Identifying the factors underlying discontinuation of triptans.

OBJECTIVE: To identify factors associated with triptan discontinuation among migraine patients.
BACKGROUND: It is unclear why many migraine patients who are prescribed triptans discontinue this treatment. This study investigated correlates of triptan discontinuation with a focus on potentially modifiable factors to improve compliance.
METHODS: This multicenter cross-sectional survey (n = 276) was performed at US tertiary care headache clinics. Headache fellows who were members of the American Headache Society Headache Fellows Research Consortium recruited episodic and chronic migraine patients who were current triptan users (use within prior 3 months and for ≥1 year) or past triptan users (no use within 6 months; prior use within 2 years). Univariate analyses were first completed to compare current triptan users to past users for: migraine characteristics, other migraine treatments, triptan education, triptan efficacy, triptan side effects, type of prescribing provider, Migraine Disability Assessment (MIDAS) scores and Beck Depression Inventory (BDI) scores. Then, a multivariable logistic regression model was selected from all possible combinations of predictor variables to determine the factors that best correlated with triptan discontinuation.
RESULTS: Compared with those still using triptans (n = 207), those who had discontinued use (n = 69) had higher rates of medication overuse (30 vs. 18%, P = .04) and were more likely to have ever used opioids for migraine treatment (57 vs. 38%, P = .006) as well as higher MIDAS (mean 63 vs. 37, P = .001) and BDI scores (mean 10.4 vs. 7.4, P = .009). Compared with discontinued users, current triptan users were more likely to have had their triptan prescribed by a specialist (neurologist, headache specialist, or pain specialist) (74 vs. 54%, P = .002) and were more likely to report headache resolution (53 vs. 14%, P < .001) or a reduction in pain intensity (71 vs. 28%, P < .001) most of the time from their triptan. On a 1-5 scale (1 = disagree, 5 = agree), triptan users felt they had more: control over their migraine attacks (2.9 vs. 2.1), confidence in their prescribing provider (4.5 vs. 4.0), and were more educated about triptan use (4.2 vs. 3.7) compared with triptan discontinuers (P < .001 for all comparisons). Although both current and prior users reported similar rates of side effects (48 vs. 43%, P = .44), of those who discontinued use, the main reasons were for lack of effect (44%) and side effects (29%). Our multivariable modeling revealed that the strongest correlate of triptan discontinuation was lack of efficacy (odds ratio = 17, 95% confidence interval [8.8, 33.0]). Other factors associated with discontinuation included MIDAS > 24 (2.6, [1.5, 4.6]), BDI >4 (2.5, [1.4, 4.5]), and a history of ever using opioids for migraine therapy (2.2, [1.3, 3.8]). Having a triptan prescribed by a specialist and using at least 1 other abortive medication with the triptan were associated with a decreased likelihood of triptan discontinuation (0.41, [0.2-0.7] and 0.44 [0.3, 0.8], respectively).
CONCLUSIONS: As expected, discontinuation was most correlated with lack of efficacy, but other important factors associated with those who had discontinued use included greater migraine-related disability, depression, and the use of opioids for migraine attacks. Compared with patients who had discontinued triptans, current triptan users felt more: educated about their triptan, control over their migraine attacks, and confidence in their prescribing provider. Current triptan users had their triptan prescribed by a specialist and used other abortive medications with their triptan more often compared with patients who had discontinued triptans. Given the cross-sectional nature of this study, we cannot determine if these factors contributed to triptan discontinuation or reflect the impact of such discontinuation. Interventions that address modifiable risk factors for triptan discontinuation may decrease the likelihood of triptan discontinuation and thus improve overall migraine control. Because lack of efficacy was most strongly associated with triptan discontinuation, future research should determine why triptans are effective for some patients but not others.