Received: 15 March 2021 | Accepted: 24 June 2021 DOI: 10.1002/ejp.1831 ORIGINAL ARTICLE Relevant factors for neurologists to define effectiveness of migraine preventive drugs and take decisions on treatment. My-­LIFE European Delphi survey Patricia Pozo-­Rosich1,2 | Raquel Gil-­Gouveia3 | Anne Donnet4 | Anne-­Christine Poole 5 | Astrid Gendolla 6 | Shazia Afridi 7 | Rainel Sanchez-­De la Rosa8 | Gisela M. Terwindt 9 | Cristina Tassorelli10,11 1 Headache Unit, Neurology Department, Hospital Universitari Vall d’Hebron, Abstract Barcelona, Spain Background: Clinical guidelines agree that preventive treatment should be consid- 2 Headache Research Group, Vall d’Hebron ered in patients with uncontrolled migraine despite acute medications or patients with Research Institute, Universitat Autònoma de ≥4 migraine days per month. However, the criteria to define the effectiveness of Barcelona, Barcelona, Spain 3 treatment and the factors that inform the decision to (dis)continue it are not clearly Headache Center, Neurology Department, Hospital da Luz, Lisboa, Portugal defined in clinical practice. 4 Pain Department, Timone Hospital, Methods: Overall, 148 healthcare practitioners from five European countries com- Marseille, France pleted a two-­wave questionnaire. The Steering Committee defined a simulated set 5 Private Practice: Oslo Headache Centre, of 108 migraine patient profiles based on the combination of five factors (frequency Oslo, Norway 6 of the attacks, intensity of the attacks, use of acute migraine medications, patient Private Practice: Essen, Germany 7 Headache Service, Department of perception and presence/absence of tolerable side effects). These profiles were used Neurology, Guy’s and St Thomas’ NHS in a Delphi survey among European neurologists to identify the criteria that should Trust, London, UK be used to decide treatment response and continuation using a conjoint analysis 8 Region Europe Medical Department, approach. Novartis Pharma AG, Basel, Switzerland 9 Results: Consensus was reached for 82/108 (76%) of profiles regarding treatment Department of Neurology, Leiden University Medical Center, Leiden, The response, and for 86/108 (80%) regarding treatment continuation. Multivariable lo- Netherlands gistic regression analysis showed that a ≥50% reduction in the use of acute migraine 10 Headache Science Center, IRCCS medications and positive patient's perception of treatment were the most important Mondino Foundation, Pavia, Italy 11 factors that lead to the decision of continuing (combined factors, OR = 18.3, 95% CI Department of Brain and Behavioral Sciences, University of Pavia, Pavia, Italy 13.4–­25.05). Conclusions: This survey identifies two relevant outcome measures: one objective Correspondence (use of acute migraine treatment medications) and one subjective (positive patient Patricia Pozo-­Rosich, Hospital Universitari Vall d’Hebron, Passeig de la Vall d'Hebron, perception) that guide the clinician decision to continue preventive treatment in mi- 119, 08035 Barcelona, Spain. graine patients. Email: ppozo@vhebron.net Significance: In clinical practice, criteria to define the effectiveness of migraine Funding information preventive treatment and factors that guide treatment stop or continuation are not Novartis Pharma AG financially supported clearly defined. In this simulated clinical setting study, a reduction in the use of acute the development of this project, the medical This is an open access article under the terms of the Creat​ive Commo​ns Attri​butio​n-­NonCo​mmerc​ial-­NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-­commercial and no modifications or adaptations are made. © 2021 The Authors. European Journal of Pain published by John Wiley & Sons Ltd on behalf of European Pain Federation - EFIC ® Eur J Pain. 2021;25:2177–2189.  wileyonlinelibrary.com/journal/ejp | 2177 2178 |    POZO-­ROSICH et al. writing assistance and the page processing charges for this article. The authors have migraine medications was the factor associated with preventive treatment effective- received no payment to write this article. ness definition. This study also revealed that factors strongly associated with the deci- sion of treatment continuation in real life are the acute migraine medications use and a positive patient's perception of treatment effectiveness. 1 | IN TRO D U C T ION cases, the decision to continue or discontinue with the pro- phylactic treatment is driven by the burden of the side effects. Migraine affects 80.8 million individuals in Western In recent years, new preventive treatments have shown Europe and more than 1.04 billion all over the world to be effective and decrease the intensity of migraine epi- (Stovner et al., 2018), being the first cause of disability sodes, together with a good tolerability and adhesion pro- in people 15–­49 years old worldwide (Steiner et al., 2018). file, in patients with episodic or chronic migraine (Sacco In Europe, 33.8% of migraine patients reported >5 head- et al., 2019). However, standards for evaluating the effec- ache days/month (Katsarava et al., 2018). The intensity, tiveness of preventive treatment exists in daily clinical duration and frequency of migraine attacks, as well as practice are lacking as are the criteria for treatment dis- the overall impact on patient's life, can be significantly continuation. The aim of the My-­LIFE Response criteria reduced with the appropriate use of acute and preventive project was to identify factors involved in the definition treatments. However, migraine treatment is not straightfor- of preventive treatment response and in the decision of ward, since it has to be patient-­tailored and the response to treatment continuation in daily clinical practice through preventive treatment is difficult to evaluate and not always a Delphi survey conducted among European healthcare complete and clear (American Headache Society (AHS), practitioners (HCP) with experience in migraine. 2019; Serrano et al., 2017). To further complicate the sit- We hypothesized that even though the main endpoint in uation, many of the available preventive treatments show CTs for migraine preventive treatments is a reduction in the limited tolerability, entailing long-­term adherence issues number of days with migraine, in clinical practice other fac- (American Headache Society (AHS), 2019). tors are also relevant to assess efficacy. Addition of preventive treatment should be considered in all patients with uncontrolled migraine despite acute treat- ment, and in patients with frequent headaches, i.e., with 4 or 2 | M ETHODS more headache days per month (American Headache Society (AHS), 2019). Patients suffering ≥4 headache days per month The My-­LIFE Response criteria project used a two-­round have poorer health-­related quality of life (HRQOL); greater modified Delphi procedure supported by a conjoint analy- impairment in work productivity and daily activities, as well sis approach. The authors met in a Steering Committee (SC) as higher healthcare resource utilization (Doane et al., 2020). to define the main factors for the evaluation of the efficacy Given the high prevalence and burden of migraine, preven- of a preventive treatment and prioritise them. The SC in- tive treatment is of the outmost importance. cluded eight members from France, Germany, Italy, Norway, Although frequency and severity of the attacks are the Portugal, Spain, The Netherlands and United Kingdom. two most objective reasons to start on preventive medication A non-­systematic literature review was conducted to in migraine patients, most guidelines suggest to monitor not identify the clinical factors that are used to define the re- only migraine frequency, but also impact and associated dis- sponse to migraine preventive treatment. On the basis of ability and/or (disease-­specific) HRQOL through migraine this literature search and their expertise, the SC agreed validated calendar tools and disability instruments (American on a list of seven factors as the most important criteria to Headache Society (AHS), 2019; Ahmed et al., 2019; Lanteri-­ evaluate treatment response and continuation. Of these, Minet et al., 2014; Steiner et al., 2019). There seems to the SC prioritized five factors and defined the categories be, however, no consensus on objective goals to determine within each of them according to their clinical experience therapeutic success. Most of the available instruments are (Figure 1). A set of simulated patient profiles was gener- Patient's reported outcome measures (PROMs; Ahmed et al., ated by the combination of two or three categories from 2019; Steiner et al., 2019; Tassorelli et al., 2018), that have each of the agreed factors: frequency of the attacks, in- been used and validated in clinical trials (CTs), but do not tensity of the attacks, use of acute migraine medications, necessarily reflect patients’ priorities in real life (Haywood patient's perception of preventive treatment effectiveness et al., 2018; Mannix et al., 2016; Smelt et al., 2014). In some and presence/absence of tolerable side effects (Figure 1). POZO-­ROSICH et al.     | 2179 F I G U R E 1 Design of patient profiles: factors and thresholds used aThese factors were not included in the defini on of pa ent profiles bTolerable side effects were defined as “side effects where the benefit from the preven ve treatment outweighs the inconveniences”. Tolerable side effects were defined as those in which the information regarding their clinical practice (Questionnaire benefit from the preventive treatment outweighs the in- in additional file). The participants were asked to classify conveniences from the side effects. All possible combi- the simulated patients according to treatment response as nations were made and a total of 108 simulated patient “responder”, “partial responder” or “non-­responder”. They profiles created. For all factors, the term “reduction” was could also select the option “more data needed” whenever referring to change from baseline. The questionnaire was they felt that data were insufficient for the classification, pre-­tested by the SC and the definition of factors and and the option “rarely found in clinical practice” if they thresholds adjusted to the final version (Figure 1). felt that it was the case. For each profile, participants were asked to decide whether preventive treatment should be continued (Figure 1). The estimated completion time of the 2.1 | Design of the Delphi questionnaire questionnaire was 2 hr. The Checklist for Reporting Results of Internet E-­Surveys The full set of 108 simulated patient profiles was integrated (CHERRIES) guidelines was followed (Eysenbach, 2004). in a questionnaire that included a section to characterize Completed CHERRIES checklist is included as supplemen- the participants’ profile and another section to collect tary material. 2180 |    POZO-­ROSICH et al. 2.2 | Panel selection and survey 3 | RESULTS administration 3.1 | Participants' profile A total of 264 HCP from five European countries (France, Germany, Italy, Spain and the United Kingdom) registered for All participants who completed both rounds of the survey the study. For participation, panelists had to fulfil the following (N = 148) were general neurologists with experience in mi- selection criteria: being either a general practitioner, neurologist, graine: 82/148 (55.4%) saw more than 50 patients per week internal medicine or pain specialist; having ≥2 years of experi- and 101/148 (68.2%) had more than 10 years of experience ence in migraine treatment, previous experience with the pre- with migraine. On average, 35.2% (SD: 21%, range: 5%–­ scription of migraine preventive treatments and prescribing them 99%) of all the patients seen by participants were suffering when deemed necessary, as well as currently having at least one from migraine, of whom 26.1% (SD: 15%, range: 5%–­76%) patient under a preventive treatment for chronic migraine. Of the from chronic migraine (Table S1). registered panelists, 205 met inclusion criteria, of whom 57 com- pleted the first round of the survey but were lost to follow-­up and 148 completed the two rounds of the Delphi questionnaire. The 3.2 | Current clinical practice for first round was conducted from July 2019 to February 2020; and episodic and chronic migraine treatment the second from April to May 2020. Before starting the study, consensus was defined as 70% or more of the panelists agreeing The use of migraine evaluation tools for diagnosis was similar on the classification regarding response to preventive treatment across countries. During the first visit of a migraine patient, or in the decision for treatment continuation for each patient 147/148 (99.3%) of participants assessed the diagnosis and profile. Only those profiles where no consensus was reached in the severity of migraine on the basis of a clinical interview the first round were held to a second round of voting without and 61/148 (41.2%) also used a patient diary. During follow- changing the questionnaire. The questionnaire was administered ­up, 131/148 (88.5%) of participants assessed the clinical situ- through an online platform that ensured data anonymity and ation with a clinical interview, 127/148 (85.8%) also used a confidentiality. As panelists had to give their opinion on their patient diary, 54/148 (36.5%) used a validated tool to assess experience without retrieving any patient data or information, chronic migraine (Table S2). The most used tools to assess no ethics committee approval or informed consent was needed. the impact of migraine were Migraine Disability Assessment (MIDAS; 83/148 (56.1%), ranging from 3/10 (30.0%) in Germany to 23/31 (74.2%) in Italy, p = .023) and Headache 2.3 | Statistical analysis Impact Test 6-­item (HIT-­6; 33.8%, similar across countries). Some differences between countries were observed in the For descriptive analysis, categorical variables were ex- use of a patient diary during follow-­up (ranging from 22/33 pressed as frequencies and percentages. Bivariate analy- (66.7%) in the United Kingdom to 30/31 (96.8%) in Italy, sis using Pearson's Chi-­square was performed to assess p = .005) and of validated scales (2/10 (20.0%) in Germany the relationship of each factor with response to preventive to 17/31 (54.8%) in Italy, p = .009; Table S2). treatment (responder, partial responder and non-­responder) Overall, 138/148 (93.3%) of participants had more than and treatment continuation or not. Stepwise multinomial 25% of their patients with migraine currently on preventive regression was performed to obtain the factors associated treatment (from 5/10 (50%) in Germany to 37/37 (100%) in with response to preventive treatment, and stepwise logistic France and Spain, without any statistical difference observed regression to obtain the factors associated with treatment among countries). When asked directly about the criteria continuation. Multiple correspondence analysis (similar that they used to assess the effectiveness of a preventive to factor analysis but for categorical variables) was per- treatment, participants rated frequency of attacks (132/148 formed. This is a technique used to graphically analyse the (89.2%) of participants reported to always use it), intensity dependence/independence relationships of a set of categor- of the attacks (108/148 (73.0%)) and occurrence of side ef- ical variables, based on the data from contingency tables. fects (treatment tolerability, 107/148 (72.3%)) as the most For this, it associates to each one of the modalities of the frequently used (Figure 2). Some differences between coun- table a point in the space with multiple dimensions, in such tries were observed in the frequency of use of the following a way that the proximity/distance between the calculated criteria: treatment tolerability (p = .002, ranging from 18/33 points reflects the relationships existing between them. (54.6%) in the United Kingdom to 9/10 (90.0%) in Germany), Data were analysed with SPSS version 22, and p < .05 was frequency of attacks (p = .027, from 25/33 (75.8%) in the considered statistically significant. No data were missing, United Kingdom to 37/37 (100%) in France and 10/10 (100%) lost or excluded of the analysis; therefore, no imputation of in Germany ), patient's perception of treatment (p = .022, missing data was conducted. from 8/31 (25.8%) in Italy to 26/37 (70.3%) in France and POZO-­ROSICH et al.     | 2181 F I G U R E 2 Criteria used to assess the 0% 20% 40% 60% 80% 100% effectiveness of preventive treatment in Frequency of aacks 89% 10% 1% patients with episodic or chronic migraine Disability (intensity of the aacks) 73% 22% 5% Treatment tolerability 72% 22% 5% Use of acute medica­on when having an aack 58% 32% 8% 1% Pa­ent percep­on of treatment 52% 33% 12% 2% 1% Dura­on of the aacks 51% 32% 16% 1% Response to acute medica­on 47% 34% 15% 5% Always Usually Some­mes Rarely Never TABLE 1 Definition of response to a preventive treatment according to the factors used to define patient profiles Consensus reached (n = 82) Responder Partial Responder Non-­responder Consensus not (n = 9) (n = 49) (n = 24) reached (n = 26) p value Frequency of the attacks .001* >50% reduction 7 (77.8%) 14 (28.6%) 2 (8.3%) 13 (50.0%) 49%–­25% reduction 2 (22.2%) 20 (40.8%) 7 (29.2%) 7 (26.9%) <25% reduction —­ 15 (30.6%) 15 (62.5%) 6 (23.1%) Intensity of the attacks .530 >50% reduction 4 (44.5%) 14 (28.6%) 6 (25.0%) 12 (46.2%) 49%–­25% reduction 3 (33.3%) 19 (38.8%) 7 (29.2%) 7 (26.9%) <25% reduction 2 (22.2%) 16 (32.6%) 11 (45.8%) 7 (26.9%) Use of acute migraine medications <.001* >50% reduction 9 (100%) 14 (28.6%) —­ 13 (50.0%) 49%–­25% reduction —­ 29 (59.2%) 3 (12.5%) 4 (15.4%) <25% reduction —­ 6 (12.2%) 21 (87.5%) 9 (34.6%) Patient's perception of treatment .038* Positive 8 (88.9%) 26 (53.1%) 8 (33.3%) 12 (46.2%) Negative 1 (11.1%) 23 (46.9%) 16 (66.7%) 14 (53.8%) Tolerable side effects .930 Absence 4 (44.4%) 25 (51.0%) 11 (45.8%) 14 (53.8%) Presence 5 (55.6%) 24 (49.0%) 13 (54.2%) 12 (46.2%) *p < .05. 7/10 (70.0%) in Germany ) and use of acute migraine medi- 3.3.1 | Factors used to classify patients' cations (p = .030, from 16/33 (48.5%) in the United Kingdom response to preventive treatment to 28/37 (75.7%) in France; Table S3). Consensus in the classification of response to treatment was reached for 82 of the 108 profiles (75.9%), with nine profiles 3.3 | Conjoint evaluation of the set of considered responders, 49 partial responders and 24 non-­ patient profiles responders (Table S4). None of the profiles reached consen- sus in being classified as “rarely found in clinical practice” or Overall, 148 participants evaluated 108 different simulated considered to lack data for the classification. profiles of patients with episodic or chronic migraine requir- Univariate analysis showed that three factors were signifi- ing preventive treatment. Overall, no differences were ob- cantly associated with response to preventive treatment: re- served between countries in terms on preventive treatment duction in the use of acute migraine medications (p < .001), response or continuation. reduction in the frequency of attacks (p = .001) and patient's 2182 |    POZO-­ROSICH et al. perception of treatment (p = .038; Table 1). Intensity of the the partial responder profiles for which agreement had been attacks was not significantly associated with treatment re- reached (22 of 29, Table 2). Lastly, for those profiles for which sponse classification (p = .530), nor was the presence of tol- no consensus on response to a preventive treatment had been erable side effects (p = .930; Table 1). previously reached, treatment continuation was agreed for 13 of The relative weight of each factor was analysed with a the 26 (55.5% of the profiles for which agreement on treatment logistic multinomial regression, a classification method used continuation had been reached; Table 2). to predict the relation between a set of nominal independent Univariate analysis showed that reduction in the use of variables. The association of each factor with treatment re- acute migraine medications (p < .001), patient's perception sponse was depicted in a multiple correspondence analysis, a of treatment (p < .001) and reduction in the frequency of at- data analysis technique used to detect and represent underly- tacks (p = .039) were associated with the decision to continue ing associations between nominal categorical data (Figure 3). or withdraw preventive treatment (Table 3). The model explained 51.0% of variability. Again, the relative weight of each factor was analysed with a multivariable logistic regression, and association of each factor with treatment continuation was depicted in a 3.3.2 | Factors used to decide continuation multiple correspondence analysis (Figure 4). The model ex- with preventive treatment plained 44.8% of variability and showed that the use of acute migraine medications and the patient's perception of treat- For each patient profile, participants evaluated whether the pre- ment were the most important factors for panelists to decide ventive treatment should be continued or withdrawn. Consensus treatment continuation. Participants were in favour of treat- was reached for 86 of the 108 profiles (79.6%): 44 (40.7%) for ment continuation in patients who had a >50% reduction in treatment continuation and 42 (38.9%) for treatment withdrawal. the use of acute migraine medications (versus. patients with Treatment continuation was agreed for all profiles classified as <25% reduction: OR = 34.7, 95% CI 30.80–­39.05), and pos- responders, and treatment withdrawal for all non-­responders itive perception of treatment effectiveness (versus. negative (Table 2; Table S4). Regarding partial responders, consensus perception; OR = 8.1, 95% CI 7.36–­8.83). When combin- was reached for 29 of the 49 partial responder profiles (59.2%). ing both factors, patients with both a >50% reduction in the Panelists were in favour of continuing treatment in 75.9% of use of acute migraine medications and positive perception of F I G U R E 3 Relationship between the factors defined to assess response to a preventive treatment and type of response to this preventive migraine treatment. Multiple correspondence plot. Total variance: 51.0%. Multiple correspondence plot representing the relationship between the factors defined to assess response to a preventive migraine treatment (frequency of attacks, intensity of the attacks, use of acute medications, patient perception with the preventive treatment and tolerable side effects) (blue dots) and the treatment response (red dots). Distance between a blue and a red dot represents their degree of relationship, the shorter the distance, the higher the relationship. Factor 1 accounts for the 28.8% of the variance and appears to be mainly capturing the variability observed regarding the treatment response (note that red dots R(responder) vs R(non-­responder) are shown as lowest and highest value across this axis. Factor 2 accounts for the 22.2% of the variance and seems to represent the variability among doubtful cases (in this case we see the red dots R(partial-­responder) vs R(rarely found) along this axis. POZO-­ROSICH et al.     | 2183 TABLE 2 Distribution of patient profiles according to treatment response and decision to continue with preventive treatment Consensus reached for treatment continuation (n = 86) Consensus not reached for treatment Decision to continue Continuation (n = 44) Withdrawal (n = 42) continuation (n = 22) Treatment response Responders, n (%) (n = 9) 9 (100%) —­ —­ Partial responders, n (%) (n = 49) 22 (44.9%) 7 (14.3%) 20 (40.8%) Non-­responders, n (%) (n = 24) —­ 24 (100%) —­ Consensus not reached for treatment 13 (50.0%) 11 (42.3%) 2 (7.7%) response, n (%) (n = 26) T A B L E 3 Factors used to decide Consensus reached (n = 86) continuation of preventive treatment in patients with episodic or chronic migraine, Continuation Withdrawal Consensus not univariate analysis (n = 44) (n = 42) reached (n = 22) p value Frequency of the attacks .039* >50% reduction 19 (43.2%) 8 (19.0%) 9 (40.9%) 49%–­25% reduction 15 (34.1%) 13 (31.0%) 8 (36.4%) <25% reduction 10 (22.7%) 21 (50.0%) 5 (22.7%) Intensity of the attacks .073 >50% reduction 18 (41.0%) 12 (28.6%) 6 (27.3%) 49%–­25% reduction 13 (29.5%) 11 (26.2%) 12 (54.5%) <25% reduction 13 (29.5%) 19 (45.2%) 4 (18.2%) Use of acute migraine <.001* medications >50% reduction 29 (65.9%) 2 (4.8%) 5 (22.7%) 49%–­25% reduction 15 (34.1%) 12 (28.6%) 9 (40.9%) <25% reduction 0 (0%) 28 (66.6%) 8 (36.4%) Patient's perception of <.001* treatment Positive 32 (72.7%) 12 (28.6%) 10 (45.5%) Negative 12 (27.3%) 30 (71.4%) 12 (54.5%) Tolerable side effects .430 Absence 25 (56.8%) 18 (42.9%) 11 (50.0%) Presence 19 (43.2%) 24 (57.1%) 11 (50.0%) *p < .05. treatment had the highest likelihood of treatment continua- all-­patient population, both patient's perception of effective- tion (OR = 18.3, 95% CI 13.41–­25.05; Table 4). ness of preventive treatment and reduction in the use of acute migraine medications were significantly associated with the decision of continuing with treatment; however, positive pa- 3.3.3 | Factors used to decide continuation tient's perception had a higher relative effect on the decision with preventive treatment in partial responders to continue treatment in partial responders than in all-­patient population. When panelists assessed partial responders, posi- In order to better understand the factors associated with tive perception (OR = 9.8, 95% CI 8.64–­11.02) and >50% re- the decision to continue treatment in partial responders, duction in the use of acute migraine medications (OR = 8.8, we specifically analysed this subset of profiles. As seen in 95% CI 7.48–­10.40) had similar weight (Table 4). 2184 |    POZO-­ROSICH et al. F I G U R E 4 Relationship between the factors defined to assess response to a preventive treatment and decision to continue with the preventive migraine treatment. Multiple correspondence plot. Total variance: 44.8%. Multiple correspondence plot representing the relationship between the factors defined to assess response to a preventive migraine treatment (frequency of attacks, intensity of the attacks, use of acute medications, patient perception with the preventive treatment and tolerable side effects) (blue dots) and the decision to continue with the preventive migraine treatment (red dots). Distance between a blue and a red dot represents their degree of relationship, the shorter the distance, the higher the relationship. Factor 1 is mainly capturing the variability regarding treatment continuation (No vs Yes) and gathers a variance of 28.0%, and Factor 2 seems to represent the variability among extreme profiles as far as the Frequency of migraine attacks. TABLE 4 Decision to continue with the preventive treatment in all patients and in partial responders All profiles (n = 108) Partial responders (n = 49) OR 95% CI p value OR 95% CI p value Use of acute migraine medications 50% reduction 34.7 30.8–­39.05 <.0001 8.8 7.48–­10.4 <.0001 25%–­50% reduction 7.0 6.33–­7.75 <.0001 3.12 2.71–­3.59 <.0001 <25% reduction —­ —­ —­ —­ —­ —­ Patient's perception of treatment Positive 8.1 7.36–­8.83 <.0001 9.8 8.64–­11.02 <.0001 Negative —­ —­ —­ —­ —­ Use of acute migraine medications/patient's perception of treatment >50% reduction/Positive 18.3 13.41–­25.05 <.0001 17.4 11.31–­26.82 <.0001 >50% reduction/Negative 2.27 1.82–­2.84 <.0001 1.79 1.31–­2.43 <.0001 25%−50% reduction/Positive 3.70 2.76–­4.97 <.0001 6.2 4.1–­9.27 <.0001 25%−50% reduction/Negative 0.46 0.37–­0.56 <.0001 0.63 0.47–­0.84 <.0001 <25% reduction/Positive 0.53 0.44–­0.64 <.0001 1.98 1.51–­2.58 <.0001 <25% reduction/Negative 0.066 0.06–­0.07 <.0001 0.20 0.18–­0.23 <.0001 and to decide on preventive treatment continuation in a simu- 4 | D IS C U SS ION A N D lated daily practice setting in five European countries. CO NC LU S ION S The main results show that the reduction in the use of acute migraine medications and a positive patient's perception of the The results from this Delphi study identify the factors used by preventive treatment are the most strongly associated factors neurologists to evaluate the response to preventive treatment with treatment response definition and treatment continuation. POZO-­ROSICH et al.     | 2185 Current guidelines of the International Headache Society The other factor significantly related to the evaluation (IHS) for controlled trials of preventive treatment of episodic of treatment response is patient's perception of treatment and chronic migraine recommend as primary endpoints the and, most importantly, to the decision to continue with the change in migraine days, change in severe-­to-­moderate head- preventive treatment. The patient's impression of change ache days, or responder rate (Diener et al., 2020; Tassorelli is a recommended secondary endpoint for controlled trials et al., 2018). When directly asked, panelists reported the use for preventive treatment of episodic and chronic migraine of frequency and intensity of the attacks as the most used cri- (Diener et al., 2020; Tassorelli et al., 2018), but is rarely used teria to assess the effectiveness of preventive treatment, con- for the evaluation of migraine patients, while it is commonly firming that international guidelines are well known among used for pain patients (Boyd et al., 2019; Derry et al., 2019). the participants. However, the conjoint analysis of simulated Indeed, the patient global impression of change scale (PGIC) patient profiles allowing to mimic the clinical practice in has been recently used only in two large controlled CTs to which the assessment is conducted considering all patient evaluate the effect of migraine therapies (Lipton et al., 2020; characteristics jointly showed that the reduction in the use Tepper et al., 2019). Another approach is the patient's assess- of acute migraine medications was deemed more relevant ment of the severity of migraine with the Global assessment when evaluating treatment response profiles, suggesting that of migraine severity (GAMS; Sajobi et al., 2019). maybe the reduction in the use of acute migraine medica- Unlike the use of acute migraine medications, patient's tions seems to be easier to quantify by clinicians in their daily perception is a subjective and general measurement based on clinical practice, than frequency or intensity of attacks. On personal opinion. Although effectiveness was the most used the other hand, a reduction in the use of acute medication outcome in published studies, the outcomes analysed in CTs could be probably related to frequency and intensity of the and in clinical practice differ. In 2014, a Delphi study con- attacks. Moreover, intensity of the attacks was not associated ducted among migraine patients on acute migraine medica- with the evaluation of treatment response in this conjoint tions showed that patients’ preferences are focused on fast analysis. This absence is surprising as, in recent years, spe- action and recurrence prevention (Smelt et al., 2014). Failure cifically with onabotulinumtoxinA and in patients who have to understand patient's preferences and perceptions may re- failed several preventive treatments, severity of pain has been duce adherence and compliance, subsequently limiting treat- proven to be important for patients, as it translates to a mean- ment success (American Headache Society (AHS), 2019). ingful improvement (Torres-­Ferrus et al., 2020). For partial responders this might be even more important, The use of acute migraine medications was also strongly given that those were a frequent scenario in our project associated with the decision to continue with preventive (45%), and panelists decided to continue treatment in 76% of treatment. Assessing acute migraine medications use is a rec- them. Therefore, a positive patient's perception had an even ommended secondary endpoint used in CTs of preventive treat- higher impact on the decision to continue with the preventive ments of episodic and chronic migraine (Diener et al., 2020; treatment than in all patient profiles population. Tassorelli et al., 2018). In this line, many trials showed a signif- Despite being a subjective measure, patient's percep- icant reduction in acute migraine medications versus placebo tion can be measured using validated PROMs. According (Andreou et al., 2018; Charles & Pozo-­Rosich, 2019; Loder & to Haywood et al. (Haywood et al., 2018), three PROMs Rizzoli, 2018). The reduction in medication consumption is have acceptable evidence of reliability and validity for likely a surrogate marker of a decrease in migraine frequency headache impact evaluation: HIT-­6 (Kosinski et al., 2003), and/or intensity of the attacks, which can be quantified in a Migraine-­Specific Quality-­of-­Life Questionnaire Version detailed and objective manner (Aurora et al., 2011). The re- 2.1 (MSQv2.1; Martin et al., 2000) and Patient Perception duction in acute migraine medications is also important for the of Migraine Questionnaire (PPMQ-­R; Kimel et al., 2008; well-­being of subjects, when considering the risks associated Revicki et al., 2006). While HIT-­6 and MSQv2.1 assess mi- with chronic use of NSAIDs or triptans. Furthermore, use of graine and headache-­specific impact, PPMQ-­R assesses the acute migraine medications >2 days per week is associated response and satisfaction with migraine-­specific treatment. with a reduction in the effectiveness of preventive treatment, Current guidance recommends HIT-­6, HALT-­30 (Headache-­ and it is a risk factor for medication overuse and chronification Attributed Lost Time-­30) and HURT (the Headache Under-­ (Bigal et al., 2008; Buse et al., 2019; Limmroth et al., 2002; Response to Treatment) for evaluating quality of life during Lipton et al., 2015). Since the use of acute migraine medica- preventive treatment (Ahmed et al., 2019; Steiner et al., 2019). tions is a factor associated with preventive treatment response Besides, MSQ v. 2.1 is also validated for patients undergoing and definition, our results suggest that neurologists are con- preventive migraine treatment (Cole et al., 2007). However, cerned about medication overuse and pay attention to con- currently used PROMs may not adequately reflect patients’ trol it. Reduction of acute migraine medications intake also satisfaction and a broader consensus on the most important involves a reduced economic burden (triptans are expensive) outcomes and how to assess them is required (Haywood and reduced acute-­medication side effects. et al., 2018; Mannix et al., 2016; Smelt et al., 2014). Thus, 2186 |    POZO-­ROSICH et al. extending education on the importance on patient's percep- conjoint analysis, regardless if an agreement was reached on tion as well as on the use of validated tools may help neurol- treatment response or treatment continuation or stop. ogists and HCP to better assess treatment response and guide Since the 108 patient profiles used in the project were arti- decisions on continuation with preventive treatment. ficially created, some profiles may not match the reality. The Panelists’ clinical practice showed that they commonly study design did not consider the classification of the sim- follow current recommendations (Ahmed et al., 2019; Steiner ulated profiles into low-­frequency episodic migraine, high-­ et al., 2019) with regard to the use of a patient diary when frequency episodic migraine and chronic migraine, and thus monitoring preventive treatment, which promotes adher- the results cannot be grouped into these three clinical scenar- ence and provides quantitative data to HCP for the decision ios. Moreover, duration of preventive treatment was not in- on treatment response and continuation. However, most cluded as a factor, and therefore conclusions cannot be drawn patient diaries are still far from being ideal, and there is an on its effect on decision to continue preventive treatment. urgent need for newly developed E-­health applications in- Other aspects, like patient clinical history and comorbidi- cluding time-­locked E-­diaries with the incorporation of the ties, concomitant medications or pregnancy or willingness Third Edition of the International Classification of Headache to become, that can influence treatment decisions, were not Disorders (ICHD-­3) criteria. Furthermore, <42% of health- considered when defining the patient profiles. Nevertheless, care providers used a validated scale to assess the impact panelists had to take into consideration five different factors of preventive treatment on migraine burden. The most used simultaneously, and this conjoint analysis approach allowed scales to assess patients with migraine in current practice us to assess the relative weight that each individual factor were MIDAS (Stewart et al., 2001) and HIT-­6. Both scales plays in their treatment decisions. are short (MIDAS, five items and HIT-­6, six items) and easy Even though all participants were general neurologists, to administer, which may increase their use. The use of scales 86% of them reported the use of a patient diary indicating a does not necessarily mean that they are valued as decision special interest and expertise in headache; subsequently, re- criteria, but it may warrant that they are probably valued and sults reflect real-­life clinical practice in Europe. Although the considered useful. sample size was not very large, we could rely on an enormous Lastly, preventive treatments are not devoid of adverse data set (108 profiles assessed by 148 panelists) to allow events (American Headache Society (AHS), 2019) and this proper and robust multivariable analysis of the simulated pro- could partially explain their low use. In our survey, the pres- files. A possible limitation was the questionnaire length (esti- ence of tolerable side effects did not emerge as a relevant fac- mated completion time for the questionnaire was 2 hr), which tor in the decision of treatment withdrawal. This suggests that might have caused some bias due to tiredness; in addition, the panelists relied more on the patients’ perception of treatment evaluation order of the simulated profiles was the same for effectiveness than on the impact of tolerable side effects, pos- all panelists. In order to minimize the risk of bias, the ques- sibly considering the former as the most reliable reflection tionnaire's platform allowed to save the answers, close the of the sum of positive and negative effects of treatment on platform and continue with the survey later on; however, the patients’ lives. In this frame, it is indeed possible that in the platform specifically programmed for this Delphi consensus panelists’ interpretation, patients are willing to endure tol- was not geared to collect who answered the questionnaire at erable side effects as long as the treatment is taking away once and how long participants were connected. a relevant part of their disease burden. This seems an inter- It would also have been of interest to evaluate if working esting working hypothesis that will need specifically aimed in a hospital or in the private setting had an impact on the investigations. answers; however, this information was not captured by the questionnaire. Lastly, our project included a limited sample size in some countries that might have influenced the differ- 4.1 | Strengths and limitations ences observed among countries. Even though general practitioners, neurologists, internal medicine and pain specialists were invited to participate, the 4.2 | Further directions inclusion criteria regarding experience in migraine and pre- ventive treatments made that only general neurologists were New preventive treatments for migraine have recently eventually valid for answering the full Delphi survey. reached the market (e.g.: monoclonal antibodies against cal- Even if consensus threshold could have been more strin- citonin gene-­related peptide [CGRP] or its receptor). These gent, when defining it, at the beginning of the project, after treatments are costly and will translate on higher pressure a non-­systematic literature review, we considered that a 70% on clinicians as regards day-­to-­day decisions about when to of agreement was appropriate. In this regard, it has to be continue or not with such expensive treatments, in particular, mentioned that all patient profiles were considered for the when dealing with partial responders. Therefore, additional POZO-­ROSICH et al.     | 2187 research on factors that determine the response to preventive Research Foundation, Novartis and PERIS. RS is an em- treatment, as well as on tools that adequately assess this and ployee and stakeholder of Novartis Pharma AG. CT has re- consider patients’ perception, would help decision making ceived personal fees from Novartis Pharma AG during the and improve the management of migraine in clinical practice. conduct of the project. Outside the submitted work, CT has As a conclusion, in a simulated clinical setting, a reduction received fees for advisory board's contribution, clinical trials in the use of acute migraine medications is the most strongly participation or medical education from Allergan, Amgen, associated factor with the preventive treatment effectiveness Eli Lilly, Teva and Medscape. CT has received independ- definition. The reduction in the use of acute migraine med- ent support from the Italian Ministry of Health, European ications, followed by a positive patient's perception of treat- Commission and Migraine Research Foundation. GT has re- ment effectiveness, is the factor that most strongly associated ceived personal fees from Novartis Pharma AG during the with the decision of treatment continuation. conduct of the project. Outside the submitted work, GT has received grants or consultancy support from Allergan, Eli ACKNOWLEDGEMENTS Lilly, Novartis, Teva, and independent support from Dutch We thank the panelists for their participation in the survey, Research Council, NIH, European Community, Dutch Heart thanks to Maite Artés, Nathalie Bofarull and Malena Águila Foundation and Dutch Brain Foundation. (Adelphi Targis S.L.) for the project management and contri- bution along the project and to Alba Gomez and Alba Llopis AUTHORS' CONTRIBUTIONS for their medical writing assistance with the methods and re- All authors were involved in the writing of introduction and sults sections. discussion and the review of all the article and meet crite- ria for authorship as recommended by the International CONFLICT OF INTERESTS Committee of Medical Journal Editors (ICMJE). SA has received personal fees from Novartis Pharma AG dur- ing the conduct of the project. Outside the submitted work, SA ETHICAL APPROVAL AND CONSENT TO has perceived personal fees from Teva as a speaker and per- PARTICIPATE sonal fees from Eli Lilly for advisory board participation. AD Given that panelists had to give their opinion based on their has received personal fees from Novartis Pharma AG during own experience without retrieving any patient data or infor- the conduct of the project. Outside the submitted work, AD mation, and the anonymity of the questionnaire, no ethics has received fees from Allergan, Eli Lilly, Grünenthal, MSD, committee approval was required. All panelists gave their Novartis, Orkyn, Pfizer, Saint-­Jude, Sanofi and Teva. AG formal consent to participate in the Delphi survey. has received personal fees from Novartis Pharma AG dur- ing the conduct of the project. Outside the submitted work, CONSENT FOR PUBLICATION AG has received fees for advisory boards and lectures from Not applicable. Allergan, Eli Lilly, Grünenthal, Hexal, Novartis, Sanofi and Teva. RG has received personal fees from Novartis Pharma DATA AVAILABILIT Y STATEMENT AG during the conduct of the project. Outside the submitted The datasets used and/or analysed during this study are avail- work, RGG has received fees for advisory board's contribu- able from the corresponding author on reasonable request. tion, clinical trials participation or medical education from Allergan, Amgen, Lundbeck, Novartis, Pfizer and Teva. RG ORCID has received support for headache research from Sociedade Patricia Pozo-­Rosich https://orcid. Portuguesa de Cefaleias and Fundação para a Ciência e org/0000-0003-0796-4702 Tecnologia. AP has received personal fees from Novartis Raquel Gil-­Gouveia https://orcid. Pharma AG during the conduct of the project. Outside the org/0000-0002-4256-4256 submitted work she has received fees for advisory board's Anne Donnet https://orcid.org/0000-0003-0034-5247 contribution, clinical trials or medical education from Anne-­Christine Poole https://orcid. Allergan, Eli Lilly, Lundbeck, Novartis and Teva. PP has re- org/0000-0002-2872-4110 ceived personal fees from Novartis Pharma AG during the Astrid Gendolla https://orcid.org/0000-0002-9884-3158 conduct of the project. Outside the submitted work, PP-­R has Shazia Afridi https://orcid.org/0000-0002-8487-5677 received fees for advisory board's contribution, clinical trials Rainel Sanchez-­De la Rosa https://orcid. participation or medical education from Allergan, Almirall, org/0000-0002-7134-1203 Amgen, Biohaven, Chiesi, Eli Lilly, Medscape, Novartis Gisela M. Terwindt https://orcid. and Teva. PP has received support for headache research org/0000-0003-3140-6882 from La Caixa Foundation, AGAUR, European Commission Cristina Tassorelli https://orcid. (ERANet Neuron), Instituto de Salud Carlos III, Migraine org/0000-0003-1513-2113 2188 |    POZO-­ROSICH et al. R E F E R E NC E S Journal of Medical Internet Research, 6, e34. https://doi. org/10.2196/jmir.6.3.e34 Ahmed F., Bahra A., Tyagi A., & Weatherby S. (2019). BASH National Haywood, K. L., Mars, T. S., Potter, R., Patel, S., Matharu, M., & headache management system for adults. British Association for the Underwood, M. (2018). Assessing the impact of headaches and the Study of Headache. outcomes of treatment: A systematic review of patient-­reported out- American Headache Society (AHS). (2019). The American headache come measures (PROMs). Cephalalgia, 38, 1374–­1386. https://doi. society position statement on integrating new migraine treatments org/10.1177/03331​02417​731348 into clinical practice. Headache, 59, 1–­18. https://doi.org/10.1111/ Katsarava, Z., Mania, M., Lampl, C., Herberhold, J., & Steiner, T. J. head.13456 (2018). Poor medical care for people with migraine in Europe –­ Andreou, A. P., Trimboli, M., Al-­Kaisy, A., Murphy, M., Palmisani, Evidence from the Eurolight study. The Journal of Headache and S., Fenech, C., Smith, T., & Lambru, G. (2018). Prospective real-­ Pain, 19, 1–­9. https://doi.org/10.1186/s1019​4-­018-­0839-­1 world analysis of OnabotulinumtoxinA in chronic migraine post-­ Kimel, M., Hsieh, R., McCormack, J., Burch, S. P., & Revicki, D. A. National Institute for Health and Care Excellence UK technology (2008). Validation of the revised Patient Perception of Migraine appraisal. European Journal of Neurology, 25, 1069–­e83. https:// Questionnaire (PPMQ-­R): Measuring satisfaction with acute mi- doi.org/10.1111/ene.13657 graine treatment in clinical trials. Cephalalgia, 28, 510–­523. https:// Aurora, S. K., Winner, P., Freeman, M. C., Spierings, E. L., Heiring, doi.org/10.1111/j.1468-­2982.2007.01524.x J. O., DeGryse, R. E., VanDenburgh, A. M., Nolan, M. E., & Kosinski, M., Bayliss, M. S., Bjorner, J. B., Ware, J. E., Garber, W. H., Turkel, C. C. (2011). OnabotulinumtoxinA for treatment of Batenhorst, A., Cady, R., Dahlöf, C. G. H., Dowson, A., & Tepper, chronic migraine: Pooled analyses of the 56-­ week PREEMPT S. (2003). A six-­item short-­form survey for measuring headache im- clinical program. Headache, 51, 1358–­ 1373. https://doi. pact: The HIT-­6TM. Quality of Life Research, 12, 963–­974. https:// org/10.1111/j.1526-­4610.2011.01990.x doi.org/10.1023/A:10261​19331193 Bigal, M. E., Serrano, D., Buse, D., Scher, A., Stewart, W. F., & Lanteri-­Minet, M., Valade, D., Geraud, G., Lucas, C., & Donnet, A. Lipton, R. B. (2008). Acute migraine medications and evo- (2014). Revised French guidelines for the diagnosis and manage- lution from episodic to chronic migraine: A longitudinal ment of migraine in adults and children. J Headache Pain, 15, 2. population-­based study. Headache, 48, 1157–­1168. https://doi. https://doi.org/10.1186/1129-­2377-­15-­2 org/10.1111/j.1526-­4610.2008.01217.x Limmroth, V., Katsarava, Z., Fritsche, G., Przywara, S., & Diener, H.-­C. Boyd, A., Bleakley, C., Hurley, D. A., Gill, C., Hannon-­Fletcher, (2002). Features of medication overuse headache following over- M., Bell, P., & McDonough, S. (2019). Herbal medicinal prod- use of different acute headache drugs. Neurology, 59, 1011–­1014. ucts or preparations for neuropathic pain. Cochrane Database https://doi.org/10.1212/WNL.59.7.1011 Systematic Review, 2019. https://doi.org/10.1002/14651​858. Lipton, R. B., Cohen, J. M., Gandhi, S. K., Yang, R., Yeung, P. P., CD010​528.pub4 & Buse, D. C. (2020). Effect of fremanezumab on quality of life Buse, D. C., Greisman, J. D., Baigi, K., & Lipton, R. B. (2019). Migraine and productivity in patients with chronic migraine. Neurology, 95, progression: A systematic review. Headache, 59, 306–­338. https:// e878–­e888. https://doi.org/10.1212/wnl.00000​00000​010000 doi.org/10.1111/head.13459 Lipton, R. B., Fanning, K. M., Serrano, D., Reed, M. L., Cady, R., & Charles, A., & Pozo-­Rosich, P. (2019). Targeting calcitonin gene-­related Buse, D. C. (2015). Ineffective acute treatment of episodic migraine peptide: A new era in migraine therapy. Lancet, 394, 1765–­1774. is associated with new-­onset chronic migraine. Neurology, 84, 688–­ https://doi.org/10.1016/S0140​-­6736(19)32504​-­8 695. https://doi.org/10.1212/WNL.00000​00000​001256 Cole, J. C., Lin, P., & Rupnow, M. F. T. (2007). Validation of the Loder, E., & Rizzoli, P. (2018). Pharmacologic prevention of mi- Migraine-­Specific Quality of Life Questionnaire version 2.1 (MSQ graine: A narrative review of the state of the art in 2018. Headache, v. 2.1) for patients undergoing prophylactic migraine treatment. 58(Suppl 3), 218–­229. https://doi.org/10.1111/head.13375 Quality of Life Research, 16, 1231–­1237. https://doi.org/10.1007/ Mannix, S., Skalicky, A., Buse, D. C., Desai, P., Sapra, S., Ortmeier, B., s1113​6-­007-­9217-­1 Widnell, K., & Hareendran, A. (2016). Measuring the impact of mi- Derry, S., Bell, R. F., Straube, S., Wiffen, P. J., Aldington, D., & Moore, graine for evaluating outcomes of preventive treatments for migraine R. A. (2019). Pregabalin for neuropathic pain in adults. Cochrane headaches. Health and Quality of Life Outcomes, 14, 143. https:// Database Systematic Review, 2019. https://doi.org/10.1002/14651​ doi.org/10.1186/s1295​5-­016-­0542-­3 858.CD007​076.pub3 Martin, B. C., Pathak, D. S., Sharfman, M. I., Adelman, J. U., Taylor, F., Kwong, Diener, H.-­ C., Tassorelli, C., Dodick, D. W., Silberstein, S. D., W. J., & Jhingran, P. (2000). Validity and reliability of the migraine-­ Lipton, R. B., Ashina, M., Becker, W. J., Ferrari, M. D., Goadsby, specific quality of life questionnaire (MSQ version 2.1). Headache, 40, P. J., Pozo-­Rosich, P., Wang, S.-­J., Houle, T. T., van den Hoek, T. 204–­215. https://doi.org/10.1046/j.1526-­4610.2000.00030.x C., Martinelli, D., & Terwindt, G. M. (2020). Guidelines of the Revicki, D. A., Kimel, M., Beusterien, K., Kwong, J. W., Varner, J. A., International Headache Society for controlled trials of preven- Ames, M. H., Mahajan, S., & Cady, R. K. (2006). Validation of the tive treatment of migraine attacks in episodic migraine in adults. revised Patient Perception of Migraine Questionnaire: Measuring Cephalalgia, 40, 1026–­ 1044. https://doi.org/10.1177/03331​ satisfaction with acute migraine treatment. Headache, 46, 240–­252. 02420​941839 https://doi.org/10.1111/j.1526-­4610.2006.00289.x Doane, M. J., Gupta, S., Fang, J., Laflamme, A. K., & Vo, P. (2020). The Sacco, S., Bendtsen, L., Ashina, M., Reuter, U., Terwindt, G., humanistic and economic burden of migraine in Europe: A cross-­ Mitsikostas, D.-­D., & Martelletti, P. (2019). European headache sectional survey in five countries. Neurology and Therapy, 9, 535–­ federation guideline on the use of monoclonal antibodies acting 549. https://doi.org/10.1007/s4012​0-­020-­00196​-­2 on the calcitonin gene related peptide or its receptor for migraine Eysenbach, G. (2004). Improving the quality of Web surveys: The prevention. The Journal of Headache and Pain, 20, 6. https://doi. Checklist for Reporting Results of Internet E-­Surveys (CHERRIES). org/10.1186/s1019​4-­018-­0955-­y POZO-­ROSICH et al.     | 2189 Sajobi, T. T., Amoozegar, F., Wang, M., Wiebe, N., Fiest, K. M., The Lancet Neurology, 17, 954–­976. https://doi.org/10.1016/S1474​ Patten, S. B., & Jette, N. (2019). Global assessment of migraine -­4422(18)30322​-­3 severity measure: Preliminary evidence of construct validity. BMC Tassorelli, C., Diener, H. C., Dodick, D. W., Silberstein, S. D., Lipton, R. Neurology, 19, 53. https://doi.org/10.1186/s1288​3-­019-­1284-­8 B., Ashina, M., Becker, W. J., Ferrari, M. D., Goadsby, P. J., Pozo-­ Serrano, D., Lipton, R. B., Scher, A. I., Reed, M. L., Stewart, W. F., Rosich, P., & Wang, S. J. (2018). Guidelines of the International Adams, A. M., & Buse, D. C. (2017). Fluctuations in episodic and Headache Society for controlled trials of preventive treatment of chronic migraine status over the course of 1 year: Implications for di- chronic migraine in adults. Cephalalgia, 38, 815–­832. https://doi. agnosis, treatment and clinical trial design. The Journal of Headache org/10.1177/03331​02418​758283 and Pain, 18. https://doi.org/10.1186/s1019​4-­017-­0787-­1 Tepper, S. J., Krege, J. H., Lombard, L., Asafu-­Adjei, J. K., Dowsett, S. A., Smelt, A. F. H., Louter, M. A., Kies, D. A., Blom, J. W., Terwindt, Raskin, J., Buchanan, A. S., & Friedman, D. I. (2019). Characterization G. M., van der Heijden, G. J. M. G., De Gucht, V., Ferrari, M. D., of dizziness after lasmiditan usage: Findings from the SAMURAI and & Assendelft, W. J. J. (2014). What do patients consider to be the SPARTAN acute migraine treatment randomized trials. Headache, 59, most important outcomes for effectiveness studies on migraine treat- 1052–­1062. https://doi.org/10.1111/head.13544 ment? Results of a Delphi study. PLoS ONE, 9, e98933. https://doi. Torres-­Ferrus, M., Gallardo, V. J., Alpuente, A., & Pozo-­Rosich, P. org/10.1371/journ​al.pone.0098933 (2020). Influence of headache pain intensity and frequency on Steiner, T. J., Jensen, R., Katsarava, Z., Linde, M., MacGregor, E. A., migraine-­related disability in chronic migraine patients treated with Osipova, V., Paemeleire, K., Olesen, J., Peters, M., & Martelletti, P. OnabotulinumtoxinA. The Journal of Headache and Pain, 21, 88. (2019). Aids to management of headache disorders in primary care. https://doi.org/10.1186/s1019​4-­020-­01157​-­8 The Journal of Headache and Pain, 20, 57. https://doi.org/10.1186/ s1019​4-­018-­0899-­2 SUPPORTING INFORMATION Steiner, T. J., Stovner, L. J., Vos, T., Jensen, R., & Katsarava, Z. (2018). Migraine is first cause of disability in under 50 s: Will health politi- Additional supporting information may be found online in cians now take notice? The Journal of Headache and Pain, 19, 17. the Supporting Information section. https://doi.org/10.1186/s1019​4-­018-­0846-­2 Stewart, W. F., Lipton, R. B., Dowson, A. J., & Sawyer, J. (2001). Development and testing of the Migraine Disability Assessment How to cite this article: Pozo-­Rosich, P., Gil-­Gouveia, (MIDAS) Questionnaire to assess headache-­ related disability. R., Donnet, A., Poole, A.-­C., Gendolla, A., Afridi, S., Neurology, 56, S20–­8. https://doi.org/10.1212/wnl.56.suppl_1.s20 Sanchez-­De la Rosa, R., Terwindt, G. M., & Tassorelli, Stovner, L. J., Nichols, E., Steiner, T. J., Abd-­Allah, F., Abdelalim, C. (2021). Relevant factors for neurologists to define A., Al-­Raddadi, R. M., Ansha, M. G., Barac, A., Bensenor, I. M., effectiveness of migraine preventive drugs and take Doan, L. P., Edessa, D., Endres, M., Foreman, K. J., Gankpe, F. decisions on treatment. My-­LIFE European Delphi G., Gopalkrishna, G., Goulart, A. C., Gupta, R., Hankey, G. J., Hay, S. I., … Murray, C. J. L. (2018). Global, regional, and na- survey. European Journal of Pain, 25, 2177–­2189. tional burden of migraine and tension-­type headache, 1990–­2016: https://doi.org/10.1002/ejp.1831 A systematic analysis for the Global Burden of Disease Study 2016.