selected patients had 12 months of continuous enrollment before and after the index date, and at least one opioid dispensed in the prior 90 days. Duloxetine and SOC patients were further classified into continuous and non-continuous users based on whether the medication possession ratio was > or =0.8. Total opioid days, number of opioid prescriptions dispensed, and cumulative morphine equivalents were examined over the 12-month pre- and post-index periods. Multivariate regressions were applied to assess the changes (pre-index minus post-index) in opioid use (total, short-acting vs. long-acting) and healthcare costs, controlling for demographic and clinical characteristics.
Results: The study sample included 1281 patients: 98 duloxetine continuous, 243 duloxetine non-continuous, 195 SOC continuous, and 745 SOC non-continuous users. Controlling for demographic and clinical characteristics, duloxetine non-continuous and SOC (continuous and non-continuous) patients had significantly less reduction in total opioid days (-24.4, -23.7, -18.5, respectively, all p < 0.05) from the 12-month pre-index to the post-index period than duloxetine continuous patients. Compared with duloxetine non-continuous, SOC continuous, and SOC non-continuous users, duloxetine continuous users had a greater reduction in short-acting hydrocodone use (difference between the 12 month pre-index and post-index periods) in terms of the total number of prescriptions dispensed (adjusted differences: 1.5, 1.7, 1.7, respectively, all p < 0.05), total supply days (adjusted differences: 28.1, 27.3, 29.7, respectively, all p < 0.05), and morphine equivalent dosage (adjusted differences: 1290 mg, 1132 mg, 1127 mg, respectively, all p < 0.05). Duloxetine non-continuous patients had significantly higher adjusted total ($12,729, p < 0.05) and inpatient costs ($14,993, p < 0.05) than duloxetine continuous patients.
Limitations: Due to the use of a retrospective administrative claims database, this study is subject to selection bias between study cohorts, misidentification of DPNP and/or other comorbidities, and an inability to confirm adherence to therapy or assess indirect costs and costs of over-the-counter medications.
Conclusions: Among commercially insured patients with DPNP, continuous treatment with duloxetine was associated with a reduction in opioid use between the 12-month pre- and post-index periods compared with treatment with SOC or non-continuous treatment with duloxetine. Duloxetine continuous patients also incurred lower subsequent healthcare costs than non-continuous duloxetine patients.