PT Classroom - Applying Evidence Based Practice: Is massage therapy effective in reducing symptoms of cervicogenic headache in adults? ׀ A Case Study by Theresa A. Schmidt, MS, PT, OCS, LMT

Question: Is massage therapy effective in reducing symptoms of cervicogenic headache in adults?

History of the Case: “DJ” is a 47 year old female with a medical diagnosis of cervicogenic headache post-whiplash injury to her neck, sustained in a MVA on 3/1/07. In the accident, DJ was driving her car at 65mph in highway traffic when she reportedly stopped suddenly to avoid the car ahead, spun and struck a cement wall on the driver’s side of her car. She struck the left side of her head on the window, but did not lose consciousness. Since the accident, she has suffered from neck pain, limited cervical range of motion and intermittent headaches, which become disabling about once per week. She is taking naprosyn 500mg 2x/day and flexiril prn at night, to control her headache. She has no other medical problems. She has had prior whiplash type injuries over the years, but none which caused persistent headaches. JD works full-time as a science teacher. She is married with no children. She has not been able to participate in her usual recreational activities tennis, running, and aerobic dance since the accident. She has tried the usual course of physical therapy with hot packs, electrotherapy, and stretching exercise with minimal temporary relief of the neck pain and headaches. Symptoms return when she is physically active, lifting over 5 pounds, pushing or pulling over 25 pounds, and after attempting to run, dance, or play tennis for over 15 minutes. She has missed several days from work due to the severity of her headaches. She said that massage therapy provides the greatest relief, to allow her to work, but the symptoms return if she exerts herself by lifting over 50 pounds, or if she inverts her head for greater than a minute or two. She had x-rays, neck and brain MRI, which showed DJD of the cervical spine. No further medical testing was performed. She reports her general health is excellent.

 

To determine if the evidence I found is applicable to this patient, I reviewed studies about manual treatments for headache relief. Some of the evidence indicates that manual therapy or massage may be beneficial for patients with headache. Most articles do not specify what type of headache they review, some group different headache types into a single study. Some studies incorporate several interventions to note the effect on headache, which does not allow us to isolate the effect of a particular intervention. Studies use a variety of different treatment frequencies, durations, and follow-up times. It is difficult to compare evidence when a variety of factors are not homogenous. I recommend future studies review the effect of a single or selected group of interventions on a specific type of headache: migraine, tension, or cluster headache. Future studies must look at the differences between varying frequencies, durations and long term effects of the interventions in homogenous subject groups. My impression of the available full text studies, which are systematic reviews, follows.

A systematic review was conducted by Bronfort, et al, entitled “Noninvasive Physical Treatments for Chronic/Recurrent Headache” in Cochrane Database of Systematic Reviews (1). The authors reviewed 21 randomized and one quasi-randomized controlled trials of 2628 patients with headaches, to quantify and compare the effects of the physical treatments on headaches. The trials were scored for methodological quality, with 100 being the highest score. Ten studies scored at least 50. The review indicated variable results for different types of headaches. One study reported that a combination of auto-massage, stretching, and electrotherapy showed weaker evidence than spinal manipulation for tension headache relief (2). Manipulation was more effective as prophylaxis than massage for short-term relief of cervicogenic headache. Few adverse effects were reported for physical treatments for headaches. The reviewers determined that additional studies of a more homogenous nature are needed to determine which types of physical treatments are effective for reducing headaches. The studies considered several types of headaches including cervicogenic, migraine, cluster, tension and, posttraumatic, and mixed headaches. The authors determined that none of the evidence was conclusive due to the heterogeneity of the samples reviewed.

The treatments reviewed compared the physical treatment to placebo or another intervention or no intervention. Outcome measures included pain level, patient satisfaction, activities of daily living/functional level, medication use, duration and frequency of headache, and level of improvement. Bronfort’s review included a description of the methods of review. Studies were analyzed by two separate reviewers who extracted specific data from each article using a checklist. The reviewers were not blinded, they standardized the effect size scores and outcomes data using percentage points. The quality of individual studies was reported on a scale of zero to 100; studies scored at least a 50 or more were considered to be of high quality. Reviewers classified twenty methodological items as “informativeness” items or “internal validity” items. Some of the items were: randomization, inclusion/exclusion criteria, comparability of variables for groups studied, reliability/validity, blinding, bias, post intervention follow-up periods, description of interventions, comparison to other interventions, study objective or hypothesis, analysis consistency with the design of the study, statistical power, confidence intervals, dropout impact, random data allocation, intention to treat analysis, adjustments made to the number statistical tests, and validity of main conclusions (3).

Researchers classified the quality levels for each study into five levels from strong, moderate, limited, preliminary, or conflicting evidence, based on validity scores. Trial results were pooled if the interventions, outcomes or patient population were homogenous. One primary outcome measure was headache index or pain intensity. Based on the exhaustive nature of the quantitative analysis of trial results and the level of evidence used to evaluate the reviewed studies, it appears that the conclusions generated were substantive.

The valid results of this review are important for many reasons. Physical therapists often use a range of individual physical treatments or a combination of interventions to manage headaches. The reviews included comparing a wide range of individual interventions to combination package interventions and also to controls or placebos. Several of the studies assessed massage as one of the interventions. It appears that the heterogeneity of the interventions used make it difficult to draw conclusions. Of the 22 trials reviewed, only one directly addressed the effect of massage on cervicogenic headache (4). The studies were very different in the number of interventions performed (1-12 over 1-6 weeks). More research is needed to determine the optimal number of interventions and duration of treatment to perform to achieve the desired outcome of headache reduction.

In studies that included massage in combination with other physical treatments such as heat, ultrasound or acupuncture, one cannot decide whether any individual intervention or the combination most influenced the outcome measures. Since massage was not tested as the only variable, no direct conclusion can be drawn about the benefit of massage alone.

The Bronfort review discussed limitations of the study, including publication bias. The authors noted that three of the studies reviewed were ones in which at least one of the authors participated(5). The review included only published research, which is more likely to have positive outcomes than unpublished research (6). The authors also mentioned that clinical trials done in languages other than English may have been missed. The authors used two methodological scoring systems in order to minimize bias. Overall, the review was very comprehensive, with few limitations. The authors made an excellent summary point, stating, “Authors too often draw inappropriate conclusions when they declare treatment effectiveness based solely on presence or absence of statistical differences between a test treatment and a control. To inform decisions about management of individual patients, it may be much more appropriate to think in terms of available treatment options which have shown a meaningful clinical effect, rather than choosing or discarding specific therapies based on mean group differences of undefined clinical importance(7).”

The authors concluded that, “No single approach to interpreting findings from RCTs and systematic reviews is perfect. To inform decisions about the management of individual patients, it may be more appropriate to think in terms of available treatment options that have shown a meaningful clinical effect, rather than choosing or discarding specific therapies solely based on mean group differences of undefined clinical importance (8).” Their conclusion is reasonable in consideration of the fact that therapists select from a wide variety of interventions to treat headaches, and the outcome of primary importance is the response of the individual patient to a given intervention. This systematic review provided some evidence to indicate that manual therapies, including massage and spinal manipulation, may help to reduce headaches, with little adverse effects. More studies must be done using specific headache classifications, subject homogeneity, and specific outcome criteria, in order to identify the types of therapies effective in reducing headache pain. The results of this review are applicable to my case study. In treating my case study patient’s headache, I utilized a combination of massage therapy, joint mobilization, (manipulation) therapeutic exercise, and relaxation training. The evidence from the review provides moderate support for the use of the manipulation, and minimal support for the use of massage, exercise and relaxation training. Overall, there are few side effects of these interventions, and if my patient shows improvement with any of them, I will choose to continue the treatment. Based on the outcomes of the review, I may focus more time on manipulation to reduce headaches, and measure the outcome for my individual patient. If the outcome is favorable, I will continue the treatment.

Another systematic review of the effect of manual therapies on headache reduction was undertaken by Fernandez de las Penas, et al. (9) The authors reviewed the computerized databases Medline, Pubmed, Ovid, Cochrane, AMED, MANTIS, CINAHL, EMBASE, and PEDro. They selected controlled clinical trials and reviews to determine the effect of manual therapy on tension type headaches. The authors reviewed publications based on specific inclusion criteria using a standardized format for data extraction. The authors stated that they agreed on the items in the form, and verified “observer reliability” using coefficient (K= 0.79) (10). Studies selected included open uncontrolled studies and randomized controlled trials, which included manual therapy treatment for tension headaches. The studies were limited to English language publications post-1994. The PEDro quality scoring method was used to determine methodological quality on a scale of 0 to 10, with a score greater than five considered high quality and a score less than five to be low quality. Authors used either effect size (ES) or an outcome quality score of P<0.05 to determine the level of scientific evidence for the trials, ranging from strong, moderate, limited, or inconclusive evidence. The primary outcome measure included headache frequency, intensity, or duration. Fifty-five articles were found, however, only six total articles met their eligibility criteria for final selection (11). The total number of subjects across all studies was 405.

The manual therapies evaluated in the studies included “spinal manipulation, classic massage, connective tissue manipulation, soft tissue massage, Dr. Cyriax’s vertebral mobilization, manual traction, and CV-4 craniosacral technique (12). The studies varied widely in the duration of intervention, from a single session to 6 weeks of therapy. “(mean=3.6+/- 1.9 weeks)…sessions ranged from 1 to 20 (mean = 11.6+/- 7.3) (13). Four of the six studies assessed a single intervention, the others used a combination of techniques. As a result of the mix of interventions, a meta-analysis could not be done. Different controls were used in each study, with variable follow-up periods ranging from immediately post-treatment to 6 months post- treatment. Some used medication groups as controls, others compared manipulation alone to a combination of manipulation with traction or massage or placebo laser, to neck exercises or to no treatment at all. Because so many types of treatments were considered, it is difficult to draw conclusions about any one type of intervention. The interventions are too heterogenous to draw a final conclusion about any one intervention.

The methodological quality of the studies reviewed was determined by use of PEDro scores, which ranged from 2-8 points (mean=5.8+/-2.1) (14). Four of the six trials reviewed scored from 6-8 points, and were therefore considered to be of high quality. The effect size (ES) could only be calculated in two trials. Effect sizes were reported as (0.3) on headache frequency and intensity values (0.49) on pressure pain threshold, and (0.1) on range of cervical motion for the Cyriax mobilization group (15). The craniosacral group has the largest ES (0.84) on pain outcomes relative to the control and exercise groups (16). The authors determined that the level of evidence for the value of spinal manipulation was inconclusive, based on the outcome of two high quality studies and one poor quality study, which had differing results. They found limited evidence for connective tissue massage and craniosacral therapy, since each had only one trial study performed. The authors concluded that due to the limited number of high quality studies on manual therapies, (RCTs) and the heterogeneity of the samples and techniques used, there is little evidence to support the use of manual therapy for tension headaches.

The authors determined that the reviewed studies were nonconclusive in regard to the efficacy of manual therapies on headache for several reasons. The studies used different outcome measures, different manual techniques, and inconsistency in the design of the clinical trails. (Some assessors were not blinded, or did not use an intention-to-treat analysis) They suggest using other standard outcome measures such as the Pressure Pain Threshold, McGill Pain Questionnaire, or Neck Disability Index, to make the outcomes measures more homogenous (17).

Publication bias may be evident in this review, since only published English studies were considered, and mostly positive outcomes may be represented. The authors recommended that future studies of higher quality design with more homogenous subjects and interventions would better represent the outcomes of manual therapies for tension headaches.

One could use the results of Fernandez de las Penas’ review to consider the type of manual therapy to use with this case study patient. They did remark that there if limited evidence for effectiveness of soft tissue manipulation for reducing headaches, and inconclusive findings for the use of manipulation. I utilize both treatments in my plan for headache reduction for my patient, with good results in pain reduction and lower duration of pain, and in the improved scores on the Neck Disability Index. Since the soft tissue intervention has low risk of side effects, (compared to medication or vertebral manipulation) I will continue to use it. I always use outcome measures in my practice, including those recommended by the authors, such as range of motion, Neck Disability Index, and the SF-36 questionnaires. When I note positive outcomes based on these objective measures, I continue to apply the manual therapy to get favorable results. Simply finding a lack of evidence in the literature to support manual therapy does not preclude its use in my practice. I find that using objective measures such as the Neck Disability Index and other standard outcome measures to assess the results of my interventions is superior to using literature review, because I can see the immediate and long term results with my patients on a one-to-one basis. I will use the literature as a guide to determine which interventions may be efficacious for a given problem, such as headaches (18-20). If I do not generate positive results within two to three visits, I will return to the literature to find alternative interventions. Since the research is constantly updated, I obtain the latest information directly in my e-mail from Medline. It is critical to know how to perform a literature review to locate recent studies which may prove beneficial to patients and to practice as a manual therapist. With practice, one can find more high-quality evidence to use in a reasonable period of time. As noted in each study reviewed, more research is needed with better methodology to determine the optimal interventions for a given problem.