Professional Documents
Culture Documents
Accepted Manuscript: 10.1016/j.msksp.2018.09.006
Accepted Manuscript: 10.1016/j.msksp.2018.09.006
PII: S2468-7812(18)30206-6
DOI: 10.1016/j.msksp.2018.09.006
Reference: MSKSP 1930
Please cite this article as: May, S., Runge, N., Aina, A., Centralization and directional preference: An
updated systematic review with synthesis of previous evidence, Musculoskeletal Science and Practice
(2018), doi: https://doi.org/10.1016/j.msksp.2018.09.006.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title page
previous evidence
PT
RI
Stephen May1*, Nils Runge1, Alessandro Aina2
SC
s.may@shu.ac.uk
U
Nils.A.Runge@student.shu.ac.uk AN
sandroaina@gmail.com
M
1
Faculty of Health and Wellbeing, Sheffield Hallam University, Sheffield, UK
2
Studio di Riabilitazione MDTC, Milano, Italy
D
TE
1
* corresponding author
Faculty of Health and Wellbeing, Sheffield Hallam University, Sheffield, S10 2 BP, UK
EP
s.may@shu.ac.uk
AC
ACCEPTED MANUSCRIPT
PT
centralization and directional preference.
RI
relevant search terms. Abstracts and titles were filtered by two authors;
relevant articles were independently reviewed by two authors for content,
SC
data extraction, and quality.
U
quality of the studies, using PEDro for randomized controlled trials, was
AN
moderate or high in six out of ten RCTs; moderate or high in six out of 12
cohort studies. Prevalence of centralization was 40%, the same as the previous
review. Directional preference prevalence was only 26%, much lower than the
M
previous review; but neither clinical response was recorded in about a third of
patients. Centralization and directional preference were confirmed as key
D
positive prognostic factors, certainly in patients with low back pain, but limited
TE
evidence for patients with neck pain. There was no evidence that these might
be important treatment effect modifiers. One study evaluated reliability, and
found generally poor levels, despite training.
EP
with chronic low back pain. But they do not occur in all patients with spinal
AC
problems, and there was no evidence that they were treatment effect
modifiers.
ACCEPTED MANUSCRIPT
Centralization and directional preference: An updated systematic review
PT
1. Introduction
RI
Centralization is defined as the abolition of distal and spinal pain in response to
SC
Directional preference is defined as the repeated movement that produces
U
centralization, or an abolition or decrease in symptoms, or an increase in
AN
restricted range of movement (McKenzie and May, 2003). Centralization and
M
response to therapeutic loading strategies and thus are clinically induced, and
TE
also as they describe a lasting change. They have been commonly referenced
(May and Aina, 2012), in fact centralization is probably the most commonly
EP
they are potentially useful prognostic and management indicators (May and
AC
Aina, 2012). For instance, the presence of centralization has been associated
with better pain, function, return to work, and non-surgical outcomes both
short and long-term (Long, 1995; Werneke and Hart, 2001; Skytte et al., 2005).
PT
Centralization has been the subject of several systematic reviews (Aina et al.,
RI
2004; Chorti et al., 2009; May and Aina, 2012). All these are reasonably dated;
SC
even the most recent is over five years old now (May and Aina, 2012).
Although now somewhat dated that last review included 54 studies relevant to
U
AN
centralization and eight studies relevant to directional preference. Since the
inclusion date of that last review a number of additional studies have been
M
published (for instance, Albert et al., 2011; Petersen et al., 2011; Edmond et
D
and non-organic signs being the exceptions (Hartvigsen et al. 2015). Thus there
C
AC
are very limited tools for the clinician to determine if a patient might recover
In terms of updating systematic reviews, it has been suggested that the median
survival time of a systematic review is 5.5 years with 23% being outdated
stipulate how the results from the previous review(s) should be included;
PT
to amalgamate the earlier individual studies with the new ones.
RI
Thus it would appear to be appropriate to update the previous systematic
SC
reviews. The aims of the present paper were to summarise previous findings
and to systematically review recent literature since June 2011 relating to all
U
AN
aspects of centralization and directional preference.
2. Methods
M
D
preference, in adults reporting low back or neck pain, with or without radiating
were followed (Harms, 2009). For data prior to the present search (June 2011
AC
to December 2017) data were synthesised from the previous review (May and
Aina, 2012).
which lists references relevant to the McKenzie method, and includes a section
on centralization, was also used. The reference lists of all included articles
PT
were also searched. Search terms used were as follows: centralization, OR
RI
directional preference; OR phenomenon; AND spine pain, OR back pain, OR
SC
neck pain, OR cervical, OR lumbar; used individually and in combination. Titles
and abstract were reviewed initially by one author (AA) to see if they might be
U
relevant and duplications removed. All potential articles were reviewed by two
AN
authors (AA, SM) to determine their final relevance, with any disagreements
M
Data extraction was done independently and blinded to each other by two
EP
quality assessment was done independently and blinded to each other by two
AC
instances there was the option, not used, to refer to the third author if a
There is not full agreement on the best methods to assess quality either in
randomized controlled trials (RCTs), or in cohort studies. There are many tools
ACCEPTED MANUSCRIPT
to choose from for such tasks (Sanderson et al., 2007; Olivio et al., 2008), but
Database (PEDro) scale was used to assess internal and external validity of the
RCTs. The PEDro scale is comprised of 11 criteria (only 10 of which are scored),
PT
has been shown to be valid and reliable, and was used in a recent systematic
RI
review, with a score of 7 or above considered high, 5 or 6 moderate, and 4 or
SC
below poor quality (Young et al., 2018). To assess the quality of cohort and
observation studies a tool for prognostic studies was used that had been
U
adopted from earlier work by Hayden et al. (2013), and was used in a recent
AN
systematic review (Hartvigsen et al., 2015). The quality criteria consist of five
M
domains, with 15 items, which was scored as yes, OR no / unsure / not stated,
D
one-year and 85% of the inception cohort; criterion 14: we made to include the
Assessing methodological quality in the other studies was not possible due to
Studies were grouped according to study design and purpose, such as case
PT
used mostly, except regarding prevalence, for which a meta-analysis was
RI
conducted, pooling individual studies for totals regarding Centralization,
SC
Directional Preference, and no Directional Preference.
U
3. Results
AN
3.1.1 Study selection and characteristics of studies
M
2486 titles and abstracts were initially screened, 101 full texts were reviewed
D
for eligibility, and 43 articles were finally included (see Figure 1). The 43
TE
additional studies since the last review (May and Aina 2012) were: randomized
EP
controlled trials or controlled trials (10), or their secondary analyses (4), cohort
studies (15), or case studies (10), and four cross-sectional studies (see Table 1
C
for full details). Seven papers related to patients with neck pain; the rest to
AC
patients with low back pain. Most studies involved patients with non-specific
neck or back pain; but studies also included specific pain syndromes, including:
sciatica or cervical radiculopathy (3), discogenic pain (2), candidates for lumbar
disc surgery (2), potential red flags (1), and spinal stenosis (1). Although
ACCEPTED MANUSCRIPT
centralization and DP were originally concepts related to the McKenzie
approach of MDT, other classification systems were also referred to, which all
involved some element of these clinical responses. Specifically these were the
PT
al., 2011), the diagnosis-based clinical decision guide (Murphy and Hurwitz,
RI
20111, 2), the Hall classification system (Gregg et al., 2014), a combined
SC
McKenzie and patho-anatomical assessment (Flavell et al., 2016), and a
U
These studies will be discussed relative to their study designs and purpose in
AN
the results section (Table 1).
M
Surprisingly, given the length of time that the concepts of centralization and
TE
directional preference have been extant, ten studies were of a case study
EP
designs (Muir Gray, 1997). One case study (Desai et al., 2012), and two cohort
AC
studies (Van Helvoirt et al., 2014; 2016) described the effect of transforaminal
PT
reversed these pathologies, and showed improvement over time (Heintz and
RI
Hegedus, 2008; Takasaki et al., 2010; Padmanabhan et al., 2011; Williams et
SC
al., 2011; Ojha et al., 2013; Elenburg et al., 2016; Robinson, 2016; Takasaki and
U
AN
3.1.3 Effectiveness of exercises based on centralization or directional
preference
M
Some of the RCTs and trials of MDT utilising centralization and directional
D
Effect and disability at two-three months, and one year (Petersen et al., 2011;
EP
Albert and Manniche, 2012; Halliday et al., 2016; Franz et al., 2017; ), and
C
But there were no significant differences in other trials (Bonnet et al., 2011;
better function, but not pain in patients with neck pain compared to non-
(Rose et al., 2016). In RCTs disability, but not pain, was significantly better
short-term (1m) compared to back school (Garcia et al., 2013); and also short-
term (8w) in Global Perceived Effect, but not other outcomes, compared to
PT
motor control exercises (Halliday et al., 2016); whereas motor control exercises
RI
had a better outcome short-term in another trial (Hosseinifar et al., 2013).
SC
3.1.4 Centralization and directional preference as treatment effect modifiers
U
Because of the nature of all study designs it was not possible to determine if
AN
either symptom response was a useful treatment effect modifier; no trial had
prognostic factors
EP
of disc lesion, but both improved more than the no pain response group
AC
(Albert et al., 2012). Other secondary analyses looked at factors that improved
outcomes. Older age was associated with better outcomes in a MDT group
compared to a back school group (Garcia et al., 2016). Age, severity of leg pain,
pain distribution, nerve root involvement and centralization were not found to
ACCEPTED MANUSCRIPT
be treatment effect modifiers favouring MDT over manipulation; however
nerve root involvement and peripheralization together did make the chance of
success greater especially for the MDT group (Petersen et al., 2015).
PT
directional preference compared to their absence or to guideline-based advice,
RI
was associated with better pain and functional outcomes, but mostly only in
SC
the short to medium-term (Werneke et al., 2011; Al-Obaidi et al., 2013;
Edmond et al., 2014; Gregg et al. 2014; Rose et al., 2016; Surkitt et al., 2016;
U
AN
Werneke et al., 2018; Yarnbowicz et al. 2018); but did not add to predictive
factors in one study (Werneke et al., 2016). (See table 1 for details)
M
from 21 studies (Table 2, which also shows the summary data from the
EP
directional preference was reported in 1716 (33.5%), and only 70 patients with
LBP were not counted in one of these groups. The total included 720 patients
with neck pain in who the following was reported: centralization, 56%,
was found in 44% of those with chronic low back pain in the 11 papers that
ACCEPTED MANUSCRIPT
reported specifically on chronic, as opposed to mixed symptom duration, of
PT
smaller proportions with lateral movements (mostly 10-14%); and less than
RI
10% for flexion (Table 2 for prevalence from previous and this review).
SC
3.1.6 Cross-sectional studies
U
Only one recent study considered the reliability of therapists to identify
AN
centralization, directional preference, and other aspects of the MDT
M
assessment process (Werneke et al. 2014). Reliability was generally weak, with
D
15 kappa values (k) all below 0.44 ; the level of training in MDT that the
TE
therapists had undertaken did not make any difference. Three judgements
were poor (k < 0.20), 10 fair(k = 0.21-0.40), and two moderate (k = 0.41-0.60)
EP
directional preference (Stanton et al., 2011). One study looked at the effect of
al., 2016)
ACCEPTED MANUSCRIPT
3.1.7 Quality of the studies
Two authors (SM, NR) independently rated the quality of the 10 RCTs against
the PEDro quality scale, and of 12 cohort studies against the quality scale
(Hartvigsen et al., 2015). There was 97% and 84% agreement between raters
PT
(85 / 88 agreements; 126 / 150 agreements) respectively. Kappa values
RI
between the two authors were respectively 0.92 and 0.78, indicating excellent
SC
to good levels of reliability in the two judgements (Altman, 1991).
U
Rated against the PEDro quality scale it was concluded that four RCTs were
AN
low, three RCTs were moderate and three RCTs were high quality (Table 3). It
was concluded that the quality of the 12 cohort studies were as follows: six
M
were low, one was moderate and five were high quality (Table 4). Regarding
D
the effect of quality, in the RCTs four of the six moderate and high quality trials
TE
had positive outcomes for the MDT groups. In the cohort studies two of the six
EP
moderate and high quality analyses had positive outcomes for the MDT
C
groups; but in the others a clear dichotomous comparison was not possible.
AC
4. Discussion
The present and previous review (May and Aina, 2012) bring together over 100
probably make them the most referenced clinical responses exposed during
ACCEPTED MANUSCRIPT
routine physical examination of specific and non-specific spinal patients. The
focuses in the previous review were on the definitions used for centralization,
prognostic indicators. There was limited evidence for them as treatment effect
PT
modifiers, variable evidence for the reliability of assessment of centralization,
RI
evidence for extension movement as the most common directional preference,
SC
and some evidence of a link between centralization and discogenic problems.
Some of the focuses in the present review were similar, but study designs were
U
different between the two reviews. The previous review contained 36 studies
AN
(N = 7113 patients) from which prevalence data could be extracted, whereas
M
prevalence of centralization was very similar, about 40%, whereas there was a
TE
present estimate is more robust. In the present review there was also a much
AC
(60-70%) of patients.
PT
In the previous review symptom duration was a definite determinant of
RI
centralization; with 77% prevalence in acute patients (N = 317), and about 40%
SC
in patients with chronic and mixed duration symptoms (4305). In the present
review very few patients with acute / sub-acute LBP were included (250); all
U
AN
the other studies reported mixed or chronic duration of symptoms. So it is not
possible to make any present judgement about the role of symptom duration
M
preference had limited evidence. However, the latter had some evidence, and
C
review there was some evidence that centralization and directional preference
were positive prognostic indicators in eight out of nine studies, although only
In the previous review reliability of the MDT assessment process had been
evaluated by six studies, and found to be very variable, mostly from moderate
PT
to good (Altman, 1991). In this review only one high quality study (Werneke et
RI
al., 2014) had evaluated reliability of several components of the MDT
SC
assessment process, including centralization and directional preference, and
U
AN
therapists are not reliable at classifying sub-groups that are purported to
determine management.
M
and centralization has been pretty consistent. Most patients appear to respond
TE
forces.
C
There are some differences from the previous review. There were 30 cohort
AC
seven criterion validity studies, six reliability, two surveys and one mini case
series - a total of 62 studies (May and Aina, 2012). In the present review there
studies, four cross-sectional studies, and ten case studies. It seems very
ACCEPTED MANUSCRIPT
surprising that at this distance from the foundation and development of the
McKenzie Method that virtually a quarter of the recent published articles are
PT
amongst practitioners is still so untrustworthy. If even trained therapists
RI
cannot agree, then this is a major problem, as MDT is a practitioner-led
SC
classification system that leads to management strategies.
In the previous review there were several studies that appeared to link
U
AN
centralization with discogenic pain, although heterogeneous definitions of
only one study attempted to explore this link, and found that type of disc
TE
lesions, such as, whether contained or extruded, were not associated with
EP
whereas the earlier study used provocation discography (Laslett et al. 2005), a
much more direct way to establish a link between pathology and symptoms. In
addition there was one case series and two cohort studies that evaluated the
PT
benefit of MDT. There is some indication for their therapeutic value, but no
RI
further evidence that they might be treatment effect modifiers. However there
SC
is still reasonably good evidence that both are a positive prognostic sign. In
U
AN
indicator of outcome, perhaps regardless of the applied management strategy.
suggesting possibly that MDT had some added value, but not with any clarity.
TE
5. Conclusion.
EP
This review has synthesised literature from 62 previous studies, but also
C
clinical response, they are still common and important prognostic indicators.
References
Aina A, May S, Clare H. The centralization phenomenon of spinal symptoms - a
PT
systematic review. Manual Therapy 2004;9:134-143.
Albert H, Hauge E, Manniche C. Centralization in patients with sciatica: are pain
RI
responses to repeated movement and positioning associated with outcome or
types of disc lesions? Eur Spine J 2011;21:630-636.
SC
Albert H, Manniche C. The efficacy of systematic active conservative treatment
for patients with severe sciatica. A single-blind, randomized, clinical controlled
trial. Spine 2012;37:531-542.
U
AN
Al-Obaidi SM, Al-Sayegh NA, Nakhi HB, Skaria N. Effectiveness of McKenzie
interventions in chronic low back pain: A comparison based on centralization
phenomenon utilizing selected bio-behavioral and physical measures. Int J
M
Altman DG. 1991. Practical Statistics for Medical Research. London: Chapman
D
& Hall.
TE
Chorti AG, Chortis AG, Strimpakos N, McCarthy CJ, Lamb SE. The prognostic
value of symptom responses in the conservative treatment of spinal pain. A
systematic review. Spine 2009;34:2686-2699.
Delitto A, Cibulka MT, Erhard RE, Bowling RW, Tenhula JA. Evidence for use of
an extension-mobilization category in acute low back syndrome: a prescriptive
validation pilot study. Physical Therapy 1993;;73:216-222.
ACCEPTED MANUSCRIPT
Desai MJ, Padmanabhan G, Simbasivan A, Kamanga-Sollo GG, Dharmappa A.
Directional preference following epidural steroid injection in three patients
with acute cervical radiculopathy. Pain Practice 2012;13:559-565.
PT
Orthop Sports Phys Ther 2014;44:68-75.
Elenburg JL, Foley BS, Roberts K, Bayliss AJ. Utilization of mechanical diagnosis
RI
and therapy (MDT) for the treatment of lumbar in the presence of known
lumbar transverse process fractures: a case study. J Man Manip Ther
SC
2016;24:74-79.
U
low back pain population using a combined McKenzie and patho-anatomical
AN
assessment. Man Ther 2016;26:201-207.
Garcia AN, Costa LdaCM, da Silva TM, Gondo FLB, Cyrillo FN, Costa RA, Costa
TE
Garcia AN, Costa LdaCM, Hancock M, Costa LOP. Identifying patients with
chronic low back pain who respond best to Mechanical Diagnosis and Therapy:
C
2016;96:623-630.
Gregg CD, McIntosh G, Hall H, Hoffman CW. Prognostic factors associated with
low back pain outcomes. J Prim Health Care 2014;6:23-30.
ACCEPTED MANUSCRIPT
Halliday MH, Pappas E, Hancock MJ, Clare HA, Pinto RZ, Robertson G, Ferreira
PH. A randomized controlled trial comparing the McKenzie Method to motor
control exercises in people with chronic low back pain and a directional
preference. J Orthop Sports Phys Ther 2016; 46:514-522.
Hagovska M, Takac P, Petrvicova J. Change in the muscle tension of erector
spinae after the application of the McKenzie Method in patients with chronic
low back pain. Phys Med Rehab Kuror 2014;24:133-140.
PT
Hartvigsen L, Kongsted A, Hestbaek L. Clinical examination findings as
prognostic factors in low back pain: a systematic review of the literature. Chiro
RI
& Man Ther 2015;23:13.
Hayden JA, van der Windt DA, Cartwright JL, Cote P, Bombardier C. Assessing
SC
bias in studies of prognostic factors. Ann Intern Med 2013;158:280-286.
Heintz MM, Hegedus EJ. Multimodal management of mechanical neck pain
U
using a treatment based classification system. J Man Manip Ther 2008;16:217-
224.
AN
Hosseinifar M, Akbari M, Behtash H, Amiri M, Sarrafzadeh J. The effects of
stabilization and McKenzie exercises on transverse abdominis and multifidus
M
provocation discography results in chronic low back pain, and the influence of
disability and distress on diagnostic power. Spine J 2005;5:370-380.
Long AL. The centralization phenomenon. Its usefulness as a predictor of
EP
PT
Spinal Publications, New Zealand, 2003.
Moncelon S, Otero J. The McKenzie Method of mechanical diagnosis and
RI
Therapy in chronic low back pain with directional preference. Kinesither Rev
2015;15:31-37.
SC
Muir Gray JA. Evidence-based Healthcare. How to Make Health Policy and
Management Decisions. Churchill Livingstone, New York, 1997.
U
Murphy DR, Hurwitz EL1. Application of a diagnosis-based clinical decision
AN
guide in patients with low back pain. Chiro & Man Ther 2011;19:26.
guide in patients with neck pain. Chiro & Man Ther 2011;19:19.
Olivio SA, Macedo LG, Gadotti IC, Fuentes J, Stanton T, Magee DJ. Scales to
D
Ther 2008;88:156-175.
Ojha H, Egan W, Crane P. The addition of manipulation to an extension-
EP
oriented intervention for a patient with chronic LBP. J Man Manip Ther
2013;21:40-47.
C
PT
manipulation: a stratified analysis in a randomized controlled trial. BMC
Musculoskel Dis 2015;16:74.
RI
Robinson M. Clinical diagnosis and treatment of patient with low back pain
using the patient response model: a case report. Physio Theory Pract
SC
2016;32:315-323.
U
outcomes for patients with centralising versus non-centralising neck pain. J
AN
Man Manip Ther 2016;24:264-268.
Sanderson S, Tatt ID, Higgins JPT. Tools for assessing quality and susceptibility
M
2007;147:224-233.
Skyte L, May S, Petersen P. Centralization: its prognostic value in patients with
EP
Surkitt LD, Ford JJ, Chan AYP, Richards MC, Slater SL, Pizzari T, Hahne AJ.
Effects of individualised directional preference management versus advice for
reducible discogenic pain: a pre-planned secondary analysis of a randomised
controlled trial. Man Ther 2016;25:69-80.
Van Helvoirt H, Apeldoorn AT, Knol DL, Arts MP, Kamper SJ, van Tulder MW,
Ostelo RW. Transforaminal epidural steroid injections influence Mechanical
Diagnosis and Therapy (MDT) pain response classification on candidates for
PT
lumbar herniated disc surgery. J Back Musculoskel Rehab 2016; 29:351-359.
Van Helvoirt H, Apeldoorn AT, Ostelo RW, Knol DL, Arts MP, Kamper SJ, van
RI
Tulder MW. Transforaminal epidural steroid injections followed by Mechanical
Diagnosis and Therapy to prevent surgery for lumbar disc herniation. Pain Med
SC
2014;15:1100-1108.
U
chronic pain or disability. Spine 2001;26:758-765.
AN
Werneke MW, Hart DL, George SZ, Deutscher D, Stratford PW. Change in
psychosocial distress associated with pain and functional status outcomes in
M
PT
Yarnbowicz R, Tao M, Owens A, Wlodarski M, Dolutan J. Pain pattern
classification and directional preference are associated with clinical outcomes
RI
for patients with low back pain. J Man Manip Ther 2018;26:18-24.
SC
Young JL, Rhon DI, Cleland JA, Snodgrass SJ. The influence of exercise dosing on
outcomes in patients with knee disorders: a systematic review. J Orth Sports
Phys Ther 2018;48:146-161.
U
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
First author Study design Population Participants Intervention: Follow-up: Outcomes - Results - only
PT
and date / Purpose from which after MDT / DP & weeks (w), clinical or SD between
participants inclusion / control OR months MDT-related groups
RI
were exclusion classification (m), year reported
recruited criteria (y)
SC
Albert & Randomised 477 with 181 Symptom- 2m, 1y (93- Global DP: Global
Manniche controlled sciatica randomized guided 95%) RMDQ (<0.008); NR
2012 trial (RCT) referred to (acute- exercises (DP) + Leg pain some SD
U
back centre chronic) stabilization NR signs
AN
exercises .V. EQ-5D
sham exercises Sick leave
M
Albert 2012 Prospective See above 181: 165 MDT Ax 3m, 1y Cent. Cent. 25.5%
cohort: (91%) back Decrease 44%
D
2ndary RCT / & referred NB 16%
TE
types of disc pain Peripheralizat Non-Cent. 7%
lesions ion 8%
EP
related to ISQ Types disc
pain MRI lesions not
C
Cent.
Al-Obaidi Prospective 297: 193 105 CLBP: MDT 5w, 10w Pain Cent. pain with
2013 cohort / Cent. eligible 62 Cent. / 43 Fear- activities
v partial Cent. partial Cent. avoidance (<0.001); NS
ACCEPTED MANUSCRIPT
PT
Apeldoorn Prospective LBP + leg pain 114 acute- MDT Ax Ax only DP with Cent. 51 (45%)
2016 cohort / test- chronic LBP DP no Cent. 23 (20%)
RI
retest No DP 40 (35%)
changes DP with Cent.
SC
spinal control better spinal
after Ax control (<0.02)
U
Bonnet 2011 RCT LBP 54 LBP MDT 1w Cent. 62% v 17%
AN
Guideline- (0.008)
based group Disability
Pain
M
Desai 2012 Case studies / NP with 3 acute- MDT Ax after 2w DP only after Full resolution
effect of ESI cervical chronic ESI ESI of symptoms
D
on DP radiculopathy
TE
Edmond Retrospective Convenience 328 NP Classified as: Discharge Function Cent. 40%
2014 cohort / sample with acute- Cent. + DP, Pain DP 70%
EP
FOTO data chronic Non-Cent + DP, Cent. or DP:
classified as Non-Cent + function
C
Cent. (NS)
Elenburg Case study / Not given 1 LBP with MDT Ax & 1m Function Almost full
2016 MDT despite lumbar treatment Pain resolution
spine fractures
fractures
ACCEPTED MANUSCRIPT
Flavell 2016 Prospective 316: 197 150 CLBP MDT Ax Ax only Cent. / Periph 32%
cross- (62%) (76%) Dysfunction 36%
sectional / Other 31%
PT
classification Postural 1%
systems
RI
Franz 2017 Pragmatic 47 consenting 44 military MDT (DP) (22) 3m Pain DP: pain, PGE,
controlled consecutive LBP Usual care (22) Disability disability 3m
SC
study LBP PGE (<0.05)
HC
U
Stability DP
AN
Garcia 2013 RCT 182 CLBP 148 CLBP MDT (DP) 1m, 3m, 6m Pain MDT: disability
(81%) Back school (BS) (99-100%) Disability 1m (0.004); all
QoL other (NS)
M
Garcia 2016 Prospective 140 / 148 Baseline 1m Pain Older age
cohort: (95%) DP characteristics Disability (0.01).
D
2ndary RCT / DP Cent. leg pain
TE
better high pain (NS)
responders
EP
DP
Gregg 2014 Retrospective Consecutive 1076 LBP Hall 6m Pain pain factors
C
PT
study / effect 24 no-LBP controls Cent. Cent. 100% 1m
Cent. controls EMG
RI
Halliday 2016 RCT 133 70 (53%) MDT 8w Pain MDT: GPE
consented CLBP with MCE GPE (0.03)
SC
DP Function
Muscle
U
Heintz & Case study / Not given 1 NP TBC 6w Pain Cent. with
AN
Hegedus use of TBC Disability mobilisation
2008 ROM
Hosseinifar RCT 75:41 (55%) 37 (90%) MDT Discharge Pain
M
2013 MCE Disability MCE (<0.05)
D
Muscle
Lopez-Diaz RCT Not given 30 LBP Mobilization Discharge Pain Mobilization:
TE
2015 Modalities ROM Cent. (<0.001)
Cent.
EP
Mazzone Cross- Not given 18 LBP and Spine Ax only MSI CPR 100%
2016 sectional / 17 no LBP kinematics in subgroups:
C
PT
classification provocation
NR 24%
RI
Myofascial 10%
Murphy Prospective Consecutive 95 acute- According to Not Red flags 1%
SC
20112 cohort / NP in one chronic NP classification recorded Cent. 27%
DBCDG year Pain 69%
U
classification provocation
AN
NR 19%
Myofascial 22%
Ojha 2013 Case study / 2 Not given 1 CLBP TBC = DP + 7w Disability
M
categories manipulation ROM
TBC
D
Otero 2014 Prospective Consecutive 349 patients MDT Discharge Classification 92% Der.
TE
cohort / MDT with LBP Cent. 71% / 76% at
syndromes, discharge
EP
Cent. DP DP 73%
Stability MDT 90%
C
Otero 2016 Prospective Consecutive 297 patients MDT Discharge Classification 92% Der.
AC
PT
2011 (27%) Cent. or Manipulation (93%) Pain Disability (0.02,
Peripheral- GPE 0.03).
RI
ization QoL Cent./Periph.
Satisfaction (NS)
SC
Further HC
Petersen 2ndary RCT / Not given 350 LBP as above Age MDT: NR +
U
2015 factors effect Duration Periph. (RR
AN
related to modifiers Pain variables 10.5)
positive NR
outcome in
M
RCT above
Robinson Case study Not given 1 sub-acute DP 4w Pain
D
2016 with DP LBP Disability
TE
ROM
MRI
EP
Rose 2016 Retrospective Not given 11 NP Cent. (6) Discharge Disability Cent: Disability:
cohort / Cent. Non-Cent (5) Cent. (0.005)
C
v non-Cent.
AC
Stanton 2011 Cross- 545 LBP > 90 250 acute or Testing out Ax only Manipulation 42%
sectional days subacute algorithm Stabilization 17.5%
study / LBP criteria for 4 DP 31%
Prevalence & sub-groups Traction: 9.5%
reliability TBC + 1 subgroup 25%
ACCEPTED MANUSCRIPT
Kappa 0.52
Surkitt 2016 2ndary RCT 2038 CLBP + 78 met DP v Guideline- 5w, 10w, Pain DP: back pain
(Ford et al., leg pain DP + criteria based advice 26w, 52w Function 10w (0.003)
PT
2016) / Discogenic* Psychosocial
discogenic* General
RI
sub-group health
Takasaki Case study / Not given 1 LBP with MDT 1m Pain Cent. & disc
SC
2010 effect on disc MRI Disability displacement
MRI resolved
U
Takasaki Case study / NP 1 NP MDT Discharge CCFT CCFT negative
AN
2016 effect on after Cent.
CCFT
van Helvoirt Prospective 132 referred 69 non-Cent Transforaminal 2w, 1y Resolved 16%
M
2014 cohort / for HLDS HLDS epidural steroid Cent. 46%
D
effect of TESI candidates injection (TESI) Non-Cent / B 16%
Surgery 22%
TE
van Helvoirt Prospective 132 referred 77 non-Cent. TESI 1y Leg pain* Surgery v non-
2016 cohort / for HLDS HLDS Disability* surgery *
EP
2ndary above candidates GPE (0.001); Cent. v
different Back pain^ non-Cent *^
C
PT
Level training
Werneke Retrospective 2066 LBP 723 for who MDT 1m Pain Cent. and DP
RI
2016 cohort / MDT, selected from complete Functional added little to
Cent. DP as FOTO data status predicting
SC
prognostic database Psychological outcomes
factors distress
U
Werneke Prospective LBP high 138 LBP DP v non-DP Discharge Disability DP (65%)
AN
2018 cohort / DP & STarT risk Other variables Psychological disability (0.03)
STarT from FOTO (pain, function, distress
MDT training)
M
Williams Case study Not given 1 discogenic MDT 2m Disability Resolution with
2011 with lateral LBP Cent.
D
component
TE
Wu 2018 Case studies Not given 3 CLBP with MDT <2m Prostate Complete
DP with LUTS LUTS Symptom resolution
EP
Index
Yarnbowicz Prospective 1006 LBP 940 initial DP Cent. Discharge Pain Cent. 20%
C
2018 cohort / Cent. consecutive 639 full data No DP no-Cent. Function Non-Cent. 39%
AC
2ndary = secondary analysis of previous study; Ax = assessment; CLBP = chronic low back pain; CCFT = Cranio-cervical flexion test; CPR = clinical prediction
rules; DBCDG = diagnosis-based clinical decision guide; Der. = Derangement; EMG = electromyography of erector spinae muscle activity; ESI = epidural
steroid injection; FOTO = Focus on Therapeutic Outcomes; GPE = Global Perceived Effect; HADS = Hospital Anxiety Depression Scale; HC = healthcare;
PT
HLDS = herniated lumbar disc surgery; HE = healthcare; LUTS = lower urinary tract symptoms; MCE = motor control exercises; MDT = Mechanical
Diagnosis and Therapy or the McKenzie Method; MSI = movement system impairment; NR = nerve root; NS = no significant difference; PT = physical
RI
therapists; QoL = Quality of Life; RCT = randomized controlled trial; ROM = range of movement; RR = relative risk; STarT = subgroups for targeted
treatment back screening tool; TBC = treatment based classification system.
SC
*Discogenic = at least 4 of: Back +leg pain; sitting limited to 60 minutes; forward bending somewhat difficult; lifting somewhat difficult; sit to stand
somewhat difficult; coughing or sneezing somewhat difficult; symptoms worse the next day; working on manual job; onset associated with flexion /
U
rotation and/or compression loading.
AN
*Hall classification = four sub-groups based on site of symptoms and DP; and fifth sub-group with heightened pain behaviours (Gregg et al. 2014)
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
PT
Duration Symptoms N (%)
RI
Sub-acute LBP 62 / 123 50%
SC
Chronic LBP 227 / 567 40%
U
Mixed LBP 1584 / 3738 42%
AN
Neck pain 62 / 168 37%
M
TOTAL 2109 / 4745 44%
Summary of previous studies - directional preference (N = 5), (May and Aina, 2012)
D
TE
TOTAL EP 1661 / 2368 70%
Edmond (2014) Mixed NP 328 40% 30% 30% (Ext. 80%, Flex. 10%, Lat. 10%)
PT
Garcia (2016) Chronic LBP +/- 148 95% 5% (Ext. 50%, Flex. 5%, Lat. 40%)
RI
Halliday (2016) Chronic LBP 133 73% 27% (Ext. 86%, Flex. 5.5%, Lat. 8.5%)
SC
Mazzone (2016) Chronic LBP 17 47% 53%
U
Murphy (2011)1 Chronic LBP +/- 264 41%
AN
Murphy (2011)2 Chronic NP +/- 95 27%
M
Otero (2014) Mixed LBP 349 76% 16% 8% (Ext. 80%, Flex. 4%, Lat. 13%)
Otero (2016) Mixed NP 297 82% 10% 8% (Ext. 84%, Flex. 3%, Lat. 14%)
D
TE
Petersen (2011) Chronic LBP +/- 350 53% 47%
PT
N44
= missing numbers; the superscript number is the discrepancy between total and accounted for
RI
SC
Duration Symptoms N Cent. DP No DP UC
U
TOTAL Mixed LBP 2655 975 788 873 19
AN
Sciatica 265 104 108 40 13
M
Sub-acute LBP 250 77 173
D
TE
Mixed Neck pain 720 401 128 191
Table 3. PEDro quality scale for randomised controlled trials (N = 10) (3 / 88 disagreements) 97% agreement
PT
RI
Albert 2012 √ √ X √ X X
√ √ √ √ √ 7 Moderate
Bonnet 2011 √ √ X √ X X
X √ X √ X 4 Low
SC
Franz 2017 X X X √ X X
X √ X √ √ 5 Low
Garcia 2013 √ √ √ √ X X
√ √ √ √ √ 8 High
U
Hagovska 2014 X X X X X X
X √ X √ √ 3 Low
AN
Halliday 2016 √ √ √ √ X X
√ √ √ √ √ 8 High
Hosseinifar 2013 √ √ X √ X X
√ X X √ √ 5 Moderate
M
Lopez-Diez 2015 √ √ √ √ √ √ √ √ X √ √ 9 High
Moncelon 2015 √ X X X X X √ √ X √ √ 4 Low
D
Petersen 2011 √ √ √ √ X X √ X √ √ √ 7 Moderate
TE
PEDro scores: 1. Eligibility criteria were satisfied; 2. Subjects randomly allocated to groups; 3. Allocation was concealed; 4. Groups similar at baseline
EP
regarding most important prognostic indicators; 5. Blinding of subjects; 6. Blinding of all therapists; 7. Blinding of all assessors; 8. Measures of jey
outcomes were obtained from more than 85% of those initially allocated to groups; 9. All subjects for who outcome measures were available received the
C
treatment or the control as allocate, or where this was not the case, data were analysed by "intention to treat"; 10. Reports of between-group statistical
AC
comparison were reported for at least one key outcome; 11. Study provided both point measures and measures of variability for at least one outcome
measure. Score is out of ten item is not included.
ACCEPTED MANUSCRIPT
PT
Author 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Total (15) Quality
RI
Albert 2012 √ √ √ √ √* √* √* √ √ √ √ √* √ √ √ 15 High
Al-Obaidi 2013 √ √ √ X X √ X √ √ √ √ √ √ X √ 11 High
SC
Edmond 2014 X X √ X X X √ √ √ √ √ X √ √ X 8 Moderate
Garcia 2016 √ √ √ √* √ √ √ X X √ √ √ √ √ √ 13 High
Gregg 2014 √ X √ X X X X √ X √ √ X X √ X 6 Low
U
Petersen 2015 √ √ √ X X √ √ √ X √ √ √ √ √ √ 12 High
AN
Rose 2016 X X X X √ √ √ X X √ √ X X X √ 6 Low
Surkitt 2016 √ √ X √ √ √ √ X X √ √ √ √ X √ 12 High
van Helvoirt 2014 √ √ √ √ √ √ X X X √ √ X X X X 8 Low
M
Werneke 2011 X X √ X X X X X X √ √ X √ √ √ 6 Low
Werneke 2016 X X √ X X X X √ √ √ √ X √ √ √ 8 Low
D
Werneke 2018 X X √ X X X X X X √ √ X √ √ √ 6 Low
TE
* with information from the accompanying study (Albert & Manniche 2012; Garcia et al. 2014; Petersen et al. 2011)
Quality items (from Hartvigsen et al. 2015): 1. Study population clearly defined; 2. Study population described; inclusion / exclusion criteria / chronicity; 3.
EP
Study population represent population of interest**; 4. Completeness of follow-up described for each point of follow-up to one year**; 5. Completeness
of follow-up adequate - 85%**; 6. Reasons for loss to follow-up adequately described; 7. No important differences (characteristics and outcomes)
C
between completers and non-completers; 8. Prognostic tests defined enough to be replicated; 9. Performance of prognostic tests are standardised; 10.
AC
Outcomes are defined; 11. Outcomes well established; 12. Method, setting, timing outcomes same for all participants; 13. Data presented sufficiently to
assess adequacy of analysis; 14. Statistical analysis sufficiently described and appropriate to account for other prognostic factors, such as multivariate
analysis**; 15. No selective reporting of results.
** these items were slightly amended from the original as described in the text.
ACCEPTED MANUSCRIPT
Figure 1. Flow chart of selection process of studies
PT
Abstracts excluded (N = 2145)
RI
Duplicates removed (N = 255)
U SC
Additional records through
hand searching reference
AN
lists and mckenziemdt.org
website (N = 15)
M
PT
modifiers was absent.
RI
U SC
AN
M
D
TE
C EP
AC