This is the first study to estimate the incremental disease burden incurred by the migraine prophylaxis group compared with that of the non-prophylaxis group in Korea, based on national health insurance claims data. To compare the two groups, propensity score matching was conducted to balance the baseline characteristics and minimize selection bias.
In this retrospective study using the Korea national health insurance claims data, 1,636,105 migraine patients were identified, of whom 638,441 patients (39%) received at least one migraine prophylactic treatment. In two previous studies, 33.8% and 38.8% of migraine patients were estimated to require prophylactic treatment, which is consistent with our findings [6, 15].
In this study, the mean annual number of migraine-related outpatient visits per patient was 2.34 and 1.70, respectively. Similar trends were observed in the number of outpatient visits among migraine patients in Finland. The number of outpatient visits was 2.4 per patient-year in migraine patients receiving prophylactic treatments and 1.3 per patient-year in migraine patients only receiving acute treatment [13]. A cross-sectional analysis of survey data from France, Germany, Italy, Spain, and the United Kingdom estimated healthcare resource utilization from visits to the healthcare system 6 months before survey participation [16]. In this study, they reported the mean number of outpatient visits to neurologists was 0.19, and the mean number of ED visits was 0.46 in migraine patients. The estimated numbers of visits were lower than those reported in our study (2.23 outpatient visits to neurologists and 1.07 ED visits per patient-year). This may reflect differences in access to healthcare. Most European countries offer gatekeeping systems in which patients first see a general practitioner before a specialized physician; however, in Korea, patients can see a specialized physician directly without having to see a general practitioner [17, 18].
Yu et al. reported that the mean annual outpatient cost per patient was 46.5 USD among migraine patients in China [3]. In the current study, the mean annual outpatient costs were 102.37 USD in the prophylaxis group and 62.46 USD in the non-prophylaxis group. Migraine patients receiving prophylactic treatment were not included in the Chinese study. Thus, it was difficult to make an accurate comparison; however, in the non-prophylaxis group, the mean annual outpatient costs were similar to those in the study from China.
The current study did not observe the effect of migraine prophylactic treatments but instead evaluated the present status of disease burden in patients receiving migraine prophylactic treatments. These study findings indicate that migraine-related healthcare resource utilization and healthcare costs were significantly higher in migraine patients who received at least one migraine prophylactic treatment than in those who never received migraine prophylactic treatment. Although international guidelines recommend prophylactic treatment to reduce the burden of migraine and number of migraine attacks, prophylactic treatments for migraine remained underutilized in patients who appear to be clear candidates [6]. Moreover, most migraine patients who receive prophylactic treatments dropped out because of adverse events and the low efficacy of drugs [8]. Therefore, these findings suggest that despite the use of migraine prophylactic treatments, there are still unmet medical needs in the migraine patients who received prophylactic treatment. These results are consistent with those of previous studies investigating the effect of migraine prophylactic treatment and estimating the burden of unmet medical needs in migraine patients [8, 19]. As a result, these findings reveal that more effective strategies and treatments to prevent migraine attacks are needed to reduce the burden of migraine patients receiving prophylactic treatment.
The current study has several limitations. First, we used propensity score matching to minimize potential confounding effects on incremental disease burden. Although we accounted for measured confounders in the matching process, unmeasured confounders, such as monthly migraine days and clinical data representing the severity of migraine, which were not included in the claims data, were not considered and may have affected the analysis. However, we considered the use of triptan and ergotamine in the matching process to balance the severity of migraine between the two groups. Second, in the current study, since we identified the migraine patients who received at least one migraine prophylactic treatment as the prophylaxis group, treatment discontinuation that may be due to adverse effects or lack of efficacy may have occurred during the follow-up period and the compliance of the treatment was not considered. A previous study reported that prophylactic treatments were associated with a high rate of discontinuation due to adverse effects or lack of efficacy [13, 20]. Third, since the migraine prophylactic medications used in this study are migraine non-specific medications, some prophylactic medications could have been prescribed for other indications [21]. Thus, it may cause overestimation of certain treatments and costs. To minimize this probability, we selected the most frequently used prophylactic treatments in Korean clinical practice based on clinical expert opinions of Korean neurologists. In addition, we only included claims with a migraine diagnosis code (ICD-10 code G43) in the analysis.
Despite these limitations, this study has several strengths. First, this study showed that a sufficient reduction in the burden of migraine was not observed in patients receiving prophylaxis, although the guidelines recommend preventive therapies to reduce the burden of migraine. It is necessary to examine whether prophylactic treatments are used properly, and more effective treatment strategies are needed. Second, this result is meaningful because evidence regarding disease burden in patients with migraine who received prophylactic treatments is very scarce [13]. In addition, the results of this study are representative and reliable because we used a nationwide claims database that represents approximately 98% of the overall Korean population [9].