Survey of Pain Medicine Specialists' Familiarity with Migraine Management

Abstract

Objective

Pain specialists treat patients with headache and interface with those who use opioids more so than neurologists and headache specialists. We assessed the headache medicine knowledge and needs of pain specialists.

Design/Setting

Cross-sectional online survey.

Subjects

Members of the American Academy of Pain Medicine.

Methods

Survey was based on a prior survey of primary care providers’ knowledge and needs and was iteratively updated by four headache specialists, two with pain medicine affiliations.

Results

Of the 105 respondents, 71.4% were physicians, 34.3% were women, and they averaged 20.0 ± 13.6 years in practice. The most common specialty was anesthesia (36.1%, n = 35/97) followed by neurology (14.4%, n = 14/97). About half of providers (55.7%, n = 34/61 and 53.3%, n = 32/60) were familiar with the American Academy of Neurology Guidelines for pharmacological migraine prevention and the Choosing Wisely Campaign recommendations for limiting neuroimaging and opioids. Less than half of all providers (39.7%, n = 23/58) were familiar with the American Headache Society guidelines for emergency management of migraine. Providers were aware of Level A evidence-based nonpharmacological therapies, with over three-fourths recognizing cognitive behavioral therapy (80.7%, n = 50/62) and biofeedback (75.8%, n = 47/62) as evidence-based interventions. About 80% of providers (n = 50/64) estimate making migraine diagnoses in ≤50% of their patients with headache. Providers consider starting preventive headache therapy at 7.1 ± 3.9 days/month and report referring 34.3%±34.2% of patients to behavioral interventions.

Conclusions

Dissemination and implementation of headache guidelines is needed for pain medicine specialists. Providers may need help diagnosing migraine based on currently accepted guidelines and referring for evidence-based behavioral therapies.

Introduction

Migraine affects over 40 million Americans and is the second most disabling condition per the World Health Organization (WHO) [1,2]. In recent years, major developments have been made in expanding available migraine therapies. In addition to well-established, evidence-based abortive and preventive treatments such as NSAIDs, triptans, antiepileptics, beta-blockers, antidepressants, and onabotulinum toxin A [3], promising new therapies targeting CGRP and 5HT_1F receptors have come onto the market in the past few years, along with several emerging pharmacological treatments and nerve stimulation devices [4, 5].

Unfortunately, patients with migraine have been historically underdiagnosed and under-treated [6]. Only 25% of patients who consulted a healthcare professional for headache received an accurate migraine diagnosis [6] and data from the American Migraine Prevalence and Prevention Study showed that although 38.8% of patients with migraine would qualify for preventative treatment, only 12.4% were actually prescribed preventive therapy [5,7]. More recent data from the OVERCOME trial showed that less than 20% of patients with at least four migraine days per month and moderate disability sought care, were diagnosed, and successfully treated for migraine, and the likelihood of undergoing these steps was correlated with health insurance coverage and level of disability, while those who sought care in emergency, retail or urgent care settings were less likely to be initiated on preventative treatment [8]. Unsurprisingly, patients with migraine oftentimes report dissatisfaction with their care [9–11]. Appropriate recognition, treatment and patient/provider education are necessary to reduce the burden of migraine and other headache disorders and improve patient satisfaction using evidence-based therapies [12].

With the shortage of providers trained in headache medicine in the United States, migraine is often managed by primary care providers and pain medicine providers [13–17]. However, numerous studies have shown that there is limited headache knowledge in various medical specialties [18–22], and many patients end up in emergency department settings due to poor outpatient care [8]; 3% of all emergency department visits are for a chief complaint of headache and most of these visits are for primary headache disorders [23]. Prior research has examined knowledge and needs assessments of primary care providers [18, 19] and women’s health providers [24]. However, to our knowledge, no knowledge and needs assessment has been done for pain medicine specialists. This is important and timely, especially with results of a recently published article on chronic opioid use in the United States which reported that pain specialists are much more likely to interface with patients with headache who use opioids (75% of patients) compared to neurologists and headache specialists (32.3% and 18.5% of patients were opioid users, respectively) [14]. As a physician diagnosis of migraine is associated with decreased opioid use and increased migraine-specific treatment [14], it is critical to close knowledge gaps for non-headache specialists to better characterize, diagnose, and manage migraines according to guidelines.

In this study we surveyed providers from the American Academy of Pain Medicine to better understand how headache was incorporated into their medical education, how providers received current information on headache, and how patients with headache were managed in practice.

Methods

Survey

A pain medicine knowledge and needs survey was developed based off of the original survey of primary care providers [18] and iteratively refined using pain medicine specialists and headache medicine specialists’ feedback. An initial team of four physicians (two with pain medicine division affiliations and active members of Pain societies and two with headache medicine division affiliations and active members of Headache societies) conducted the initial iterations. Following this, the survey was reviewed by the Chair of the Headache Special Interest Group (SIG) of the American Academy of Pain Medicine (AAPM) and then the AAPM survey review committee. It was also open to review by the Editor-in-Chief of the journal Pain Medicine. The final survey, consisting of questions about providers’ knowledge, including awareness of the American Academy of Neurology guidelines [3], American Headache Society guidelines [25], and Choosing Wisely recommendations [26], and providers’ practice of headache medicine, was then distributed via email as an anonymous survey by the AAPM, as a joint sponsorship by the Headache Special Interest Group of the AAPM and the journal Pain Medicine. The survey was online 7/17/20–8/21/20. This was initially a society survey for which IRB approval was not obtained; post survey analysis we sought IRB approval for the study and while the Northwell IRB does not approve studies post hoc, the IRB reviewed the study and concluded that the study would have been deemed exempt if it had been submitted a priori.

Study Sample

All members of the AAPM received invitations via email to participate. There were two additional reminders sent over email by AAPM. In addition, the Headache SIG members received additional reminders to complete the survey. There were also social media posts to remind people to complete the survey, though no links were publicly distributed via social media in order to capture the response rate.

Statistics

Data for the study were entered into Qualtrics and analyzed in Excel version 16.41. Data were summarized using descriptive statistics such as means and standard deviations (SD), and frequency counts with percentages. Survey responses with missing data were included.

Results

A total of 2,848 emails were sent out, and there were 635 total opens with the email address de-duplicated (meaning if the user got both emails and opened one or both, they were only counted once).

Survey Participants

We received 105 survey responses eligible for the analysis. Of these respondents, 36 (34.3%) were women and the majority (75; 71.4%) were physicians (MD or DO). Among 97 responding physicians, the most common specialty was anesthesia (37, 36.1%) followed by neurology (14.4%, n = 14/97). Seventy-two respondents reported an average of 19.98 ± 13.62 mean years in practice, ranging from 0 to 55 years. Table 1 summarizes the breakdown of provider type, medical experience, specialty, and membership in professional societies.

Table 1.

Characteristics of comprehensive survey respondents (N = 105)

Background	n/Total responses	%
Women	36/105	(34.3%)
Provider type
MD/DO	75/105	(71.4%)
Attending	69/74*	(93.2%)
Trainee	5/74*	(6.8%)
NP	12/105	(11.4%)
PA	5/105	(4.8%)
Psychologist	9/105	(8.6%)
Other	4/105	(3.8%)
Medical experience (attendings)	Total responses
Mean years in practice; range	72	19.98 years (SD ± 13.62); 0–55 years
Primary specialty	n/Total responses	%
Anesthesia	35/97	(36.0%)
Neurology	14/97	(14.4%)
Physical medicine and Rehab	11/97	(11.3%)
Internal medicine	4/97	(4.1%)
Family medicine	7/97	(7.2%)
Psychiatry	5/97	(5.2%)
Other	21/97	(21.7%)
Pain management	9/97	(9.3%)
Psychology	8/97	(8.3%)
Member of a pain society	58/62	(93.6%)
Member of a headache society	11/62	(17.7%)

^*One respondent in the MD/DO group did not disclose level of training.

PMR = Physical Medicine and Rehabilitation.

Knowledge

Pain medicine providers estimated that 20.7% ± 14.7% of the overall US population has migraine, within the range of 16.6–22.7% national prevalence reported in current literature. Three-fourths of respondents (75.0%, n = 51/68) were aware that barbiturates contributed to medication-overuse headache (MOH), and a majority were also aware that narcotics/opioids (63.2%, n = 43/68) and acetaminophen/aspirin/caffeine combination pills (54.4%, n = 37/68) were causes of MOH. About half of responses (51.7%, n = 35/68) additionally identified NSAIDs as a top culprit. Just over half (55.7%, n = 34/61 and 53.3%, n = 32/60) of respondents answered “yes” when asked if they were familiar with the American Academy of Neurology (AAN) Guidelines for pharmacological treatment for migraine prevention [3] and with the Choosing Wisely recommendations for limiting neuroimaging and opioid prescriptions [26]. Less than half (39.7%, n = 23/58) were familiar with the American Headache Society (AHS) guidelines for the management of migraine in an emergency setting [25]. Providers were aware of the Level A evidence-based nonpharmacological therapies for migraine [27], with over three-fourths recognizing cognitive behavioral therapy (80.6\7%, n = 50/62) and biofeedback (75.8%, n = 47/62) as evidence-based interventions. The survey results are summarized in Table 2 and Figure 1. We stratified provider specialty to examine whether that might affect knowledge around certain headache guidelines (Supplementary Data E-Table). In order to account for residency in their primary specialty, the Supplementary Data E-Table includes only physician responses and excludes responses from nurse practitioners, physician assistants, psychologists, a pharmacologist and a dentist.

Table 2.

Survey results: Headache knowledge

	Total Responses		Mean ± SD	Correct Response
On average, what percent of the overall US population do you think has migraine?	75		20.7% (14.7%)	16–22%*
		n/Total responses		%
Which are the top three culprit medications that contribute to rebound/medication overuse headache?
Butalbital containing medications (Fioricet, Fiorinal)		51/68		(75.0%)
Narcotics		43/68		(63.2%)
Acetaminophen/aspirin/caffeine		37/68		(54.4%)
NSAIDs		35/68		(51.5%)
Triptans		19/68		(27.9%)
Tylenol		15/68		(22.1%)
Are you familiar with the American Academy of Neurology Guidelines for prescribing preventative medications for migraine?		34/61		(55.7%)
Physicians (MD/DO)		23/43		(53.5%)
Are you familiar with the Choosing Wisely Campaign recommendations to limit opioid use and imaging in people with migraine?		32/60		(53.3%)
Physicians (MD/DO)		24/43		(55.8%)
Are you familiar with the American Headache Society Guidelines for the management of migraine in the ED?		23/58		(39.7%)
Physicians (MD/DO)		20/41		(48.8%)

^*Smitherman 2013.

Figure 1.

Survey results: Level A evidence-based nonpharmacological treatments. Y-axis displays the number of respondents who selected that choice. Percentages are displayed above the bar.

Education

Medical School

The majority of respondents (63.8%, n = 37/58) estimated that 1–5 hours was spent in medical school learning about headache diagnosis and management. A subset of respondents (n = 36) estimated on a subsequent question that an average of 4.70 ± 6.45 hours were spent on didactic instruction and 4.25 ± 6.03 hours were spent caring for patients with headache during rotations.

Residency

In total, 40% of respondents (n = 22/55) estimated that 1–5 hours were spent learning to manage headache in residency, with the majority of respondents in this category identifying as anesthesiologists (63.6%, n = 14/22). A quarter of respondents (25.5%, n = 14/55) indicated that over 26 hours were spent on headache management, with neurologists comprising the majority of responses in this category (57.1%, n = 8/14). Responses to a subsequent question (n = 33) indicated more time was dedicated to clinical supervision of patients with headache (24.36 ± 37.64 hours) than to didactic instruction (6.23 ± 7.85 hours). Further work could elucidate in which specific setting (i.e., inpatient or outpatient) that clinical experience occurred.

Fellowship

Over half (52.4%, n = 22/42) of respondents indicated 26+ hours were spent on headache training during fellowship. (The vast majority of these respondents (90.5%, n = 38/42) had completed ACGME-accredited pain medicine fellowships.) Nearly 60% of the respondents in this category of were anesthesiologists (n = 13/22), and around one-third were neurologists (n = 7/22), along with one family medicine provider and one PMR provider. Fellowships involved some didactic time (12.84 ± 31.30 hours) and more clinical care time with patients with headache (121.86 ± 348.17 hours). Large variations were reported across programs, most likely due to the small minority of respondents who completed headache-specific fellowships.

Clinical Practice

In continued clinical practice, half of the respondents estimate between 6–10 (30.5%, n = 18/59) and 11–25 hours of headache training (23.7%, n = 14/59), and around 40% (n = 23/59) estimated that they spent over 26 hours on headache training. These findings are summarized in Tables 3 and 4.

Table 3.

Survey results: Headache education during medical training

If MD, about how many hours of training were dedicated to learning about headache diagnosis and management when in:	1–5	N	6–10	N	11–25	N	26+	N	Total
Medical school	63.8%	37	27.59%	16	3.5%	2	5.2%	3	58
Residency	40.0%	22	23.64%	13	10.9%	6	25.5%	14	55
Fellowship	11.9%	5	26.19%	11	9.5%	4	52.4%	22	42
Clinical practice	6.8%	4	30.51%	18	23.7%	14	39.0%	23	59

Table 4.

Survey results: Didactic and clinical exposure to headache during medical training

	Hours (Mean ± SD)	N
If hours were dedicated in medical school,
How many hours were spent on didactics?	4.7 (6.45)	36
How many hours were spent helping to care for patients on rotations?	4.25 (6.03)	34
If hours were dedicated in residency,
How many hours were spent on didactics?	6.23 (7.85)	32
How much time was spent helping to care for patients with headache with clinical supervision?	24.36 (37.64)	33
If hours were dedicated during fellowship,
How much time was spent on didactics?	12.84 (31.30)	25*
How much time was spent helping to care for patients with headache with clinical supervision?	121.86 (348.17)	25*
If hours were dedicated during clinical practice, please indicate how you received this training:
Conference time	30.93 (70.65)	38
Shadowing other faculty	19.10 (74.25)	30

If a closed range was given, the average of the upper and lower limit was used. If an open-ended range was given (e.g., <10, 5+), the benchmark number was used. Responses were excluded if they gave nonquantifiable answers, e.g., “Can't remember,” “N/A,” “whole year.”

^*1 outlying response was excluded from these calculations due to irregular answers.

Information Sources

The majority of headache providers surveyed identified pain meetings (81.3%, n = 61/75), medical journals (81.3%, n = 61/75) and online summary resources such as UptoDate and Medscape (56%, n = 42/75) as their primary sources of information on headache. Thirty-two percent (n = 24/75) reported that they used headache meetings as a source of information. See Figure 2 for details.

Figure 2.

Headache information sources used by providers. Percentage of question responses displayed on the Y-axis.

Practice

Pain medicine providers reported that patients with headache made up approximately 20.2% ± 23.9% of their practice (n = 66). Providers estimated prescribing opioids and barbiturates to few patients with migraine (2.9%±6.4% and 3.8%± 10.4% respectively). Almost 80% of providers (n = 50/64) estimate that they diagnose migraine in up to half of their patients with headache. Providers would consider starting preventive headache therapy at 7.09 ± 3.94 days a month. Most providers (85.3%, n = 58/68) have made the diagnosis of MOH in their career.

The most common comorbidities that providers discuss with patients with headache included sleep conditions (67.6%, n = 46/68), depression (67.6%, n = 46/68), anxiety (61.8%, n = 42/68), and overlapping pain conditions (60.3%, n = 41/68). Sixty-two percent of providers (n = 42/68) would recommend prescribing triptans as abortive therapy and 51.5% (n = 35/68) would recommend NSAIDs. No providers said that they would prescribe opioids as first-line abortive therapy.

On average, providers referred over one-third of patients (34.3%±34.18%) to behavioral interventions. The most notable obstacles to recommending nonpharmacologic therapy were lack of insurance coverage (79.4%, n = 50/63), patient skepticism (73.0%, n = 46/63), patient willingness to engage in behavioral interventions (71.4%, n = 45/63), and lack of available providers trained in these therapies (66.7%, n = 42/63). Three-fourths of providers (75.8%, n = 47/62) ask patients to keep a headache diary, and close to half (47.5%, n = 29/61) coordinate management of headache with the patient’s primary care provider. This information is presented in Table 5.

Table 5.

Survey answers: Medical practice

Comprehensive survey questions	Total Responses	Mean ± SD*
% of patients in a provider's practice who are seen for headache	66	20.23 (23.90)
% of patients with migraine that are prescribed opioids for attacks	61	2.85 (6.37)
% of patients with migraine prescribed butalbital-containing medications (Fioricet, Fiorinal) for migraine	62	3.78 (10.40)
% of patients providers refer to psychological intervention	59	34.25 (SD 34.18)
Number of headache days a month for which provider would consider starting patient on preventive therapy (Mean ± SD and range)†	54	7.09 (3.94); 0–20
	N/Total responses	% responding yes
Have you ever made the diagnosis of medication overuse (rebound) headache?	58/68	(85.29%)
Which comorbidities do you discuss with your patients with headaches?
History of concussion	36/68	(52.9%)
Anxiety	42/68	(61.8%)
Sleep	46/68	(67.6%)
Depression	46/68	(67.6%)
Substance use disorder	38/68	(55.9%)
Obesity	19/68	(27.9%)
Overlapping pain conditions	41/68	(60.3%)
All of the above	35/68	(51.5%)
Other: (Neck pathology, chronic pain conditions, CTD, POTS, inflammatory  conditions, TMD, OSA, occipital neuralgia)	3/68	(4.4%)
What medications would you recommend/prescribe as first line for your patients to stop an acute migraine? Check all that are applicable.
Triptans	42/68	(61.8%)
NSAIDs	35/68	(51.5%)
Acetaminophen	22/68	(32.4%)
Caffeine	14/68	(20.6%)
Magnesium	11/68	(16.2%)
Gepants	7/68	(10.3%)
Antihistamines	4/68	(5.9%)
Butalbital	3/68	(4.4%)
Opioids	0/68	(0.0%)
Other (ditans, lidocaine nasal spray, ketamine nasal spray; or does not prescribe meds)	6/68	(8.8%)
When you are evaluating a patient with a complaint of headache, in approximately what percentage do you make the diagnosis of migraine?
0–25%	27/64	(42.2%)
26–50%	23/64	(35.9%)
51–76%	8/64	(12.5%)
76–100%	6/64	(9.4%)
Obstacles to recommending nonpharmacologic therapy
Insurance coverage does not cover the non-pharmacologic therapy	50/63	(79.4%)
Patient skepticism about efficacy of behavioral approaches	46/63	(73.0%)
Patient willingness to engage in behavioral approaches	45/63	(71.4%)
Lack of availability of providers trained in this treatment	42/63	(66.7%)
Patients state that they do not have time	33/63	(52.4%)
Do not know to whom to refer the patients for non-pharmacologic treatment	17/63	(27.0%)
Other	2/63	(3.2%)
Do you ask patients to keep a headache diary?	47/62	(75.8%)
Do you reach out to a patient's PCP to coordinate migraine management if they have one?	29/61	(47.5%)

^*Responses were excluded in this question if they gave nonquantifiable answers, e.g., “refer these out.”

^†If a closed range was given, the average of the upper and lower limit was used. If an open-ended range was given (e.g., <10, 5+), the benchmark number was used.

The top medications prescribed for migraine prevention by this sample of providers is listed in Table 6. The most common treatments were topiramate (76.2%, n = 48/63), tricyclic antidepressants (63.5%, n = 40/63), beta blockers (49.2%, n = 31/63), and anti-calcitonin gene-related peptide targeted therapies (39.7%, n = 25/63). Few (4.8%; 3/63) listed medical cannabis as one of their preferred options for migraine prevention. Figure 3 illustrates that most providers would try abortive or preventive therapy first before referring to other specialists.

Table 6.

Survey responses: Practices in pharmacological migraine prophylaxis

What are the top 5 medications you prescribe for migraine prevention?	N = 63
Topiramate	48 (76.2%)
Tricyclic antidepressants (amitriptyline, nortriptyline, desipramine)	40 (63.5%)
Beta blockers (Propranolol, timolol, metoprolol)	31 (49.2%)
Calcitonin gene-related peptide antibodies (Aimovig, Ajovy, Emgality)	25 (39.7%)
Magnesium	18 (28.6%)
Onabotulinum toxin	18 (28.6%)
Gabapentin	14 (22.2%)
SNRIs (Venlafaxine, duloxetine)	13 (20.6%)
Valproic acid	9 (14.3%)
Other	8 (12.7%)
Verapamil	6 (9.5%)
Vitamin B*	4 (6.4%)
Does not prescribe	4 (6.4%)
Medical cannabis	3 (4.8%)
Coenzyme Q	2 (3.2%)
ARBs (Vandesartan, Candesartan)	1 (1.6%)

Values are n and percentages. Petasides hybridis, zonisamide, lisinopril, and imipramine were included in the survey but was not selected by any respondents. One survey response to this question was excluded due to excessive answer selections.

^*Type of Vit B was not specified.

Figure 3.

Point in care when providers refer to headache specialists.

Improved Headache Care

When asked about the needs that pain medicine providers would prioritize to improve headache care, the majority identified patient advocacy for behavioral treatment coverage (61.3%, n = 38/62), documentation of best headache practices (53.2%, n = 33/62), clearer statements on effectiveness and side effects of available treatments (53.2%, n = 33/62) and guidelines for patients resistant to clinical care (53.2%, n = 33/62) as areas of additional support. This information is displayed in Table 7. Figure 4 illustrates a summary paraphrasing open-ended feedback about improving headache care that respondents left at the end of the survey.

Table 7.

Survey answers: Improvement areas in headache care

What do you think are the top needs for pain medicine providers to improve headache care?	N = 62
Patient advocacy for behavioral treatment coverage	38 (61.3%)
Documentation of “Best Practices” for prevention and treatment interventions	33 (53.2%)
Clearer statements regarding effectiveness and side effects of pharmacologic and non-pharmacologic treatments	33 (53.2%)
Guidelines for treatment of non-cooperative or clinically resistant patients	33 (53.2%)
Patient advocacy for pain medicine headache procedure reimbursement	28 (45.2%)
Patient advocacy for medication coverage	23 (37.1%)
Patient advocacy for device coverage	19 (30.7%)
Other	2 (3.2%)

Figure 4.

Survey answers: Open-ended comments and suggestions.

Discussion

Our study evaluated patterns and tendencies of pain medicine specialists in terms of medical knowledge and medical practice. In our sample of providers, we found: 1. Pain medicine providers are aware that migraine is prevalent, that MOH exists and can identify the most common culprit medications. 2. Pain providers were likely to utilize pain meetings and pain journals for up-to-date information on headache. 3. Headache exposure during training and familiarity with guideline-recommended migraine management varied based on the provider’s primary specialty. 4. Approximately three out of five providers will include triptans in first line abortive therapy, and most providers do not view opioids or barbiturates as first line agents. 5. Providers are appropriately identifying evidence-based non-pharmacological therapies such as cognitive behavioral therapy and biofeedback but find that one of the greatest needs is more patient advocacy for insurance coverage of behavioral treatment.

Our survey showed that overall, pain medicine providers responding to the survey had a good grasp of migraine prevalence and migraine management with evidence-based treatments. It also confirmed that they provide care for a large number of patients, often with multiple comorbidities and a high level of complexity. The pain medicine providers in this survey cited pain-specific meetings and journals, each chosen by 82% of respondents, as a primary source of continuing headache education, followed by online summary resources such as UpToDate and Medscape. In the open-ended portion of the survey, respondents also noted that continued headache lectures at pain meetings would be helpful. The pain medicine providers were also much less likely to use headache meetings as a source of information compared to pain meetings, so it would therefore be important to include adequate and headache courses at pain society meetings or find novel methods to distribute information from headache meetings to pain medicine providers.

Clearer standardization and distribution of guidelines on migraine management was also a need highlighted in the survey and open-ended responses. Our survey showed that pain medicine providers were somewhat familiar with the current AAN and Choosing Wisely guideline recommendations on preventative treatment and opioid use in migraine, but less familiar with the guidelines for emergency management of migraine, although it varied based on provider specialty. For example, neurologists tended to be far more familiar with the AAN guidelines for pharmacologic prevention of migraine compared to the other specialties, in which closer to one-third or one-fourth were aware of preventative treatment guidelines. It is also important to note that headache teaching experience among MDs during residency and fellowship varied widely based on provider specialty and training program, with most neurologists gaining significant headache experience in residency (26+ hours), while other non-neurology trainees may have received only a few hours of headache experience. In fellowship as well, a subset of anesthesiologists received 26+ hours of headache management while other anesthesiologists received 1–5 hours of headache exposure. Given the variability in formal headache training, our data indicate there is a need for increased awareness of consensus guidelines and more provider education on migraine management for pain medicine providers.

Pain medicine providers were most likely to select triptans (61.8% of respondents selected) as a first-line abortive medication, followed by NSAIDs (51.5% of respondents selected). However, our survey findings suggest that 40%–50% of respondents were unaware that triptans and NSAIDs should be used as first-line treatments in acute migraine, highlighting the need for clear, accessible practice bulletins or guidelines that non-headache specialists can reference. Another reassuring finding is that none of the pain medicine providers in this survey selected opioids as a possible first-line agent for acute migraine, only prescribing opioids and butalbital for an average of <4% of their patients overall. Compared to prior studies that showed opioid and barbiturate use in as many as 18–53% of patients with migraine [28, 29] and that pain specialists are likely to see a higher proportion of people with migraine treated with opioids [14, 30], the respondents in our survey were relatively conservative in their use of opioids and butalbital-containing medications for migraine. Prior studies may have been conducted with pain specialists who see patients with headaches or migraine and other pain conditions for which opioids might need to be used, for example, post-operative or cancer pain. Other possibilities are that they aware of increasing concern for opioid and barbiturate abuse potential in recent years, knowledge of the Choosing Wisely recommendations against opioid and butalbital use as first-line agents for headache, and awareness of medication overuse headache; over half of the respondents were aware that butalbital and narcotic medications were top contributors in MOH. These changes may also be attributed to an increase in availability and advertisement of other abortive treatments. Such influencing factors may be exaggerated in our survey sample, as members of the AAPM Headache Special Interest Group received additional reminders to complete the survey and these providers may have increased awareness of headache treatment and more self-report bias.

Prior studies on migraine have not gauged provider knowledge and attitudes toward newer therapies such as calcitonin gene related peptide (CGRP) targeted monoclonal antibodies, and few studies have examined provider attitudes toward evidence-based non-pharmacologic interventions. A survey among primary care providers showed that over one-third to one-half of PCPs were unable to identify biofeedback, cognitive behavioral therapy, and relaxation therapy as evidence-based interventions [19]. This survey showed that more pain medicine providers were able to identify cognitive behavioral therapy and biofeedback, which are also used to treat other pain conditions, as evidence-based interventions. Relaxation therapy was still overlooked as an intervention. Unfortunately, lack of insurance coverage of behavioral treatments is cited by pain medicine providers as the most common barrier to recommending nonpharmacological therapy, followed by patient receptiveness of behavioral approaches. Patient advocacy for insurance coverage and patient education resources were also highlighted by pain medicine providers as top needs in improving headache care.

Limitations

Recognizing the high prevalence of migraine and that migraine is the number one type of headache for which patients present for care, the knowledge assessment was more focused on migraine and medication overuse headache. However, the field of headache medicine is very broad, with hundreds of different headache types, and within one survey with a limited number of questions, we were unable to assess for other also very disabling headache types, including tension-type headache and cervicogenic headache. Further studies may look at the other categories of primary and secondary headaches elucidated within the International Headache Classification to further understand the breadth of knowledge with less common conditions.

Another limitation of the study is that we did not examine familiarity with the latest recommendations for use of novel preventive agents [31].

Our study was limited by a small proportion of respondents to the online survey (105 responses/2,848 emails sent out), recall bias, self-report bias, and self-selection bias among members of the American Academy of Pain Medicine, which has a total of 1,472 members at the time of manuscript submission. Our sample size was too small to conduct subgroup analyses based on level of training. In addition, neurologists, who make up <5% of the current AAPM body, were overrepresented in 14% of our survey responses. Likewise, anesthesiologists, who make up almost half of the AAPM, and physical medicine and rehabilitation (PMR) providers, who comprise almost a quarter of the AAPM, were underrepresented in our survey responses at 36% and 11%, respectively. However, this distribution of primary specialties may be more representative of the distribution of pain specialists who treat more patients with headache and the information captured in this survey is still valuable. There are also several other prominent pain societies in addition to the AAPM with differences in membership, and this limits our ability to generalize these findings across all pain medicine providers. However, this study was able to capture in detail several parameters of pain medicine provider training, knowledge, practice and attitudes and elicited some of the priorities of pain medicine providers with regards to improving headache care.

The amount of missing data may overestimate medical knowledge on migraine and underestimate first-line opioid and barbiturate use, as percentages are largely reported out of the number of question-specific responses rather than the total survey responses. However, the data acquired are still useful for the sake of a descriptive study on pain medicine providers’ attitudes toward migraine treatment. With some consensus shown through both the close- and open-ended responses, our survey is still able to provide a valuable glimpse into the knowledge and needs of pain specialists that can direct future planning for provider education and support.

Conclusion and Perspectives

Pain medicine providers come from different primary specialties and have varying formal training in headache management. In our sample of pain medicine providers, although pain providers are aware of migraine prevalence, mostly knowledgeable about current treatments, and appropriately avoid opioid and butalbital medications as first-line treatments in migraine, distribution of clear consensus guidelines or practice bulletins was identified as a need for better provider and patient education. Pain medicine providers prefer continued headache education through pain societies and pain journals. Therefore, local and national headache societies should be aware of the increasing need to provide ongoing headache education to pain specialists by including them in our academic programs. Pain providers also identified patient advocacy for insurance coverage of pharmacological and non-pharmacological treatments as a major area for improvement in headache care.

Supplementary Data

Supplementary data are available at Pain Medicine online.

 

Funding sources: The journal Pain Medicine handled the fees for the American Academy of Pain Medicine’s survey distribution. Dr. Minen’s time was supported by a National Institutes of Health (NIH) grant: NCCIH K23 AT009706.

Disclosure and conflicts of interest: Dr. Minen reports no disclosures. Ms. Yang reports no disclosures. Dr. Ashina advises/consults for Allergan/Abbvie, Amgen, Eli Lilly, Impel NeuroPharma, Novartis, Satsuma, Supernus, Percept, and Theranica, serves as associate editor for Neurology Reviews, and is on the advisory board for Journal of Headache and Pain. Dr. Rosen advises for Allergan/Abbvie, Amgen/Novartis, Biohaven, Curelator, Eli Lilly, Impel, Lundbeck, Supernus, Teva; speaks for Allergan/Abbvie; research support/clinical trial site: Eli Lilly, Allergan; serves on the Editorial Board of Current Pain and Headache Reports. Dr. Duarte consults for Lilly, Allergen, AbbVie, and Biohaven.