Persistence with migraine prophylactic treatment and acute migraine medication utilization in the managed care setting.

OBJECTIVE: The aim of this study was to describe persistence with migraine prophylactic treatment and acute migraine medication utilization in patients prescribed migraine prophylaxis.
METHODS: For this retrospective cohort study, the Health Core Integrated Research Database provided pharmacy/medical claims data from 5 commercial health insurance plans (ie, excluding Medicare and Medicaid) on adult patients with migraine. Eligible patients had >or=1 pharmacy claim for a migraine prophylactic medication between July 1, 2000, and May 31, 2005, and >or=12 U of any combination of acute treatment (serotonin receptor agonist [triptan], ergotamine, or ergotamine combination) dispensed during the 180-day period preceding a first pharmacy claim for a prophylactic medication (index date). The prophylactic medication identified at index date was used for categorizing patients into 1 of 4 cohorts: amitriptyline, propranolol/timolol, divalproex sodium, or topiramate (reference). Kaplan-Meier curves were used for evaluating unadjusted risk for discontinuation over time, and a multivariate Cox proportional hazards model was developed to analyze factors associated with discontinuation of prophylactic medication.
RESULTS: A total of 12,783 patients met the inclusion criteria and were included in the analysis (amitriptyline, 3749; propranolol/timolol, 2718; divalproex sodium, 1644; and topiramate, 4672). The mean (SD) ages were not significantly different across cohorts (43.9 [11.3], 42.0 [11.1], 43.1 [11.3], and 43.9 [10.6] years, respectively). The mean duration of treatment was significantly longer (131 [184] days) with topiramate compared with amitriptyline (94 [152] days), propranolol/ timolol (119 [180] days), and divalproex sodium (109 [158] days) (P < 0.001, P = 0.005, and P<0.001,respectively). The risks for discontinuing prophylactic treatment were 23%, 6%, and 11% higher with amitriptyline, propranolol/timolol, and divalproex sodium, respectively, compared with topiramate (P<0.001, P = 0.024, and P <0.001). Patients prescribed topiramate had a higher mean consumption rate of triptans preindex; postindex, decreases in triptan use were observed in all cohorts, although the magnitude of the decrease was greatest in patients prescribed topiramate compared with the other cohorts.
CONCLUSIONS: In this study, prescription of topiramate was associated with greater persistence with prophylactic treatment than the other prophylactic drugs. Furthermore, greater reductions in acute treatment utilization, particularly triptans, were observed among patients prescribed topiramate compared with the other prophylactic cohorts.