This systematic review aims to give an update on the reliability of the MDT system for spinal conditions. Only the cervical and lumbar spine were studied as there has been no research published on MDT's reliability for the thoracic spine as of yet.

After the screening process, a total of 12 studies were retained to evaluate MDT's reliability with regards to classification of main syndromes, sub-syndromes, directional preference, centralization and lateral shift. By "sub-syndromes" the authors mean either the direction used for remodeling a Dysfunction or the direction used for reduction of the Derangement.

Summary of the results:

Main syndromes:
Lumbar (6 studies) - conflicting evidence for inter-rater reliability
Cervical (3 studies) - conflicting evidence for inter-rater reliability

Sub-syndromes:
Lumbar (3 studies) - strong evidence of acceptable inter-rater reliability
Cervical (2 studies) - conflicting evidence for identifying sub-syndromes

Directional preference (DP):
Lumbar (2 studies) - conflicting evidence for identifying DP
Cervical (1 study) - limited evidence of unacceptable inter-rater reliability for identifying DP

Centralization:
Lumbar (2 studies) - conflicting evidence for identifying centralization
Cervical (0 study) - no studies found regarding the reliability of centralization

Lateral shift:
Lumbar (2 studies) - limited evidence of unacceptable intra-rater reliability and conflicting evidence for inter-rater reliability for identifying lateral shift

Although the authors point out that there have been relatively few studies on the subject of reliability, it is important to put that in context. There may be few studies in absolute numbers, but in comparison to other musculoskeletal systems, MDT is ahead by far. However, most of the research has been conducted for the lumbar spine therefore the results should be interpreted with some caution. They also identify a few factors that may explain why conflicting results were found:

1 - Training

Although one would expect there to be increasingly better reliability as a clinician completes Parts A through D, this has been shown to be untrue.¹ As there is no examination after each course, there is no formal way to assess if participants have retained the information learned on those courses. In addition, there is no way of knowing if the clinicians use the MDT system correctly in their practice, or with what proportion of their patients. Rather than the amount of hours spent on the courses, it seems that practice and/or the process of studying for the credentialing exam are beneficial in producing more reliable classifications as the authors mention:

" Overall, we observed acceptable inter-rater reliability of the MDT system for identifying main and sub-syndromes in patients with low back pain when performed by credentialed MDT therapists and physiotherapists who have received a MDT diploma."

"This review shows based on high methodological quality studies that the MDT system appears to have acceptable inter-rater reliability for classifying patients with back pain into main/sub-syndromes when applied by therapists who have completed the credentialing examination, but unacceptable reliability in other therapists."

A parallel can be drawn between learning to be an MDT clinician and learning to drive a car. One can attend driving lessons, but it takes hands-on practice to become a proficient driver. Therefore, if we conducted an experiment where we wanted to study a group people's decision-making at the wheel in various scenarios, it would make sense to use licensed drivers, not student drivers. The results would be more representative.

2 - Methodological flaws

Depending on the study, there were different ways to assess reliability: video assessment, a live assessment with a clinician and an observer or two consecutive assessments of the same patient by two different clinicians. In the cases where the patient was assessed by two therapists back-to-back, reliability may have been compromised depending on the nature of the patient's condition. For example, with Derangements, certain baselines being present for the first clinician and absent of the second. Also, since the assessment involves tracking symptom behavior through repeated movements it may become difficult for the patient to tell if they are getting worse or better due to the sheer amount of movements performed.

In their suggestions for future research, the authors mention the option to "include multiple raters in independent rooms observing someone’s assessment of the patient in real time, but with the opportunity to interact with the patient during the assessment". These interactions would then be exclusive between the patient and the rater in question. In addition, another element that can lead to confusion for researchers is the labelling of Derangements 1-7, which is no longer formally taught. To prevent this, identification of sub-syndromes should be made in terms of the direction of reduction/remodeling. Finally, if we are to measure the true reliability of the MDT system, future studies should consider using only Certified or Diplomaed clinicians to get a more accurate picture.

1.Werneke MW, Deutscher D, Hart DL, Stratfoed P, Ladin J, Weinberg J, Hebowy S, Resnik L., McKenzie lumbar classifications: inter-rate agreement by physical therapists with different levels of formal McKenzie post-graduate training., Spine, 39(3):E182-90, 2014

Lumbar: Studies into assessment, diagnosis and procedures