Predictors of Duloxetine versus Other Treatments among Veterans with Diabetic Peripheral Neuropathic Pain: A Retrospective Study
Disclosures: Shi, Zhao YN, and Liu are WOC employees/investigators for the Southeast Louisiana Veterans Health Care System. Shi has received research support (to Tulane) from American Diabetes Association, Eli Lilly and Company, Takeda, and Blue Cross Blue Shield of Louisiana. Zhao Y is an employee of Eli Lilly and Company and owns the company stocks. Zhao YN drafted the article. Liu, a certified advanced programmer for SAS®9, primarily performed data analysis. All authors participated in data interpretation. Shi, Liu, Zhao Y, Thethi, and Fonseca provided comments on the article. All authors contributed to the revisions of the article.
Abstract
Objective: This study used medical and pharmacy records from the Veterans Affairs (VA) health system to explore the predictors of duloxetine versus other treatments for patients with diabetic peripheral neuropathic pain (DPNP).
Methods: The electronic medical and pharmacy records from January 2004 to December 2008 were requested from the Veterans Integrated Service Network 16 data warehouse. All select patients received either duloxetine or other treatments [tricyclic antidepressants (TCAs), venlafaxine, gabapentin, and pregabalin] over the study period, with the first dispense date of the index agent as the index date. All patients must have 1+ prior DPNP diagnosis (ICD-9-CM: 250.6x or 357.2), but no diagnoses of prior depression (ICD-9-CM: 296.2, 296.3, 300.4, 309.1, or 311.0), fibromyalgia (ICD-9-CM: 729.1), or neuralgia (ICD-9-CM: 729.2). Logistic regression was used to examine the predictors of receiving duloxetine versus other treatments, controlling for demographics, comorbidities, prior pain level, prior use of other medications, and opioid use.
Results: The analytical sample included 2,694 patients (duloxetine cohort, n = 216; other-treatment cohort, n = 2,478). Prior uses of gabapentin (odds ratio [OR] = 13.66, 95% confidence interval [CI]: 9.70–19.24), TCAs (OR = 5.40, 95% CI: 3.73–7.82), or venlafaxine (OR = 3.67, 95% CI: 1.67–8.06) were strong predictors of duloxetine. Other comorbidities associated with duloxetine were anxiety (OR = 2.08, 95% CI: 1.40–3.08), cerebrovascular disease (OR = 1.44, 95% CI: 1.01–2.07), and substance abuse (OR = 2.11, 95% CI: 1.10–4.03). Prior opioid users were 1.47 (95% CI: 1.02–2.12) times as likely to receive duloxetine as those without prior opioid use. Patients with self-reported severe pain were 1.66 (95% CI: 1.11–2.50) times as likely to receive duloxetine as those with no pain reported.
Conclusion: DPNP patients in the VA healthcare system with prior other treatment use, select comorbid conditions, prior substance abuse, prior opioid use, and higher pain level were more likely to receive duloxetine.
