Factors Associated with Patient Adherence to Biofeedback Therapy Referral for Migraine: An Observational Study

Abstract


Introduction
Grade A. nonpharmacological therapies for migraine include cognitive behavioral therapy (CBT), relaxation, and biofeedback (Campbell, Penzien, & Wall, 2000).Biofeedback, a therapeutic approach used for over 50 years (Giggins et al., 2013), involves presenting an individual with visual and/or auditory feedback from their autonomic physiologic processes.Individuals then learn how to use that feedback to modulate their conscious physical actions (e.g., controlled breathing) and gain insight into how thoughts and feelings impact their bodies.A meta-analysis of 55 studies (including randomized controlled trials (RCTs) and pre-post trials) revealed a medium effect size for several types of biofeedback-based interventions for migraine (d = 0.58, 95% CI = 0.52, 0.62; migraine frequency and perceived self-e cacy showed the most signi cant improvements), which was stable over an average of 17 months (Nestoriuc & Martin, 2007).Non-pharmacologic interventions, including biofeedback, are commonly administered in person by a trained provider.
However, despite proven e cacy and availability, few patients engage in nonpharmacological treatments for migraine.In our studies of primary care providers, only 1.4-10% report referring patients with migraine for biofeedback (M.Minen et al., 2016; M. Minen et al., 2016) though referrals for CBT range from 2.53 to 10% and relaxation therapy up to 40% (M.Minen et al., 2016; M. T. Minen et al., 2016).
When a patient is referred for nonpharmacologic migraine intervention, there are additional barriers to engaging with the treatment.In most cases, such treatments are provided by psychologists.Patients report di culty nding and accessing trained providers and challenges associated with the time and expense required for treatment.There is also a stigma of seeking treatment from a mental health professional, which is enough to discourage many from initiating therapy (Corrigan & Penn, 1999;Kaiser, Mooreville, & Kannan, 2015).Chronic pain, like migraine itself, has long carried an associated stigma, with patients fearing a perception that the disorder is "all in the head" (Young et al., 2013).In one study of > 1,000 people with chronic pain, approximately 30% were concerned that their pain might be felt to be "not real" or "psychological" or expressed doubt that a pain psychologist might be helpful to them (Engel, 1977).
In this study, we sought to understand patient perspectives for adhering or not adhering to the recommendation to pursue biofeedback.We also learned how this safe and effective migraine treatment could improve engagement rates.

Methods
Between 1/21/2020 and 11/1/2022, 119 patients presented to pre-designated providers for care at an urban headache center and were referred for biofeedback.Notably, due to healthcare disruptions from COVID-19 in New York City during 2020 and 2021, recruitment methods changed from in-person (from 1/21/2020 to 3/9/2020) to virtual visits (from 10/26/2020 to 11/1/2022).This study, and all modi cations to accommodate COVID-19 disruptions, were approved by NYU Langone Health's Institutional Review Board (see Fig. 1).Of note, patients in the study could be new or returning Headache Center patients.

Pre-Covid-19 Pandemic
Between January and March 2020, members of the Headache Research Team approached in-person potentially eligible patients waiting for their headache center appointments.Inclusion criteria comprised being greater than or equal to 16 years of age, speaking English, having a diagnosis of migraine, and being referred to biofeedback therapy.Informed consent was completed for interested, qualifying patients.Patients completed a paper questionnaire consisting of baseline questions and measures.Completed questionnaire responses were transferred from paper questionnaires into REDCap (Harris et al., 2009) by a research team member.A second research team member reviewed the original questionnaires and REDCap records to screen for typos or manual errors.
Of note, the study was paused during the height of the COVID-19 pandemic in New York City and then transitioned to entirely virtual.

During the Covid-19 Pandemic
Between October 2020 and November 2022, patients presenting for virtual (telehealth) headache center appointments to pre-designated providers were screened by the study team coordinator after completing their appointments via their EMR.Eligibility criteria remained the same.Patients referred to biofeedback during their appointments were called by the study coordinator and invited to participate.Interested patients completed the informed e-consent process virtually over the phone with the study coordinator using REDCap.Once the informed e-consent was completed, patients completed baseline questionnaires using the same REDCap link.The study coordinator remained on the phone with the patient to answer any questions that arose while completing the questionnaires.All virtually recruited patients were moved to the follow-up arm as they had all been referred to biofeedback.

Measures
Across all patients, the baseline questionnaires included demographic questions, including name, date of birth, age, gender, race, ethnicity, height, weight, years of formal education, highest degree achieved, health insurance coverage, drug prescription coverage, and annual income (Table 1).The baseline questionnaire also included questions speci c to the patient's headache experience, prior history of behavioral therapy treatment for headache, headache medications used, and any medical or psychiatric comorbidities (See Appendix 1 for complete baseline questionnaire).At their 1-month (+/-10 days) follow-up, patients were asked whether they followed up with the referral for biofeedback and their experiences with biofeedback therapy.Patients were also asked to respond to questions regarding their satisfaction with treatment from their biofeedback healthcare provider (see the Appendix for the complete list of questions).
We report the results of baseline and follow-up data collection, with the follow-ups being completed virtually, as initially intended, by the study coordinator via phone or REDCap email.Data were captured via REDCap questionnaire either as completed by patients or manually entered by the study coordinator if patients preferred to dictate responses over the phone.For patients recruited in person and whose study period was disrupted by COVID-19, a brief, additional survey was added to differentiate between traditional and COVID-19-speci c barriers to accessing behavioral therapy treatment.

Results
Of the 119 patients who presented for care between January 2020 and November 2022, 51 were fully enrolled in the study; 25% (13/51) were enrolled in person by a member of the Headache Research Team before the COVID-19 pandemic, and most 75% (38/51) were enrolled virtually after the onset of the COVID-19 pandemic (See Fig. 1).
Of those who had been referred previously for behavioral therapy for headache but did not pursue the recommendation, most (65%, 11/17) felt that they did not have time, and nearly all (88%, 15/17) endorsed nancial obstacles such as treatment cost and/or insurance coverage, and 47% (8/17) found it too di cult to schedule an appointment due to limited provider availability.Only 30% (5/17) of patients did not think the treatment would work, and just 6% (1/17) did not know enough about it.

Qualitative Results
Patients addressed four text-response questions upon completing their follow-up questionnaire (see Table 3 for qualitative themes and subthemes).4) Hesitation due to prior experience with biofeedback; 5) Time constraints.
To gauge the kinds of providers patients might prefer, they then answered, "Do you have a preference for seeing a physical therapist or a behavioral therapist for treatment of migraine?"and the following themes emerged: 1) Any provider that accepts insurance; 2) Any provider that is quali ed and/or experienced; 3) Physical therapist because of familiarity; 4) Preference for a physical therapist due to prior experience with physical therapy; 5) Preference for a behavioral therapist due to prior experience with behavioral therapy.
Patients then answered, "What do you think is the most important reason you did not make an appointment for biofeedback?"Themes included: 1) Insurance coverage and expense; 2) Pursuing medication-based treatment; 3) Did not think it would help; 4) Lack of response from biofeedback provider; 5) Too much work to nd a provider; 6) Time constraints.
For patients who had made appointments with a biofeedback provider, "How many appointments have you attended?What are your impressions thus far?" Themes about appointment impressions included: 1) Signi cantly positive results; 2) Enjoyed educational aspects.All six patients who had attended appointments endorsed positive impressions.

Discussion
In patients with four or more headache days each month, migraine treatment characteristically involves a combination of acute and preventive medications.Still, RCTs have found that the ideal treatment for migraine typically comprises a blend of pharmacologic and non-pharmacologic treatment (Holroyd et  Our prior studies found that reasons for the gap between treatment e cacy and treatment utilization include few trained providers, di culty with insurance coverage (being billed under mental health bene ts), and patient-perceived barriers.Further, our study may reinforce that stigma, and access may be barriers to behavioral therapy because, interestingly, more patients reported trying acupuncture for headache treatment at baseline, followed by physical and behavioral therapy.
Within the neurology setting, patients with more signi cant migraine-related disability are likely to pursue behavioral therapy (M.T. Minen, Jalloh, Begasse de Dhaem, & Seng, 2020) though the preference not to have to pay for behavioral therapy was identi ed (M.Minen et al. et al., 2020).In this study, we found that patients who had used, on average, eight migraine preventive medications still had not pursued biofeedback compared to those who used four migraine preventive medications on average.A theme emerged that patients need clari cation regarding what biofeedback is and what it entails.
We sought to understand better whether patients with migraine might prefer a therapist with a background in psychology or a related eld, e.g., a psychologist or a therapist with a background in physical therapy (PT).We found that our patients had mixed responses; some preferred one over the other for various reasons, e.g., they were more familiar with one eld than another or had already tried a provider in a eld.They wanted to try a provider in a different eld with different perspectives based upon the additional training, or they might want the most experienced provider.

Strengths
This is the rst study to examine whether people with migraine pursue biofeedback and their reasons for doing so.This is also one of the rst studies to explore patient perspectives on the types of providers from whom they would want to see to learn the biofeedback technique.

Limitations
This was a small study, and the Covid-19 pandemic may limit our results; fewer patients may have wanted to engage in in-person therapy and may have yet to realize that, in some instances, biofeedback was being delivered at times via telehealth.Some patients may also have felt that the pandemic would improve shortly and that they could defer the in-person therapy until they thought it was safe to partake in it.Other limitations were that our patient population was not heterogeneous.Most patients identi ed as non-Hispanic white females (92%, 47/51), and over half (55%, 28/51) had obtained an advanced degree.

Table 3
al., 2010; Powers et al., 2013).Nevertheless, patients with migraine rarely engage in Grade A evidence-based behavioral therapies (M.T. Minen et al., 2016; M. T. Minen et al., 2018; M. T.Minen et al., 2020)and thus, in this study, we sought to determine if they might be more willing to engage in biofeedback speci cally.