Blog

Ultrasonography use in the diagnosis of neck pain

Critical Appraisal

Stecco, A., Meneghini, A., Stern, R., Stecco, C., & Imamura, M. Ultrasonography in myofascial neck pain: randomized clinical trial for diagnosis  and follow-up. Surgical and Radiologic Anatomy, 2014;36(3), 243-253.

A. ARE THE STUDY RESULTS VALID?

1. Was there an independent “blind” comparison with a reference standard?

The preclinical study done to collect normative data from 25 painfree subjects was performed by clinicians blinded to the patient’s condition (or lack thereof). Initial and follow up measures were taken by blinded physician(s) in the neck pain group (n=28), but it is not mentioned how many clinicians took these measures and what the variability may be amongst those clinicians.

According to Weinstein et al “If the test relies on a trained observer to interpret it, then two or more such observers are needed to independently interpret the test “.¹  There is no mention if more than one tester was involved in interpreting the ultrasound images.

In this study the purpose was to determine if thickness of fascia could be used to diagnose myofascial neck pain. Myofascial neck pain is not  a well defined disease process and the author  mentions that  it is often used for the purposes of exclusion and is based only on clinical determinants.² The current reference standard for the diagnosis of myofascial neck pain is criteria set out by Travell and that is what is still frequently used.³   A systematic review states that further research is needed to test the reliability and validity of this diagnostic criteria and that until this is established, claims for effective interventions in treating this condition should be viewed with caution. ⁴  It is not mentioned if patients in the pain group met this criteria, only that they were excluded of more serious pathology.

What the author presents as a reference standard is a group of test results (pain, ROM, disability) and the patient’s response to “fascial manipulation”.

I feel in order to establish normal values of fascial thickness  in painfree controls we need further data to account for differences in age, gender, lifestyle, etc. They don’t consider other possible anthropometric  variables that may affect fascial thickness in people.

2. Did the patient sample include an appropriate spectrum of patients to whom the diagnostic test will be applied in clinical practice?

Yes. The patient group included were those that had been diagnosed with chronic nonspecific  neck pain. The authors postulate that a diagnosis of myofascial neck pain can be made based on thickening of the fascia in the neck. Subjects chosen had chronic neck pain without concomitant neurological, rheumatological, or orthopaedic syndromes. Patients’ xrays of the past three months were viewed to exclude disorders related to severe spondylo-arthritis or severe decreases in intraforaminal spaces.

3. Did the results of the test being evaluated influence the decision to perform the reference standard?

As the author’s reference standard was  the result of measures (ROM, pain, and disability) and a patient’s response to treatment, no, the results of the ultrasound did not influence the decision to perform the treatment. The authors followed these patients and documented their symptomatology and response to treatment. All subjects had the same ultrasound images performed and none of the images were discarded.

4. Were the methods for performing the test described in sufficient detail to permit replication?

Using an Aloka Prosound manchine with 38mm linear array transductors, 5-10MHz, patients supine,the following areas were examined:

–              Sternal end of SCM lateral to the cricoid cartilage.

–              Scalene medius midway between the mastoid process and the first rib

–              Fascia and muscle thickness was averaged over the width of the transducer.

–              100micrometers of gel was used with no skin compression.

–              Transducer was moved in the coronal plane until the thickest value of the muscle belly was found, the axis of the transducer was moved until the direction of the perimysium was parallel to the transducer and the two extremities of the belly in the monitor were thickest.

–              Layers of dense, loose connective tissues of the deep fascia were measured.

B. WHAT WERE THE RESULTS?

1. Consider the pre -test and post-test possibilities. i) Are likelihood ratios for the test results presented or data necessary for their calculation provided?

A person was considered to have the diagnosis of myofascial disease if the SCM fascia measured > 0.15 cm. This is based on a value higher than 2 SD in the norms. The data presented are averages of thickness for each group so we are unable to calculate sensitivity or specificity.  Predictive values can not be determined as data is not included. We cannot know  the likelihood that a patient has the disease if their fascia is over 0.15 cm.

C. CAN THE RESULTS BE APPLIED TO YOUR PATIENT(S)?

1. Will the reproducibility of the test result and its interpretation be satisfactory in my setting?

Inter and intratester reliability of this test is not reported.

2. Are the results applicable to my patient?

Guyatt et al state that “Test properties may change with a different mix of disease severity or with a different distribution of competing conditions.”⁵  “If you practice in a setting similar to that of the study, and if the patient under consideration meets all the study eligibility criteria, you can be confident that the results are applicable.” The subjects in this study are individuals with chronic neck pain without concomitant neurological, rheumatologic or orthopedic syndromes aged 27-52. This represents a  portion of my caseload, although most neck pain sufferers I see are over 55.

3. Will the results change my management?

If a patient comes to me with a history of chronic neck pain whereby other conditions have been excluded such as arthropathies and foraminal compression, the pretest probability of myofascial pain or nonspecific neck pain would already be quite high. I would also be inclined to rule out discogenic pain that responds to repeated movements. This addition of a diagnostic ultrasound would not likely add value to that.

Imaging of these patients is probably not practical clinically. Myofascial trigger points have also been identified using MRE (magnetic resonance elastography) which is more useful in terms of research. It may be useful to use diagnostic ultrasound to identify trigger points for the purposes of needling or massaging; however, palpation is faster and has been found to be reliable among trained clinicians.6

This study suggests that reductions in ROM may be related to the thickened fascia and thickened fascia may be a useful biomarker for pathophysiological  processes that could predispose one to chronic pain, but one cannot assume this at this point.

4. Will patients be better off as a result of the test?

Guyatt et al state “The value of an accurate test will be undisputed when the target disorder is dangerous if left undiagnosed, if the test has acceptable risks, and if effective treatment exists.”⁵  In Stecco’s study the target disorder described is “chronic neck pain”. This disorder is not dangerous and through appropriate history and examination we can identify a TBC. Ultrasound results will not likely alter this. Further testing with imaging is expensive. Patients may not be better off as performing imaging in patients with chronic pain may increase catastrophizing, and shift locus of control onto the practitioner.

Although we see a relationship between decreased range of motion and fascial thickening we cannot infer cause and effect.  Is the thickened fascia causing the pain and range of motion restriction, or is fascia thickened because of pain and altered movement patterns?

A perfect study to determine diagnostic accuracy of a test will compare it against a gold standard. I feel a problem with this study is that there is no reference standard for chronic nonspecific pain. Weinstein states that a reference standard should be close to infallible,¹ otherwise we have an imperfect gold standard bias which can mislead us to think a test is more accurate than it is. Stecco et al have combined multiple tests (pain, disability, ROM measures) to construct a reference standard outcome.⁷ His pre-specified rule for the target condition (chronic neck pain) may result in misclassification of patients.

He makes the following assumptions to create his reference for ultrasound diagnosis:

• Chronic myofascial pain is caused by fascial thickness;
• FM is an effective way of treating myofascial pain
• Therefore FM is an effective way of treating fascial thickness

His method of inductive reasoning causes weakening of his argument. He has used one assumption to conclude a premise (increased fascial thickness causes MFPS) and then used that premise to build a new induction.(fascial manipulation reduces fascial thickening and therefore reduces the symptoms of MFPS) Just because fascial manipulation and MEL decrease fascial thickness doesn’t mean that a diagnosis of MFPS can be made based on the results of his RCT. Inductive reasoning is useful to create testable hypotheses but each step must be tested. One cannot infer causality.

His paper says fascial thickness causes pain and decreased ROM, therefore, treat the fascial thickness and you can decrease the pain and increase ROM. He is trying to validate the ultrasound test by associating it with a response to treatment to see if the test can predict who will benefit from fascial manipulation. By proving that this test is meaningful in practice perhaps over time we may reach a point where it is considered valid and information gathered is sufficient to allow use of the test clinically with confidence.

According to Weinstein et al  there are three phases to developing a diagnostic test: the first is the exploratory phase which is typically small, involving patients with disease and healthy controls.¹ If there is no difference between these two populations then the test is not usually pursued further. Stecco et al identified a difference in 0.15cm in the thickness of fascia amongst those with chronic pain.  The second phase of validating this test would be to test it on a large group of patients with varying degrees of the disorder.

What we may get out of this paper is that reductions in ROM may be related to the thickened fascia and thickened fascia may be just one useful biomarker for pathophysiological  processes that could predispose one to chronic pain. Alternatively, it may just be a biomarker for processes that have already occurred in patients that have chronic pain and may have no prognostic value. Perhaps there is a TBC group that responds to fascial manipulation.

Further questions regarding this study:

• Do other anthropometriic variables affect fascial thickness? Ie male/female, BMI, smoking, diabetes, etc
• Is fascial thickening a cause of pain or is it the result of altered/abnormal movement patterns or muscle recruitment?
• Stecco also states in his paper that many of these CC points used in the technique of fascial manipulation correspond to trigger points and acupuncture points so this mechanism of action may be neurophysiological?

–             Could the changes in ROM seen just represent normal variability since some values increased and some decreased?

–             I thought the tables of information included were unusual. He includes a table of the number of points used in the FM group, yet doesn’t provide baseline demographics, or tables of ROM or VAS measures with SD so we can analyze the data.

• He correlates fascial thickness with pain in Figure 9, but no mention is made for other factors that influence pain in chronic pain patients. How significant is this really in patients with chronic pain?
• The method of reporting change in function is unusual.  In each question on the NDI one needs a difference of >1 to represent a meaningful change, yet in very few categories is this met, and if it is, it is not exceeded by much. Cumulative NPDQ scores may have been more useful. My estimate of cumulative NPDQ scores suggests no significant difference between the treatment groups, and that it barely exceeds the MCID of the NPDQ

 

References

1. Weinstein S., Obuchowski, N. A., & Lieber, M. L. (2005). Clinical evaluation of diagnostic tests. American Journal of Roentgenology, 184(1), 14-19.
2. Gerwin, R. D. (2001). Classification, epidemiology, and natural history of myofascial pain syndrome. Current pain and headache reports, 5(5), 412-420.
3. Simons DG, Travell J, Simons LS. Travell & SimonsMyofascial Pain and Dysfunction: The Trigger Point Manual.Vol. 1. Upper Half of Body: 2nd ed. Baltimore: LippincottWilliams& Wilkins; 1999.
4. Tough, E. A., White, A. R., Richards, S., & Campbell, J.  Variability of criteria used to diagnose myofascial trigger point pain syndrome—evidence from a review of the literature. The Clinical journal of pai n 2007; 23(3), 278-286.
5. Guyatt G, Rennie D, Meade MO, Cook DJ. 2008. User’s Guide to the Medical Literature. A Manual for Evidence-Based Practice.. 2nd ed. United States of America: American Medical Association.
6. Bron C, Franssen J, Wensing M, Oostendorp RA. Interrater reliability of palpation of myofascial trigger points in three shoulder muscles. J Man Manip Ther. 2007;15:203-215.
7. Reitsma, J. B., Coomarasamy, A., Khan, K. S., & Bossuyt, P. M. M. (2007). Evaluation of diagnostic tests when there is no gold standard: a review of methods. Gray Pub..