OBJECTIVES : To characterize and compare health resource utilization (HRU) and costs among adherent and non-adherent chronic migraine (CM) patients prescribed prophylactic oral migraine-preventive medications (OMPMs).

METHODS

: We used de-identified data from the Optum^® Clinformatics^® Data Mart database to identify adult patients with CM who initiated an OMPM (subgroups based on antidepressant, antihypertensive, or antiepileptic use) between January 2010 and December 2017. Patients with depression, congestive heart failure, hypertension, or seizures were excluded. Patients were stratified into adherent and non-adherent groups based on a proportion of days covered greater than 0.80 in the year after OMPM initiation. The analysis followed 'intent-to-treat' design, with 1-year follow-up concurrent with adherence assessment. HRU and costs (2017 USD) during the 12 months follow-up were compared in 1:1 propensity score-matched adherent and non-adherent samples.

RESULTS

: Data for 1,733 matched pairs were analyzed. The likelihood of an emergency room (ER) visit was somewhat lower for the adherent versus non-adherent groups in the overall population [relative risk (RR) of 0.89, 95% confidence interval (CI) 0.80, 1.00] but was significantly lower for adherent patients among antidepressant users (RR=0.78, 95% CI, 0.67,0.91). Adherence among antihypertensive users was associated with lower risk of hospitalization (RR of 0.48, 95% CI, 0.26, 0.89) and fewer ER visits (mean 0.56 vs. 0.38, p=0.05). Among non-adherent antihypertensive users, inpatient and overall costs were $1,138 and $797 higher, respectively, although differences were not statistically significant. Migraine/headache-related costs represented approximately 30% of the inpatient cost differential.

CONCLUSIONS

: Migraine patients who are adherent to OMPMs may have reduced ER and inpatient utilization, as well as decreased costs, although these associations may depend on the type of OMPM used. Further investigation is warranted to understand how OMPM adherence affects healthcare utilization.