A systematic review of economic evaluations of pharmacological treatments for adults with chronic migraine

After deduplication, we identified 2927 citations from the database searches and reviewed them at the title/abstract review stage. After title/abstract review, 75 articles were reviewed in the full-text stage and nine articles [16,17,18,19,20,21,22,23,24] ultimately met the inclusion criteria for published peer-reviewed journal articles. One article was translated from Bulgarian [24] and one article was translated from Italian [22]. Our translation is available on request from the corresponding author. We also found an additional 51 reports and after the screening, we were left with seven reports [25,26,27,28,29,30,31] which were identified through other methods, such as Google, Google Scholar and websites of NICE, SMC, AWMSG and CADTH that were not published in peer review journals. The flow of studies through the systematic review process is presented in Fig. 1. We have narratively synthesised the reporting of the nine published journal articles separately from the seven reports.

Fig. 1figure 1

PRISMA flow chart diagram

Context, types of economic evaluation, interventions and comparators

Among the nine included studies from the database search, the majority of studies (n = 7) [17,18,19,20,21, 23, 24] were published during the last 5 years. All studies originated from high-income countries, mainly countries in European Union (n = 5) [17, 18, 21, 22, 24], including two each in the UK [16, 19] and the United States [20, 23]. All the evaluations were model-based cost-utility analyses involving hypothetical cohorts of the participants with chronic migraine [16,17,18,19,20,21, 23]. Among the nine included studies, four studies [16, 18, 19, 22] evaluated Botox and five studies [17, 20, 21, 23, 24] evaluated Erenumab for the treatment of chronic migraine. All the studies evaluating Botox [16, 18, 19, 22] compared the treatment with placebo or the best supportive care. As no established standard of care existed participants took their usual acute headache medications, placebo treatment (injection of saline water) was also considered as a comparator and assumed to represent consultant appointments to tailor acute medication, such as triptans. Those studies evaluating Erenumab compared the outcomes with Botox as a primary comparator or as a part of a sensitivity analysis. Further details about the interventions and comparators are presented in Table 1.

Table 1 Summary of the characteristics of included studies

In addition to the journal databases search, there were seven additional reports identified. Four reports were from the United Kingdom [28,29,30,31], two from Canada [25, 26] and one from the United States of America [27]. The reports evaluated the cost-effectiveness of Botox [25, 31], Erenumab [26, 27, 29], Fremanezumab [28], and Galcenzumab [28]. The treatments were compared either with placebo (best supportive care) [25, 27,28,29,30,31] or Botox [26, 28,29,30] and all of them were cost-utility analyses.

Modelling approach, perspective and time horizon used in the models

All of the journal articles were cost-utility studies. The majority (n = 6) [16, 18,19,20, 22, 24] of studies utilised a Markov state-transition model structure. Markov models commonly included health states stratified by the number of migraine or headache days per 28 days and allowed for treatment discontinuation or treatment on/off periods. One study used a decision tree, and the other two studies were based on the hybrid models: the first, a hybrid of decision tree and Markov model and the second, a hybrid of a Monte Carlo simulation and Markov model. The time horizon most utilised was two-years, however, the length varied from 1 year to 10 years. The studies from the United Kingdom were conducted from the National Health Service (NHS) perspective, the European studies were based on the societal perspective and the American studies were based on societal and payers’ perspectives.

Examination of the reports suggested two different types of model being implemented: Markov model with health states stratified by number of headache days per 28 days as mentioned above [31] or hybrid model with decision tree for 12-week assessment period, classifying participants as responders and non-responders, and Markov model for post-assessment with 12-week cycle lengths [25,26,27,28,29,30]. There were two studies which used a lifetime horizon (25 years) [29, 30] and the rest of them used a time horizon ranging from two to 10 years [25,26,27,28, 31]. Further details about the modelling approach, perspective and time horizon used in the models are presented in Table 2.

Table 2 Details of the Models and Model Inputs in the studies included in the Systematic ReviewResource utilisation and cost

In most of the studies published in the journal articles, the cost included the cost of pharmacotherapy and healthcare resource utilisation. Direct costs included the cost of the medicine and the administration costs, the costs of acute drugs used under usual care, and the costs of hospitalisation, physician, neurology appointment, specialist nurse and/or emergency department visits. Indirect costs for the societal perspective analyses included wages lost on workdays. In some studies (n = 4), the resource use data were obtained from the International Burden of Migraine Studies (IBMS) [16,17,18,19, 22, 24] and for other studies, they were either obtained from other published studies or local databases. The NHS reference costs were used as the source of cost inputs in the UK studies [16, 19], whereas other studies relied on other specific trials or publicly available sources. All studies included in the reports [25,26,27,28,29,30,31] also included similar resource usage data and unit cost data. The summary of the resource utilisation and the associated cost is illustrated in Table 2.

Price year/currency, the discount rate used and willingness-to-pay (WTP threshold)

All studies published in the journal articles reported the currency and price year. The UK studies [16, 19] used a 3.5% discount rate for costs and benefits, all other European studies [17, 18, 21, 22], except one [24] and the American studies [20, 23] applied a discount rate of 3% for both costs and outcomes. The studies from the UK [16, 19] us

留言 (0)

沒有登入
gif