This policy was originally based on a 1996 TEC Assessment  which concluded that evidence was insufficient to demonstrate the effectiveness of biofeedback for treatment of tension or migraine headaches. The available evidence did not clearly show whether biofeedback’s effects exceeded nonspecific placebo effects. It was also unclear whether biofeedback added to the effectiveness of relaxation training alone.

 

Between the 1996 TEC Assessment and policy update in 2002, controlled studies of biofeedback for headaches, using various types of relaxation therapy or no therapy as controls, reported conflicting results. The majority of studies did not support a benefit to biofeedback in terms of frequency and severity of headache symptoms. Negative studies suggested that there was no difference between the effects of biofeedback and other types of relaxation therapy alone. Studies reviewed in detail are described here.

 

Kroner-Herwig et al selected 50 pediatric patients with either tension headaches or combined tension-migraine headaches. Four treatment groups were created, based on combinations of the presence or absence of parental involvement in treatment and whether patients received either relaxation training or biofeedback. A waiting-list control group was also included. Several analytic approaches were used, one of which found biofeedback to have better effects on pain than relaxation. Another study by Bussone et al compared biofeedback-assisted relaxation training in adolescents with a control group, finding better pain improvement in the former group. Scharff et al enrolled 36 children and adolescents and randomized them to hand-warming biofeedback, to hand-cooling biofeedback, or to a waiting list. Patients treated with hand-warming biofeedback achieved greater degrees of clinical improvement than either of the other two groups. Hand-cooling biofeedback could be considered a placebo. Sartory et al randomly assigned 43 children to either relaxation training plus stress management, biofeedback plus stress management, or drug therapy with a beta-adrenergic blocking agent. Both the relaxation and biofeedback groups had better therapeutic outcomes than the drug therapy group.

 

Relaxation training was not considered to be an appropriate control for the non-specific effects of biofeedback. In addition, some studies indicated that the physiologic parameter "fed back" to the patient may not be related to the pathophysiology of headache. For example, Andrasik and Holroyd examined the correlation between success in controlling scalp muscle tension and the reduction in headache symptoms. Thirty patients with tension headaches were taught to either decrease, keep stable, or increase frontal muscle tension, but were all led to believe that they were decreasing muscle tone. Despite changes in muscle tone in the intended direction, the degree of headache relief was the same in all groups. In another similar study, patients who were told that they were successful at decreasing muscle tension, regardless of the actual results, achieved greater reduction in symptoms.  Similar results were reported in patients with migraine undergoing thermal biofeedback.

 

In 2007 and 2008, Nestoriuc and colleagues published systematic reviews of biofeedback for migraine and tension-type headaches. The meta-analysis for treatment of migraine included 55 studies (randomized, pre-post, and uncontrolled) and 39 controlled trials, reporting a medium effect size of 0.58 (pooled outcome of all available headache variables) for treatment of migraine. (9) Effect sizes were computed using Hedges’ g, which refers to the mean difference between the experimental and control groups divided by the pooled standard deviation. For treatment of tension-type headaches, 53 studies met criteria for analysis; these included controlled studies with standardized treatment outcomes, follow-up of at least 3 months, and at least 4 patients per treatment group.  Meta-analysis showed a medium-to-large effect size of 0.73 that appeared to be stable over 15 months of follow-up. Biofeedback was reported to be more effective than headache monitoring, placebo, and relaxation therapies. Biofeedback in combination with relaxation was more effective than biofeedback alone, and biofeedback alone was more effective than relaxation alone, suggesting different elements for the two therapies. Although these meta-analyses are limited by the inclusion of studies of poor methodological quality, the authors did not find evidence of an influence of study quality or publication bias in their findings.

 

Another meta-analysis assessed psychological treatments of recurrent tension headache or migraine in children.  Three studies were included that compared relaxation combined with biofeedback versus relaxation training alone. In general, small standardized effect sizes (0, 0.5, and 0.25) were reported from the 3 studies for the addition of biofeedback on headache symptoms (frequency, intensity, and duration of headache). Small standardized effect sizes were also reported for clinically significant changes (> 50% reduction) in headache symptoms (0.20, 0.34, and 0). A 2006 systematic review of non-pharmacological treatments for migraine concluded that the current literature does not show clear effectiveness of biofeedback for migraine in children.  

 

Martin et al compared cognitive behavioral therapy (CBT) versus temporal pulse amplitude (TPA) biofeedback (8 weekly sessions plus homework) or wait-list control among patients who volunteered for a study of psychological treatments.  Thirty patients with migraine and 21 with tension-type headaches were randomized to 1 of the 3 treatments; 51 completed the protocol (20% dropout) with no significant difference in loss to follow-up among the groups. Patient logs showed an average reduction in headaches of 68% for the CBT group, 56% for biofeedback, and 20% for the control condition. Clinically significant improvement, defined as at least 50% reduction in either headache rating or medication use, was observed in 78% of the CBT group, 63% of the biofeedback group, and 23% of the control group. The cognitive mediators (self-efficacy and locus of control) that had been hypothesized to underlie efficacy of both biofeedback and CBT were not found to be associated with improvement for either treatment. Statistical analysis was limited by the small group sizes.

 

In response to requests, input was received through 3 physician specialty societies and 3 academic medical centers (4 inputs) while this policy was under review. While the various physician specialty societies and academic medical centers may collaborate with and make recommendations during this process, through the provision of appropriate reviewers, input received does not represent an endorsement or position statement by the physician specialty societies or academic medical centers, unless otherwise noted. Clinical input considered biofeedback to be a reliable and appropriate nonpharmacologic option for treatment of headaches.

 

Based on clinical input, physician specialty society recommendations, and the evidence available at this time, biofeedback may be considered an appropriate technique to treat migraine and tension-type headaches. Biofeedback, along with other behavioral techniques such as relaxation training, may be particularly useful for children, pregnant women, and other adults who are not able to take medications.

 

The National Institute of Neurologic Disorders and Stroke (2008) state that when headaches occur 3 or more times a month, preventive treatment is usually recommended.  “Drug therapy, biofeedback training, stress reduction, and elimination of certain foods from the diet are the most common methods of preventing and controlling migraine and other vascular headaches. Drug therapy for migraine is often combined with biofeedback and relaxation training.”

 

The American Academy of Family Physicians (AAFP) 2000 guidelines on preventive therapy for migraines, based on evidence review by the U.S. Headache Consortium, recommend relaxation training, thermal biofeedback combined with relaxation training, EMG biofeedback, and cognitive-behavioral therapy as treatment options for prevention of migraine (Grade A recommendation).  Relaxation techniques and biofeedback may be combined with preventative drug therapy to achieve additional clinical improvement (Grade B recommendation). According to the guidelines, nonpharmacologic therapy may be well suited for patients who have exhibited a poor tolerance or poor response to drug therapy, who have a medical contraindication to drug therapy, and who have a history of long-term, frequent or excessive use of analgesics or other acute medications. Nonpharmacologic intervention may also be useful in patients with significant stress or in patients who are pregnant, are planning to become pregnant, or are nursing.

 

The American Academy of Neurology’s (AAN) recommendations for the evaluation and treatment of migraine headaches states that behavioral and physical interventions are used for preventing migraine episodes rather than for alleviating symptoms once an attack has begun.  Although these modalities may be effective as monotherapy, they are more commonly used in conjunction with pharmacologic management. Relaxation training, thermal biofeedback combined with relaxation training, electromyographic biofeedback, and cognitive-behavioral therapy may be considered treatment options for prevention of migraine. Specific recommendations regarding which of these to use for specific patients cannot be made.

 

 

 

 

 

 

 

A literature search conducted through April 2014 did not reveal any new information that would prompt a change in the coverage statement   

 

 

A literature search conducted through April 2016 did not reveal any new information that would prompt a change in the coverage statement.