Acute Migraine Treatment With Oral Triptans and NSAIDs in a Managed Care Population

Headache 2008 Merck and Co., Inc. Journal compilation 2008 American Headache Society ISSN 0017-8748 doi: 10.1111/j.1526-4610.2007.01055.x Published by Wiley Periodicals, Inc. Research Submission Acute Migraine Treatment With Oral Triptans and NSAIDs in a Managed Care Population Daisy S. Ng-Mak, PhD; X. Henry Hu, MD, PhD; Ya-Ting Chen, PhD; Larry Ma, PhD Objective. To describe the use of oral triptans with or without nonsteroidal anti-inflammatory drugs (NSAIDs) for acute migraine treatment in a managed care population and its potential impact on functionality. Background. Prescription or over-the-counter NSAIDs with or without oral triptans are commonly used for treatment of acute migraine pain. Little is known about patients treatment strategy when they have had experiences using NSAIDs and oral triptan cotherapy and the relationship between treatment strategy and migraine symptoms and functionality. Method. Migraineurs identified from an administrative claims database were surveyed for their use of oral triptans and NSAIDs during their last attack in the screening phase and during the subsequent migraine attack in the follow-up phase of the study. Treatment regimens were classified into 6 categories: simultaneous coadministration of triptans and NSAIDs (T&N); triptans first followed by NSAIDs (T_N); NSAIDs first followed by triptans (N_T); triptans only (TRP_only); NSAIDs only (NSAID_only); and others. Headache experience, reasons for treatment regimens, and treatment satisfaction were crosstabulated by treatment regimens. The log-rank test was used for the analysis of time-to-event data. Results. Among 8440 oral triptan users surveyed during the screening phase, 2307 (27%) reported using triptans and NSAIDs combination therapy during their last migraine attack. Of those, 1502 experienced a subsequent migraine attack and completed the follow-up survey; 38% of these 1502 patients who used triptans and NSAIDs cotherapy during their last migraine attack continued to use combination therapy for their next attack. The most common treatment regimen, excluding others (n = 354, 24%), was TRP_only (n = 403, 27%), followed by N_T (n = 345, 23%), NSAID_only (n = 170, 11%), T&N (n = 152, 10%), and T_N (n = 75, 5.0%). More TRP_only patients became nausea-free within 1 h after an initial dosing. TRP_only, T&N, and N_T had significantly shorter median hours of suffering from migraine and limited functioning, as compared with other treatment regimens. Substantially more patients taking TRP_only (34.7%) were very satisfied with their current treatment regimen than other regimens. Conclusions. Migraine patients frequently change their treatment regimens in response to headache profiles. For patients with migraine associated nausea symptom, combination of therapy with triptan and NSAIDs appears to be less effective in relieving nausea than triptan monotherapy. Triptan montherapy remains a common and an effective migraine treatment strategy. Key words: acute migraine, oral triptans, NSAIDs, combination therapy, treatment regimens, treatment patterns (Headache 2008;48:1176-1185) Outcomes Research & Management, Merck & Co., Inc., West Point, PA, USA (D.S. Ng-Mak), Global Qutcomes Research, Global Human Health, Merck & Co., Whitehouse Station, NJ, USA (X.H. Hu, Y.-T. Chen, and L. Ma). Address all correspondence to D. S. Ng-Mak, 770 Sumneytown Pike, MS WP39-166, West Point, PA 19486, USA. Accepted for publication December 10, 2007. INTRODUCTION A large number of acute migraine medications are commercially available for migraineurs Clinical practices suggest that migraine-specific treatments, as compared with nonspecific treatment, provide faster pain relief and pain freedom at 2 h. Within the class Financial support information: This research study was funded by Merck & Co., Inc. 1176

Headache 1177 of migraine-specific treatments, triptans are more efficacious than ergotamines. 1 About 20% of migraine sufferers treat their acute migraine episodes with triptans. 2 Nonsteroidal anti-inflammatory drugs (NSAIDs), a class of nonspecific treatments, are among the most commonly used medications to alleviate headache pain, despite their well-known risk of gastrointestinal side effects. While many treatment options are available to migraine sufferers, the way that migraine sufferers use their migraine medication is not well understood. Despite extensive literature showing treatment efficacy of cotherapy with triptans and NSAIDs for most migraine patients, 3-6 little is known about the likelihood that patients use cotherapy treatment regimen when oral triptans and NSAIDs are available to them. The goals of this study are 3-fold: (1) to examine the usage of triptans and NSAIDs, and its combination for treatment of acute migraine attack; (2) to explore reasons for their choice of treatment regimen; (3) to describe the potential differences of use of various regimens on functionality. METHODS Study Design. A prospective survey was conducted from April to July 2006. Potential study participants were initially identified through a large managed care database, which contains a broad clinically rich and geographically diverse spectrum of claims data from health plans in the southeastern, mid-atlantic, central, and western regions of the United States. A total of 64,165 adult migraine patients (35% with an ICD-9 code for migraine diagnosis and 65% with claims for at least 1 oral triptan) were initially selected as oral triptan users during the last 6 months of year 2005. These patients were invited to participate in the survey through the mail and were informed of telephone eligibility screening interview if they did not opt out. A sample size of 1500 patients was determined for the study as it provides an appropriate level of precision around the proportion estimates of patients with different treatment regimens (the primary objective). Patient enrollment discontinued when the predetermined sample size was reached. Verbal consents were obtained and eligible patients were invited to participate in the follow-up study during the screening phase. Consented patients received patient package which contained a follow-up questionnaire and an instruction guide in the mail. The study was approved by the New England Institutional Review Board. Study Population. To be eligible to participate in the follow-up phase of the study, migraine sufferers must have used both an oral triptan medication* and NSAIDs (prescription or over-the-counter ) during their recent migraine attack. During the screening phase, phone interviewers first asked migraine sufferers Which one of the following oral triptans did you use for your last migraine attack? and then read a list of oral triptan medications. Next, migraine sufferers were asked if they used pain medication (either prescription or over-the-counter). If they did, migraine sufferers provided the name of such pain medication. Patients who used oral triptan or NSAIDs for conditions other than acute migraine treatment were excluded from the study. Between April and July 2006, consented participants were asked to complete a follow-up questionnaire within 7 days but preferably within 24 h after their next migraine attack via one of the following venues: mail, fax, telephone, or web. Participants who completed the follow-up survey received $30 in recognition of their time. Measurements. Treatment Regimens at Screening. In addition to providing names of medications used for acute treatment of migraine attacks, migraine sufferers also reported the sequence in which they took the combination therapy of oral triptan medications and NSAIDs during the screening interview in one of the following responses: (1) took (name of an oral triptan) first, then took another pain medication; (2) took (name of an oral triptan) *Oral triptan medications include Almotriptan, Electriptan, Frovatriptan, Naratriptan, Rizatriptan, Sumatriptan, and Zolmitriptan. Prescription NSAIDs include Aspirin, Bextra, Celebrex, Diclofenac, Diclofenac sodium, Flurbiprofen, Indomethacin, Ketorolac, Mobic, Naproxen, and Naproxen sodium. Over-the-counter NSAIDs include Actron, Advil/Advil Migraine, Aleve, Anacin, Aspirin/Bayer, Bufferin, Ecotrin, Ibuprofen, Midol IB, Motrin, Nuprin, and Rufen.

1178 September 2008 and another pain medication at the same time; (3) took another pain medication first, then took (name of an oral triptan). Treatment Regimens at Follow-Up. At the follow-up survey, migraine sufferers were asked to describe in detail the names and sequences of medication used for their migraine attack. Using the first 2 treatment sequences, migraine sufferers were categorized into 6 treatment regimens, namely (1) simultaneous coadministration of an oral triptan and NSAIDs (T&N); (2) oral triptan first, then NSAID (T_N); (3) NSAID first, then oral tripan (N_T); (4) oral triptan only (TRP_only); (5) NSAID only (NSAID_only); and (6) others (Others). Headache Severity and Nausea Symptoms at First Medication. Migraine sufferers reported headache severity at the time of first medication on a 4-point Likert scale (no headache but suspected one would start soon; mild headache; moderate headache; and severe headache). A binary-coded headache severity was created by categorizing headache severity level as none or mild and moderate or severe. For those experiencing nausea before taking the first medication, migraine sufferers reported time to become nausea free after the first medication, which were then grouped into less than 1 h and 1 or more than 1 h. Reasons for Choosing Current Treatment Regimen. Migraine sufferers checked the following statements (whenever it is applicable) that described the way they treated this migraine episode. Dichotomous response was recorded for each of the following statements: (1) your headache intensity has been severe lately and has been difficult to deal with; (2) your headache normally has a slow start and you took the wait-and-see approach; (3) your headache usually came on very quickly and you needed to take medicine early; and (4) your headache frequently came back after initial relief and you needed to take additional medication to keep the headache from recurring. Also, migraine sufferers quantified the extent to which their treatment decision was affected by the availability of drug supply at home. Patients responded to the question For this migraine attack, how much was your treatment decision affected by the availability of drug supply at home? in a 4-point Likert scale (not at all; a little bit; somehow; and very much). A dichotomized variable indicating their treatment decision was somehow or very much affected by the drug availability at home was created. Medication Satisfaction and Functional Levels. On a 5-point Likert scale (very satisfied; satisfied; neither satisfied nor dissatisfied; dissatisfied; and very dissatisfied), migraine sufferers were asked to report their level of overall satisfaction with the medications they took for this migraine attack. To measure the functional impact of migraine, migraine sufferers reported the total number of hours suffering from migraine for this attack and not being able to perform normal activities such as work or leisure. Statistical Analysis. Frequency distributions of headache symptoms, reasons for treatment options, and overall satisfaction levels were cross-tabulated with 6 types of treatment regimens. Adjusting for the effect of headache severity at dosing, a logistic regression was performed to further examine the relationship between treatment satisfaction (very satisfied vs all others combined) and treatment regimens. Pairwise comparisons between TRP_only and all other treatment regimens were produced by the logistic regression. The log-rank test was used for the analysis of time-to-event data, which included hours suffering from migraine and total hours not functioning. All analyses were performed using SAS V8. 7 Results. Sample Recruitment Statistics. Among all oral triptan users identified in a managed care claims database, the extent to which migraine sufferers used oral triptans and NSAIDs for 3 migraine episodes (in 2005, at screening, and at follow-up) is displayed in the Figure. From the managed care claims data, we identified 64,165 patients who received an ICD-9 code for migraine diagnosis and/or had a claim of an oral triptan during the last 6 months of December 2005. We attempted to contact 13,168 migraine sufferers and recorded their medications used for their last migraine episode. In total, 4728 (36%) of these records were discarded for the following reasons: wrong names on files (63%), not suffering from migraine (22%), opting out of research studies (15%), and refusal of participating in migraine research (0.7%). Among contacted oral triptan users (N = 8440), 2307 (27%) used combination therapy of

Headache 1179 Claims Data Base Extraction (June-Dec 2005) Migraine Patients Identified in Claims Data Base N=64,165 Contacted Migraine Patients N=13,168 Discarded Migraine Patients N=4,728 (36%) Names not on files N=2,957 (63%) Not suffered from migraine N=1,030 (22%) Screening Phase (April-July 10, 2006) Screened Migraine Patients N=8,440 (64%) Opt out of research studies N=708 (15%) Refused to participate in migraine research N=33 (0.7%) Patients took both oral triptan & NSAID for their last migraine attack N=2,307 (27.3%) Patients did not take NSAID for their last migraine attack N=6,133 (72.7%) Follow-up Phase (April-July 31, 2006) Participated in follow-up study N=1,502 (65%) Not experienced a migraine attack N=734 (32%) Dropouts N=71 (3%) Oral triptans only N=403 (26.8%) (95% CI: 24.6-29.0%) NSAIDs first, followed by oral triptan, N=345 (22.9%) (95% CI: 20.8-25.0%) Oral tritpan & NSAIDs co-adminstered N= 152 (10.1%) (95% CI: 8.6-11.6%) Oral triptan first, followed by NSAIDs N=75 (4.9%) (95% CI: 3.8-6.0%) NSAIDs only N=170 (11.3%) (95% CI: 9.7-12.9%) Others * N=354 (23.6%) (95% CI: 21.5-25.7%) No treatment N=3 (0.3%) (95% CI: 0-0.6%) Co-therapy users (37.9%) (95% CI:35.4-40.4%) Figure. Patient recruitment status. *Others category include Excedrin/Excedrin Migraine (45%), Acetaminophen (39%), Darvocet (2%), Hydrocodone (2%), Vicodin (1.6%), Anacin (1.3%), Butalbital/ASA/Caffeine (1.3%), Fiorinal (1%), Butalbital (1%), Oxycodone (0.7%), Fioricet (0.7%), Phenergan (0.7%), Ecotrin (0.7%), Ultracet (0.7%), and all others (2.4%) includes Tylenol #3, Butalbital/ASA/Caffeine/ Codeine, Fioricet with Codeine, Fiorinal with Codeine, Prenilin, Duradrin, Midrin, and Zanaflex. NSAID = nonsteroidal antiinflammatory drug. oral triptans and NSAIDs while 73% (6133 patients) used only triptans, but not NSAIDs for their last migraine attack. Among eligible migraine sufferers (N = 2307) who took combination therapy of oral triptan and NSAIDs for their most recent migraine attack, 1502 (65%) of them participated in the follow-up study; 71 (3%) were dropouts; and 734 (32%) did not experience a migraine attack during the follow-up period. Prevalence of Oral Triptan and NSAID Combination Therapy. Three patients did not use any migraine treatment for their migraine episode during the follow-up period. Oral migraine treatment regimens used for the next migraine attacks were reported in descending order of use: 26.8% (95% CI: 24.6-29.0%) TRP_only; 23.6% (95% CI: 21.5-25.7%) Others; 22.9% (95% CI: 20.8-25.0%) N_T; 11.3% (95% CI: 9.7-12.9%) NSAID_only; 10.1% (95% CI: 8.6% 11.6%) T&N; and 4.9% (95%CI: 3.8-6.0%) T_N (Fig.). About 38% (95% CI: 35.4-40.4%) of migraine sufferers who used triptan and NSAIDs for their most recent migraine attacks used combination therapy of triptan and NSAID for their next attacks.that is, 62% of cotherapy users would switch to other treatment regimens when they experienced another migraine attack. About 1 in 4 migraine sufferers who used combination therapy at screening phase used identical treatment regimen for their next migraine attack (Table 1). Specifically, 29% of patients taking N_T regimen at screening continued to use the same N_T regimen at follow-up. Only 10% of patients taking T_N and 45% of patients taking T&N were consistent with their treatment regimen over 2 migraine attacks. Characteristics of Participants in the Follow-Up Phase. The majority of the participants in the follow-up phase were female (N = 1387; 92.3%) and white (87%), with an average of 44 (SD 10.7) years of

1180 September 2008 Table 1. Consistency of Treatment Regimens Over Time: From Screening to Follow-Up Treatment regimens at screening Treatment regimens at follow-up N_T N = 973 n T_N N = 342 n T&N N = 166 n No report of treatment regimen N = 21 n Total N = 1502 n T&N 45 (4.6) 30 (8.8) 75 (45.2) 2 (9.5) 152 (10.1) T_N 31 (3.2) 36 (10.5) 8 (4.8) 0 75 (4.9) N_T 278 (28.6) 49 (14.3) 17 (10.2) 1 (4.7) 345 (22.6) TRP_only 234 (24.1) 120 (35.1) 43 (25.9) 6 (28.6) 403 (26.8) NSAID_only 127 (13.1) 34 (9.9) 5 (3.0) 4 (19.1) 170 (11.3) Others 256 (26.3) 72 (21.1) 18 (10.8) 8 (38.1) 354 (22.9) No treatment 2 (0.2) 1 (0.3) 0 (0.0) 0 3 (0.2) Twenty-one patients did not report their treatment regimens at screening. Column percentages were shown in the table. Overall, 389 (278 + 36 + 75) out of 1481 (ie, 26%) patients used the same treatment regimen at screening and at follow-up. age. About two-thirds of the participants were part or full-time employed. Fifty-four percent were at least bachelor degree holders. Almost all patients (95%) had health insurance coverage. Twenty-eight percent received care from headache specialists or neurologists (Table 2). Headache Severity and Nausea Symptoms at Time of First Medication. At the time of first medication, significant proportions of patients taking T&N (74%) and TRP_only (72%) experienced moderate/severe headache than other patients (Table 3). Likewise, more patients of T&N (32%) and TRP_only (34%) experienced nausea symptoms before taking the first medication. Among those who experienced nausea symptoms, more TRP_only (42%) patients became nausea-free within 1 h after taking their first medication, as compared with other 5 groups (T&N 29%; Others 19%; NSAID_only 18%; N_T 15%; and T_N 11%). Controlling for the effect of headache severity at dosing, similar finding was observed such that more patients taking TRP_only became nausea-free within 1 h after dosing. About 43% of nausea-patients reported taking NSAIDs. Reasons for Current Treatment Regimens. Compared with N_T (15%) and NSAID_only (15%), a significant proportion of patients who used oral triptans as their first treatment medication (ie, T&N [22%], T_N [21%], and TRP_only [22%]) reported severe headache lately as one of the reasons for their current treatment regimen (Table 4). On the contrary, more patients taking NSAID as their first treatment (N_T [61%], NSAID_only [53%]) reported slow-start headache profile and adopted a wait-and-see approach as one of the reasons for their current treatment regimen. Approximately one-third of T&N, as compared with other 5 treatment regimens (ranging from 20% to 26%), reported that their headache came on very quickly and needed to take medicine early. Across all treatment regimen groups, over 50% of patients reported drug supply at home was a concern for treatment choices. Significantly more proportion of T_N (69%) patients, as compared with 51% T&N, claimed drug availability as one of the reasons for their regimen choice. Satisfaction With and Functional Impacts of Current Treatment Regimens. Significant differences in satisfaction level were observed across 6 treatment regimens (Table 5). Substantially more patients taking TRP_only (34.7%) were very satisfied with their current treatment regimen than other regimens (T&N 23%, N_T 22%, NSAID_only 12%, T_N 9%,

Headache 1181 Table 2. Demographic Characteristics of Migraine Sufferers Demographic characteristics N = 1502 % Female 1387 92.3 Age in years, mean (SD) 44 (10.7) Age at migraine diagnosis, 28.5 (11.9) years, mean (SD) Ethnicity White 1308 87.1 Hispanic 67 4.5 Asian 34 2.3 Black 24 1.6 American Indian 18 1.2 Others 50 3.3 Employment status Full time employed 738 51.1 Part time employed 222 15.4 Unemployed but looking for 27 1.9 work Homemaker 246 17.0 Student 50 3.5 Retired 76 5.3 Unable to work because of 27 1.9 migraine Unable to work because of 58 4.0 health reasons other than migraine Education level Less than 8th grade 2 0.1 Some high school 18 1.2 High school graduate 213 14.2 Some college 455 30.3 College graduate 514 34.2 Postgraduate 299 19.9 Health insurance coverage Yes 1431 95.3 Current physician for migraine Primary care physician 1004 71.2 Headache specialist/ neurologist 407 28.8 Fifty-eight missing data for employment status. Ninety-one missing data for current physician for migraine. and Others 13%). Across all pairwise contrasts between treatment regimens and TRP_only, patients taking TRP_only were statistically significantly very satisfied with their treatment than their counterparts (Table 6). This finding sustained after the adjustment of the effect of headache severity at dosing. About 73% of patients taking T&N, N_T, and TRP_only were satisfied/very satisfied with their medication while approximately 1 in 10 patients were dissatisfied/somewhat dissatisfied with their medication. About 52% to 56% of patients taking T_N, NSAID_only, and Others were satisfied/very satisfied with their treatment while 25% to 33% of them were dissatisfied/somewhat dissatisfied with their treatment. Table 7 displays the median hours suffered from migraine and median hours not functioning by treatment regimens. Patients taking triptans alone experienced the least number of hours suffering from migraine (TRP_only: median 4 h) while patients taking NSAID only reported the longest hours of suffering from migraine (NSAID_only: median 8 h) (P <.0001). Patients of 3 treatment regimens (T&N, TRP_only, and N_T) reported the shortest median hours (2 h) not functioning due to migraine (P =.0002). Comments. This study provided an overview of patterns of treatment regimens for migraine sufferers. The prevalence of combination therapy of both an oral triptan and NSAID for acute migraine was 27% among oral triptan users identified from a large claims database. Among migraine sufferers who used both oral triptans and NSAIDs for their most recent migraine episode, only 38% of them continued to use any type of combination therapy for their next migraine attack. A lesser proportion (26%) of them chose an identical combination therapy over 2 migraine attacks. Oral triptan alone was the most common type of treatment regimens, seconded by step-care approach (NSAID first, then triptan). In this large prospective observational study, migraine sufferers had many options for their migraine treatment. In total, 23% of migraine sufferers chose migraine medications other than triptan and/or NSAIDs. They tailored their treatment therapy according to their individual headache episode. It is imperative to educate patients about medication efficacy as well as potential side effects of migraine medications. Our data showed that a large proportion of migraine sufferers modified their combination therapy to noncombination treatment regimens from one migraine attack to another. Responding to their changing nature of migraine attacks, migraine sufferers appreciate the flexibility of treatment regimens. Potential advantages of a fixed dose combination of

1182 September 2008 Table 3. Headache Severity and Nausea Symptoms T&N N = 152 n T_N N = 75 n N_T N = 345 n TRP_only N = 403 n NSAID_only N = 170 n Others N = 354 n Headache severity at the first medication No or mild 40 (26.3) 29 (38.7) 148 (42.9) 115 (28.5) 61 (35.9) 111 (31.4) Moderate or severe 112 (73.7) 46 (61.3) 197 (57.1) 288 (71.5) 109 (64.1) 243 (68.6) Experienced nausea before taking the first medication No 102 (67.6) 57 (76.0) 275 (79.7) 264 (65.7) 126 (74.1) 251 (70.9) Yes 49 (32.4) 18 (24.0) 70 (20.3) 138 (34.3) 44 (25.9) 103 (29.1) Became nausea-free after taking the first medication Within 1 h 14 (28.6) 2 (11.1) 10 (15.2) 57 (41.6) 8 (18.2) 19 (19.2) More than 1 h 35 (71.4) 16 (88.9) 56 (84.9) 80 (58.4) 36 (81.8) 80 (80.8) Analysis was performed for the sample who reported nausea experience (n = 422). Four patients from the others did not report timing for nausea-free. Among patients reporting nausea, 43% (ie (49 + 18 + 70 + 44)/422) took NSAID. oral triptan and NSAID such as treatment efficacy, compliance, and tolerability await to be proven. 8 While available data have shown that a fixed combination of an oral triptan and an NSAID is more efficacious than the corresponding monotherapy alone, 3-6 it is unknown if the fixed dose combination regimen would be superior to coadministration of an oral triptan and NSAID. Treatment compliance seems to be less of an issue for migraine sufferers who have taken their acute migraine medication. Speed and efficacy of aborting migraine may override any potential inconvenience of taking 2 types of medications. Use of NSAIDs is known to be associated with increased gastrointestinal side effects, including dyspepsia, bleeding, and, in rare cases, perforation. We observed in our data that 43% of the study partici- Table 4. Reasons for Current Treatment Regimens Reasons for current treatment decisions T&N N = 152 T_N N = 75 N_T N = 345 TRP_only N = 403 NSAID_only N = 170 Others N = 354 Your headache intensity has been severe lately and 21.7 21.3 14.5 22.3 15.3 29.1 difficult to deal with Your headache has a slow start and you took a 39.5 40.0 61.2 47.2 52.9 51.4 wait-and-see approach Your headache came on very quickly and you needed to 30.9 25.3 19.7 26.3 20.0 25.1 take medicine early Your headache frequently came back after initial relief 30.3 44.0 32.2 32.8 39.4 33.6 and you needed to take additional medication to keep the headache from recurring Availability of drug supply at home 51.3 69.3 56.8 57.3 58.8 65.3

Headache 1183 Table 5. Overall Satisfaction With Medication Used by Treatment Regimens Satisfaction level T&N N = 151 T_N N = 75 N_T N = 344 TRP_only N = 403 NSAID_only N = 170 Others N = 352 Very satisfied 23.2 9.3 21.8 34.7 12.4 13.1 Satisfied 49.7 42.7 51.2 37.9 34.1 39.2 Neither 17.2 20.0 15.1 15.6 20.6 23.0 Dissatisfied 7.3 21.3 9.9 8.4 27.1 16.8 Very dissatisfied 2.7 6.7 2.0 2.9 5.9 7.9 Seven patients were excluded: 4 did not provide answers on satisfaction question and 3 did not take any medication. pants experienced nausea before taking the first medication took NSAIDs. These patients endured extended duration of nauseated symptoms than patients taking triptans only. Apart from gastrointestinal side effects, NSAIDs may induce hypertension in normotensives. 9 The cost advantage of NSAIDs as migraine treatment must be weighed against its benefits and side effects. Physician patient education should consider extending its focus on general health status of patients. Migraine sufferers made rational decisions on their treatment regimens according to their prior experience of headache profiles. Our data suggest that migraine sufferers were more likely to take oral triptans as their first medication when they experienced severe headache lately. However, when migraine sufferers experienced slow-start headache profile, they would rather adopt a wait-and-see approach by taking NSAIDs as their first medication.these patients have, on average, 16 years migraine history. Over the years of experiencing migraine, they have learned to observe their headache symptoms profile and experiment treatment regimens. Headache profiling may be an important indication for taking certain types of treatment regimen. However, patients with slow-start and slow-progression migraines are the least to benefit from treatment because of the loss of opportunity to treat a migraine with an oral triptan at an earlier phase. Table 6. Treatment Regimens and Satisfaction Levels Pairwise Contrasts Unadjusted Adjusted Contrast OR (95% CI) P value OR (95% CI) P value TRP_only vs N_T 1.91 (1.38-2.65).0001 1.95 (1.40-2.72) <.0001 TRP_only vs T&N 1.76 (1.15-2.71).009 1.76 (1.15-2.71).01 TRP_only vs T_N 5.17 (2.31-11.56) <.0001 5.26 (2.35-11.76) <.0001 TRP_only vs NSAID_only 3.78 (2.29-6.23) <.0001 3.82 (2.52-6.31) <.0001 TRP_only vs Others 3.54 (2.44-5.13) <.0001 3.56 (2.46-5.17) <.0001 Adjusted variable was headache severity at the initial medication. Logistic regression was performed to examine the relationship between satisfaction level (very satisfied vs others) and treatment regimens.

1184 September 2008 Table 7. Functional Impacts by Treatment Regimens Functional Impacts T&N N = 151 T_N N = 75 N_T N = 344 TRP_only N = 403 NSAID_only N = 170 Others N = 352 Log-rank test, df = 5 P value* Total hours suffered from migraine Median hours (95% CI) Total hours not functioning Median hours (95% CI) 4.5 7.0 5.0 4.0 8.0 6.0 50.3 <.0001 (4.0-6.0) (5.0-10.0) (5.0-6.0) (4.0-5.0) (7.0-10.0) (6.0-6.5) 2.0 4.0 2.0 2.0 3.0 3.5 24.6.0002 (2.0-3.0) (2.0-5.0) ( - ) (2.0-3.0) (3.0-4.0) (3.0-4.0) *P values were generated from the log-rank test. Migraine patients can indeed identify the coming of migraine at an early phase of a migraine attack. 10 With proper education, patients can be taught to recognize symptoms for a migraine attack and to optimize treatment efficacy. The study has several limitations. Our results can only be generalizable to migraine patients who had experienced with oral triptans and NSAIDs. The prevalence of combination use of oral triptans and NSAIDs among this highly selected patient sample was comparatively higher (38%) than the prevalence of cotherapy among oral triptan users (27%). Female, in general, is more represented in research.the majority of our sample was female and more matured in age, any inference regarding findings should take gender and age into consideration. As an observational study, we did not interfere what and how patients took migraine treatment. For some unknown and unmeasured reasons, patients self-selected themselves into various treatment regimens. Inferences about outcome measures among these treatment regimens should be exercised with great caution. Following our primary objective of the study, we only described the prevalence and patterns of migraine treatment regimens. Strengths of the study are worth mentioning. This study involves a large sample of patients across a broad range of geographical regions in the United States to calibrate migraine treatment sequences. To the best of our knowledge, this is the first large prospective study describing the use of triptans and NSAIDs when migraine sufferers had experience with using both medications. Malik et al 11 surveyed 109 migraine patients regarding the sequence and timing of taking migraine medication during year 2001. Similar to our finding, Malik et al reported triptan usage was the most common type of migraine medication. The study by Malik, however, did not calibrate the sequence of migraine medication and therefore did not report the concomitant usage of triptans and NSAIDs. Prospective nature of capturing patients migraine experience and treatment regimens avoids unnecessary memory lapses of events. In conclusion, oral triptans continue to be the most commonly used acute migraine treatment regimen. Many migraine sufferers are very satisfied with oral triptans. Oral triptan and NSAID combination therapy is considered to be a sideline treatment regimen. Most of triptan and NSAID combination users chose not to use combination therapy for their next attack. Being responsive and proactive about their migraine symptoms and profiles, migraine sufferers would value the flexibility of when and how to use oral triptans and NSAIDs. Use of NSAIDs alone or in combination with oral triptans should be exercised with caution given their side-effect profiles. REFERENCES 1. Lipton RB, Bigal ME, Goadsby PJ. Double-blind clinical trials of oral triptans vs other classes of

Headache 1185 acute migraine medication a review. Cephalalgia. 2004;24:321-332. 2. Lipton RB, Scher AI, Kolodner K, Liberman J, Steiner TJ, Stewart WF. Migraine in the United States: Epidemiology and patterns of health care use. Neurology. 2002;58:885-894. 3. Krymchantowski AV. Naproxen sodium decreases migraine recurrence when administered with sumatriptan. Arq Neuropsiquiatr. 2000;58:428-430. 4. Krymchantowski AV, Barbosa JS. Rizatriptan combined with rofecoxib vs. rizatriptan for the acute treatment of migraine: An open lable pilot study. Cephalalgia. 2002;22:309-312. 5. Krymchantowski AV, Bigal ME. Rizatriptan versus rizatriptan plub rofecoxib versus rizatriptan plus tolfenamic acid in the acute treatment of migraine. BMC Neurology 2004;4:10. [www document]. Available at: http://www.biomedcentral.com/1471-2377/ 4/10 (accessed 26 June 2007). 6. Smith TR, Sunshine A, Stark SR, Littlefield DE, Spruill SE, Alexander WJ. Sumatriptan and naproxen sodium for the acute treatment of migraine. Headache. 2005;45:983-991. 7. SAS Institute, Inc. SAS/STAT User s Guide, Version 8. Cary, NC: SAS Institute, Inc; 1999. 8. Lober E. Fixed drug combinations for acute treatment of migraine. CNS Drugs. 2005;19:769-784. 9. Stöllberger C, Finsterer J. Nonsteroidal antiinflammatory drugs in patients with cardio- or cerebrovascular disorders. Zeitschrift für Kardiologie. 2003;92:721-729. 10. Ng-Mak DS, Cady R, Chen YT, Ma L, Bell CF, Hu XH. Can migraineurs accurately identify their headaches as migraine at attack onset? Headache. 2007;47:645-653. 11. Malik SN, Hopkins M, Young WB, Silberstein SD. Acute migraine treatment: Patterns of use and satisfaction in a clinical population. Headache. 2006; 46:773-780.