Professional Documents
Culture Documents
L
ow back pain (LBP) is the worldwide leading cause of years lived A variety of clinical practice guide-
with disability, with an estimated point prevalence of 9.4% and lines have been developed for the treat-
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
a lifetime prevalence of up to 39%.25,52,62 This negatively impacts ment of LBP.6,29,43 These guidelines
propose a shift away from treatment of
the psychosocial health of those affected.48 Moreover, with an
LBP primarily based on pathoanatomi-
aging population, LBP is expected to become more widespread.26 cal principles in favor of a classification-
based approach. This suggestion is
UUSTUDY DESIGN: Literature review with meta- patients with acute LBP, there was no significant largely based on several studies report-
analysis. difference in pain resolution (P = .11) and disability ing that classifying patients led to im-
UUBACKGROUND: The McKenzie Method of
(P = .61) between MDT and other interventions. In proved clinical results.14,15,31 However, a
patients with chronic LBP, there was a significant
Mechanical Diagnosis and Therapy (MDT), a recent review has questioned the clinical
difference in disability (SMD, –0.45), with results
classification-based system, was designed to effectiveness of subgrouping claims, due
Journal of Orthopaedic & Sports Physical Therapy®
1
Physiotherapy Department, Faculty of Medicine and Health Science, Sherbrooke University, Sherbrooke, Canada. 2Physiotherapy at Concordia Physio Sport, Montreal, Canada.
3
Physiotherapy at Physio Multiservices, Chateauguay, Canada. 4Physiotherapy private practice, Saint-Laurent, Canada. 5Centre for Interdisciplinary Research in Rehabilitation,
Constance Lethbridge Rehabilitation Centre, and the School of Physical and Occupational Therapy, McGill University, Montreal, Canada. The Edith Strauss Rehabilitation
Research Project at McGill University provided grants to support its authors. The Edith Strauss Rehabilitation Research Project of McGill University took no part in the design,
implementation, analysis, or production of the manuscript for this meta-analysis. The authors certify that they have no affiliations with or financial involvement in any organization
or entity with a direct financial interest in the subject matter or materials discussed in the article. Address correspondence to Olivier Tri-Thinh Lam, 7985 Salomon, Brossard,
Quebec, Canada J4X 1J2. E-mail: olivierlam.qc@gmail.com t Copyright ©2018 Journal of Orthopaedic & Sports Physical Therapy®
T
directional preference of a patient is the he methodology for this review search was performed on May 26, 2016,
direction in which a repeated movement was based on the PRISMA state- and a third search was performed on Sep-
and/or sustained position produces ment,39 and the data extraction form tember 6, 2017 to provide an update of
an improvement in symptoms. Those was informed by the Cochrane meta- articles published since the first search.
improvements may include centraliza- analysis guidelines.27 Additionally, references from the includ-
Downloaded from www.jospt.org at Grand Valley State University on April 12, 2018. For personal use only. No other uses without permission.
manifest differently, the treatment ef- (κ = 0.7-0.9) than are therapists without
fect could be different. 19,44,51 A cutoff certification (κ = 0.17-0.39).28,49,65 Stud- Data Extraction
of 12 weeks to differentiate acute from ies in which an MDT classification was Data extraction was performed by 2 in-
chronic LBP has been used in previous not completed prior to the treatment vestigators (P.T.P., M.C.F.), who each
systematic reviews and clinical prac- were excluded, as a priori classification independently extracted the data from
tice guidelines. 4,37 Also, the previous is an essential characteristic of the MDT all studies with the use of an extraction
meta-analysis compared MDT to pas- approach.36 Last, the comparator inter- form. A customized data extraction form
sive therapy, which included a variety of vention had to be a typical rehabilitation was developed for each of the 2 outcomes
interventions that might have different intervention, such as manual therapy, ex- of interest, pain and disability. The data
effects. Because the relative effective- ercise, or education. There was no review extraction form was a Microsoft Ex-
ness of MDT could change based on the protocol published for this meta-analysis. cel spreadsheet designed according
comparator intervention, MDT should to the Cochrane meta-analysis guide-
be compared to each intervention type Information Sources lines and adjusted to the needs of this
separately. The level of MDT training Six electronic databases (MEDLINE, meta-analysis.27
should also be considered, as it may Embase, CINAHL, Cochrane Database The following information was ex-
impact interventions and risk-adjusted of Systematic Reviews, PsycINFO, and tracted from each study: (1) charac-
functional outcomes. 10 The objective the Physiotherapy Evidence Database teristics of the study (study duration,
of this meta-analysis was to determine [PEDro]) were searched using 3 pri- therapist MDT training, and the number
the effectiveness of MDT provided by mary search strings: (1) MDT therapy, of patients allocated to each group) and
trained therapists compared to that of (2) low back/lumbar pain, and (3) ran- inclusion criteria, (2) type of intervention
different types of comparator interven- domized controlled trials. Related terms (including duration and frequency of the
tions for improving pain and disability were included for each search string, different interventions), and (3) type of
in patients with acute and chronic LBP and an example for the MEDLINE outcome measures (including pain scores,
separately. search is provided (APPENDIX, available at disability scores, definitions and time of
subdivision of other interventions. Other Assessment, Development and Evalua- tervention that was considered to con-
interventions were defined as nonsurgi- tion (GRADE) approach was used to as- tribute most (eg, manual therapy) was
cal and noninvasive interventions within sess the quality of the body of evidence included in the primary analysis. How-
the scope of physical therapy practice (eg, for each outcome of this meta-analysis ever, in these cases, a sensitivity analysis
exercise, manual therapy, and education). (pain and disability).27 This evaluation was completed where the comparator
These interventions could be performed was conducted by 2 raters (D.S., P.T.P.), groups were substituted. Both compara-
by physical therapists or other health and a third reviewer (O.L.) made the tor groups could not be included in the
professions. Other interventions were final decision if a consensus could not same analysis to avoid artificially inflat-
further subdivided into manual therapy, be reached. The quality of evidence was ing the sample size. When medians and
exercise, a combination of manual ther- initially considered “high” and could be interquartile ranges (first and third)
apy and exercise, or education. Chronic downgraded based on the following 5 were provided, means were calculated by
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
LBP was defined as pain in the lumbar factors: (1) limitation of design, (2) in- summing the median, first interquartile
spine lasting more than 12 weeks. Acute directness of evidence, (3) inconsistency range, and third interquartile range and
LBP was defined as having a duration of results, (4) imprecision of results, and then dividing by 3. Standard deviation
of pain less than 12 weeks. After hav- (5) high probability of publication bias. estimates were calculated from inter-
ing completed the extraction process, Studies that did not reach a score of 5 on quartile values and consideration of the
the investigators compared results and the PEDro scale could be downgraded study sample size.63
reached consensus on any discrepancies. for a limitation of design41; studies that
A third investigator (S.R.) resolved dis- possessed differences in populations, in- RESULTS
agreements if a consensus could not be terventions, outcome measures, and in-
T
reached. Once the extraction form was direct comparisons could be downgraded he literature search resulted
Journal of Orthopaedic & Sports Physical Therapy®
completed, the 2 investigators indepen- for indirectness; studies with effect esti- in the identification of 758 publi-
dently tested the form with the first 3 mates that were heterogeneous could be cations, 678 from databases and
included studies. The results were then downgraded for inconsistency; and stud- 80 from reference lists (FIGURE 1). After
compared to ensure uniformity of the ies that had fewer than 400 participants removing duplicates, 2 independent
extraction process. When relevant data could be downgraded for imprecision. reviewers screened 354 abstracts and
were missing from a study, the authors selected 51 articles for full-text review.
and coauthors were contacted via e-mail Statistical Analysis After review, 17 articles were retained for
to request the missing information. If Analyses were completed separately for the meta-analysis; however, of these 17
the data could not be obtained, the study patients with acute and chronic LBP. The studies, 4 did not provide sufficient data
was excluded from the analyses. For each effectiveness of MDT compared to other to be included in the statistical analy-
study, pain and disability measures were interventions, subdivisions of other in- ses. These 4 studies are summarized in
extracted immediately after the MDT in- terventions, or placebo were examined TABLE 1.1,20,46,53 No significant between-
tervention or the comparison interven- using random-effects models with sta- group differences were observed in pain
tion, when the intervention was assumed tistical significance set at P<.05.8,67 The and disability in 3 and 4, respectively, of
to have the largest treatment effect. standardized mean difference (SMD) and the 4 studies excluded from the meta-
95% confidence interval (CI) were calcu- analysis.1,20,46,53 Attempts to contact the
Risk of Bias and Strength of Evidence lated for each analysis. Random-effects authors to provide additional data were
To evaluate risk of bias in individual stud- models were utilized, because it was ex- not successful. One study that met the
ies, the methodological quality of the in- pected that there would be heterogene- inclusion criteria was excluded from
cluded studies was rated on the PEDro ity of the comparator interventions. The data analysis, because participants who
scale.34 The PEDro scale has demonstrat- heterogeneity among studies was deter- were noncentralizers post randomization
ed acceptable reliability for the overall mined using the chi-square statistic with were only excluded from the intervention
score (intraclass correlation coefficient = significance set at P<.10 and I2. These group.41 This could have biased the treat-
0.680)34 and validity.7 The ratings were analyses proceeded even if statistical het- ment effect toward the MDT group, as
group difference in improvement in pain included 2 comparator interventions.3 moderate evidence of no significant differ-
and disability favoring MDT should be When the education booklet was includ- ence (P = .36) in disability after the inter-
interpreted with caution.41 One study ed instead, no significant differences re- vention period between MDT and manual
with a mix of individuals with acute and mained (P = .16). therapy plus exercise (SMD, –0.24; 95%
chronic LBP45 was included in the data CI: –0.77, 0.28). Ratings were also down-
analyses for chronic LBP, because most Acute LBP: Subgroup Analysis graded because of imprecision of results.
participants had recurrent episodes of MDT Versus Manual Therapy Plus Ex- MDT Versus Exercise None of the in-
LBP. For 1 study, medians and inter- ercise Three studies compared MDT cluded studies compared MDT to exer-
quartile ranges were converted to means to manual therapy plus exercise.3,54,55 cise alone in participants with acute LBP.
and standard deviations, respectively, as Comparator interventions included spi- MDT Versus Education Two studies
described in the Methods.63 A summary nal manipulative thrusts with lumbar compared MDT to an intervention that
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
of the meta-analysis is shown in TABLE 2. range-of-motion exercises,54 joint mo- included only education in participants
bilizations,55 and manipulations with with acute LBP.3,33 In 1 study, education
Acute LBP: Primary Analysis home exercises.3 Only 2 of 3 studies were was described as “first line care,” and in-
of MDT Versus Other Interventions included in the pain intensity analysis.54,55 cluded advice to avoid bed rest and to
Four studies compared MDT to other Tests of heterogeneity were not signifi- remain active, assurance of a favorable
interventions in participants with acute cant (FIGURE 2B). There was moderate evi- prognosis, and advice to take acetamino-
LBP.3,33,54,55 The other interventions in- dence of a significant (P = .04) difference phen.33 This first-line care was provided
cluded spinal manipulative thrusts,
lumbar range-of-motion exercise,54 joint
MEDLINE, Embase, CINAHL, Cochrane, McKenzie Institute lumbar spine trials,
mobilizations,55 and first-line care (eg,
Journal of Orthopaedic & Sports Physical Therapy®
Summary of Inclusion and Exclusion Criteria,
TABLE 1
Intervention Groups, and Outcome Measures
Acute Pain
Study Participants (<12 wk) or
(PEDro (MDT Participants (Other Chronic Pain MDT Level
Score) Intervention)* Interventions)* Inclusion Criteria (>12 wk) Intervention of Training Outcomes
Bonnet et al1 n = 28; men, n = n = 26 Nonspecific LBP Mix MDT: directional-preference exer- Parts A and B Pain: visual analog
(7/10)† 17; women, n men, n = 12; women, with or without cises, can modify positions and/or scale
= 11; age, 48.8 n = 14; age, 45.9 ± radiation to lower add manual techniques Disability: Oswestry
± 4.75 y; mean 5.1 y; mean symptom extremity, ≥18 y Manual therapy plus exercise: active Disability Question-
symptom dura- duration, 49.2 mo of age mobilizations in weight bearing naire
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Acute Pain
Study Participants (<12 wk) or
(PEDro (MDT Participants (Other Chronic Pain MDT Level
Downloaded from www.jospt.org at Grand Valley State University on April 12, 2018. For personal use only. No other uses without permission.
Score) Intervention)* Interventions)* Inclusion Criteria (>12 wk) Intervention of Training Outcomes
Machado n = 73; men, n = n = 73; men, n = 38; wom- Acute nonspe- Acute MDT: first-line care, directional- Credentialed Pain: numeric rating
et al33 35; women, n en, n = 35; age, 45.9 ± cific LBP, pain preference exercises, postural scale
(8/10) = 38; age, 47.5 14.9 y; mean symptom between the 12th correction and education, Treat Disability: Roland-
± 14.4 y; mean duration, 67% <2 wk, rib and buttock Your Own Back book, lumbar roll, Morris Disability
symptom 33% 2-6 wk crease, with or home exercise program Questionnaire
duration, 66% without leg pain, Education: physician advice, Function: Patient-
<2 wk, 34% <6 wk in dura- acetaminophen; follow-up visit in Specific Functional
2-6 wk tion, preceded 3 wk, earlier if necessary Scale
by at least 1 mo Outcomes evaluated
without LBP in after 1 and 3 wk
which the patient
did not consult
a health care
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
practitioner,
18-80 y of age
Moncelon n = 7; men, n = n = 7; men, n = 5; women, Chronic nonspecific Chronic MDT: directional-preference exer- Parts A and B Disability: Oswestry
and 4; women, n n = 2; age, NA; symp- LBP, directional cises, home exercise program, Disability Question-
Otero40 = 3; age, NA; tom duration, NA preference, 18-70 pool therapy naire
(5/10) symptom dura- y of age Manual therapy plus exercise: Outcome evaluated
tion, NA; age of diaphragmatic breathing, after 1 wk
both groups, 47 lumbopelvic and coxofemoral mo-
± 11 y bilizations, paravertebral muscle
strengthening, pool therapy
Murtezani n = 111; men, n = n = 109; men, n = 42; Nonspecific LBP, Chronic MDT: directional-preference 50 h of Pain: visual analog
Journal of Orthopaedic & Sports Physical Therapy®
et al41 83; women, n = women, n = 67; age, pain between exercises, can add manual training, scale
(8/10)† 28; age, 48.8 ± 47.5 ± 8.8 y; symptom lower angle techniques, avoid motions that equivalent Disability: Oswestry
8.9 y; symptom duration, NA of scapulae peripheralize symptoms, home to 2 courses Disability Question-
duration, NA and above the exercise program naire
buttocks, with or Modalities: interferential current, Outcomes evaluated at
without leg pain ultrasound, heat 2 and 3 mo
or neurological
signs, >3 mo
duration, 18-65 y
of age
Paatelma n = 52; men, n = Education: n = 37; men, n Nonspecific LBP Mix MDT: exercises with or without Credentialed Pain: back and leg
et al45 37; women, n = 24; women, n = 13; with or without sustained end-range positions, pain, visual analog
(7/10) = 15; age, 44 ± age, 44 ± 15 y; symp- radiation to one manual techniques, education, scale
9 y; symptom tom duration, NA or both lower Treat Your Own Back book, home Disability: Roland-
duration, NA Manual therapy: n = 45; extremities, em- exercise program Morris Disability
men, n = 26; women, ployed, acute or Manual therapy plus exercise: Questionnaire
n = 19; age, 44 ± 10 y; chronic duration, high-velocity, low-amplitude Outcomes evaluated
symptom duration, NA 18-65 y of age thrust manipulation; specific after 3, 6, and 12
mobilizations; stretching; spinal mo
stabilization exercises; home
exercise program
Education: good prognosis of LBP,
pain tolerance and remaining
active, medication, early return to
work, booklet
Table continues on page 7.
Acute Pain
Study Participants (<12 wk) or
(PEDro (MDT Participants (Other Chronic Pain MDT Level
Downloaded from www.jospt.org at Grand Valley State University on April 12, 2018. For personal use only. No other uses without permission.
Score) Intervention)* Interventions)* Inclusion Criteria (>12 wk) Intervention of Training Outcomes
Petersen n = 132; men, n = n = 128; men, n = 72; LBP with or without Mix MDT: directional-preference exer- Credentialed, Pain: back and leg
et al46 70; women, n women, n = 56; median leg pain of >8 cises, can modify positions and/or parts A-D pain, Low Back Pain
(7/10)† = 62; median (10th, 90th percentiles) wk; radiograph, add manual techniques Rating Scale
(10th, 90th per- age, 35 y (24.0, 51.6 CT scan, or MRI Exercise: stationary bike and Disability: Low Back
centiles) age, y); median symptom taken within the low-resistance exercises for Pain Rating Scale
34.5 y (23.0, duration (10th, 90th preceding 2 y; lumbopelvic muscles, dynamic Outcomes evaluated
52.1 y); median percentiles), 14 mo (2.7, 18-60 y of age back strengthening exercises, after 2, 4 , and 12
(10th, 90th 113.5 mo) stretching trunk and hip muscles mo
percentiles) Both groups: asked to continue
symptom dura- exercising for a minimum of 2 mo
tion, 8 mo (2.0, after intervention
95.7 mo)
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Petersen n = 175; men, n= n = 175; men, n = 83; LBP, with or without Chronic MDT: directional-preference exercise, Screening Pain: numeric rating
et al47 72; women, n women, n = 92; age, 37 leg pain, >6 wk; no manual vertebral mobiliza- preran- scale
(7/10) = 103; age, 38 ± 9.4 y; symptom dura- able to speak tions, educational booklet and/or domization: Disability: Roland-
± 10.4 y; symp- tion, 94 ± 181 wk and understand lumbar roll at therapist discretion diploma Morris Disability
tom duration, Danish; clinical Manual therapy plus exercise: manu- Treatment: cre- Questionnaire
97 ± 230 wk signs of disc-re- al techniques at therapist discre- dentialed SF-36
lated symptoms; tion (eg, vertebral mobilization/ Outcomes evaluated
18-60 y of age manipulation), self-manipulation, after 3, 5, and 12
flexion/extension exercises and mo
stretching, educational booklet
Both groups: given stabilization/
strengthening exercises at
Journal of Orthopaedic & Sports Physical Therapy®
Downloaded from www.jospt.org at Grand Valley State University on April 12, 2018. For personal use only. No other uses without permission.
Acute Pain
Study Participants (<12 wk) or
(PEDro (MDT Participants (Other Chronic Pain MDT Level
Score) Intervention)* Interventions)* Inclusion Criteria (>12 wk) Intervention of Training Outcomes
Schenk n = 15; men, n = 7; n = 10; men, n = 8; Lumbar radiculopa- Acute MDT: directional-preference exercises Credentialed Pain: visual analog
et al55 women, n = 8; women, n = 2; mean thy: symptoms Manual therapy plus exercise: scale
(5/10) mean age, 40.1 age, 44.8 y; symptom originating in mobilization: passive movement Disability: Oswestry
y; symptom duration, 7 d to 7 wk disc, peripheral to spinal segments Disability Question-
duration, 7 d to to lumbar region, Both groups: postural correction, naire
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
according to grouping
Halliday n = 35; men, n = n = 35; men, n = 7; LBP localized Chronic MDT: directional-preference Credentialed Pain: visual analog
et al22 7; women, n = women, n = 28; age, between the exercises, postural education scale
(7/10) 28; age, 48.8 ± 48.3 ± 14.2 y; median 12th rib and the and lumbar roll, Treat Your Own Disability: Patient-
12.1 y; median symptom duration, 37.7 buttock crease, Back book Specific Functional
symptom dura- wk (IQR, 28.8) with or without Exercise: motor control exercises of Scale
tion, 26.6 wk referred pain deep lumbar stabilizers, home Outcomes evaluated
(IQR, 22.3) into one or both exercise program after 8 wk
legs and with or
without sensory
and or motor
changes, for >3
mo; directional
preference
Garcia et al18 n = 74; men, n = n = 73; men, n = 19; Chronic nonspe- Chronic MDT: directional-preference exer- Part A Pain: numeric pain-
(8/10) † 16; women, n = women, n = 54; age, cific LBP, pain cises, specific end-range motion rating scale
58; age, 57.5 ± 55.5 ± 13.7 y; symptom intensity of 3/10 exercise, postural education, Disability: modified
12.2 y; symp- duration, 48 ± 96 mo on a numeric home exercise program, and Treat Roland-Morris
tom duration, pain-rating scale, Your Own Back book Disability Question-
36 ± 102 mo 18-80 y of age, Placebo: detuned pulsed ultrasound, naire
and able to read detuned shortwave diathermy Outcomes evaluated
Portuguese Both groups: given educational after 5 wk and 3, 6,
booklet The Back Book and 12 mo
Abbreviations: CT, computed tomography; IQR, interquartile range; LBP, low back pain; MDT, Mechanical Diagnosis and Therapy; MRI, magnetic resonance
imaging; NA, not available; NSAID, nonsteroidal anti-inflammatory drug; PEDro, Physiotherapy Evidence Database; SF-36, Medical Outcomes Study 36-
Item Short-Form Health Survey; TENS, transcutaneous electrical nerve stimulation.
*Values are mean ± SD unless otherwise indicated.
†
Not included in meta-analysis.
with exercise as the comparator inter- cluded in the meta-analysis comparing of heterogeneity were significant (FIGURE
vention from the study that included 2 MDT to modalities (heat, ultrasound, 4B).45,47 There was moderate evidence of
no significant (P = .30) difference in pain
after the intervention period between
TABLE 2 Summary of Meta-analysis Results interventions (SMD, –0.26; 95% CI:
–0.73, 0.22). Ratings were downgraded
because of unexplained heterogeneity.
Number of Studies Mean Difference (95% CI) P Value
All 3 studies measured disability, and
Acute LBP
tests of heterogeneity were not signifi-
MDT versus other interventions
cant (FIGURE 5B).40,45,47 There was high-
Journal of Orthopaedic & Sports Physical Therapy®
were downgraded because of impreci- less than 5. Due to the nature of the inter- tively; therefore, there is good-quality
sion of results. These 4 studies also ex- ventions, the providers could not be blind- evidence showing that MDT is not clini-
amined disability. Tests of heterogeneity ed to the interventions in any of the studies, cally superior to other interventions in
were not significant (FIGURE 5C). There which lowered the PEDro scores of the acute LBP to improve pain or disability.
was high-quality evidence of a signifi- included articles. Blinding of the patients In patients with chronic LBP, (1) MDT
cant difference (P<.01) in disability after was reported in only 1 study.18 Blinding of was more effective at reducing pain and
the intervention period, with the results the assessor was reported in 9 of the stud- disability than other rehabilitation inter-
favoring MDT (SMD, –0.45; 95% CI: ies.1,17,20,22,31,33,41,47,55 The mean PEDro scale ventions, (2) MDT was superior to exercise
–0.64, –0.25. score for all studies was 6/10 (TABLE 1). for reducing disability but not pain, and
One of the included studies did not (3) MDT was not superior to combined
provide sufficient data to be included in DISCUSSION manual therapy and exercise, or to educa-
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
the meta-analysis; the authors found no tion. Although superior, the effect size was
I
significant between-group differences in n patients with acute LBP, (1) MDT small to moderate, indicating at least min-
change in pain and disability between pa- was no more effective than other inter- imal clinical significance. The strength of
tients treated with MDT versus exercise.47 ventions, (2) MDT yielded statistically evidence of these findings was moderate to
MDT Versus Education Only 1 study and clinically significant better improve- high and was downgraded mainly due to
compared MDT to an education inter- ments in pain intensity compared to significant heterogeneity between the in-
vention in participants with chronic LBP, manual therapy plus exercise (though cluded studies. The strength of evidence is
and thus a meta-analysis could not be only 2 studies with small sample sizes further demonstrated by the PEDro scale
completed.45 Education included advice were included in the analysis), and (3) scores higher than 5 for all studies contrib-
to remain active. There was no significant
Journal of Orthopaedic & Sports Physical Therapy®
The Back Book (Baltimore, MD: Johns Heterogeneity: τ = 0.13, χ = 4.39, df = 2 (P = .11), I = 54%.
2 2 2
ence between groups for change in pain Heterogeneity: τ = 0.09, χ = 1.48, df = 1 (P = .22), I = 32%.
2 2 2
intensity at the end of treatment, with re- Test for overall effect: z = 2.03 (P = .04).
sults favoring the MDT group (adjusted FIGURE 2. Forest plot of the effectiveness of MDT for improving pain in patients with acute low back pain in
mean difference, –1.00; 95% CI: –2.09, comparison to (A) other physical therapy interventions, and (B) a combination of manual therapy with exercise.
–0.01); however, the difference was only The other physical therapy interventions included a combination of manual therapy with exercise or education.
1 point on a 0-to-10 visual analog scale, Abbreviations: CI, confidence interval; IV, independent variable; MDT, Mechanical Diagnosis and Therapy; SMD,
standardized mean difference.
and likely not clinically significant.
ferences in pain and disability in patients providing care based on MDT principles, cific treatment to be considered an MDT
with LBP. Nine studies included in the trained therapists obtained better treat- treatment. The classification process was
current study were published after the last ment outcomes than untrained thera- omitted in 5 of the included studies in the
meta-analysis,32 published in 2006 (TABLE pists.10 From the previous meta-analysis, previous systematic review.9,11,35,57,60 Thus,
1).1,18,22,33,40,45,47,53,54 There are 4 main differ- 2 studies included therapists who were the current findings provided an updated
ences between the previous and current not trained in MDT.5,12 Third, only stud- meta-analysis of the effectiveness of MDT,
meta-analyses. First, in the current me- ies in which classification was conducted and ensured that the included studies
ta-analysis, acute and chronic LBP were a priori were included in the current more closely followed the MDT program
as intended.
A In patients with acute LBP, we ob-
Study Weight SMD IV, Random (95% CI) served statistically significantly greater
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Cherkin et al3 45.4% 0.09 (–0.16, 0.34) improvement in pain intensity when
Machado et al33 35.0% 0.02 (–0.32, 0.35) utilizing the MDT approach compared
Schenk et al55 9.5% –0.61 (–1.43, 0.21) to the combination of manual therapy
Schenk et al54 10.1% –0.58(–1.37, 0.22) and exercise. Two studies in which di-
Total 100.0% –0.07 (–0.34, 0.20) rectional-preference exercises were the
primary means of treatment in the MDT
–2 –1 0 1 2
Favors MDT Favors other interventions group were analyzed.54,55 Directional
Heterogeneity: τ = 0.03, χ = 4.60, df = 3 (P = .20), I = 35%.
2 2 2
preference implies a rapid improve-
Test for overall effect: z = 0.51 (P = .61).
ment in patient symptoms in response
B to a specific exercise.36 This could ex-
Journal of Orthopaedic & Sports Physical Therapy®
provided greater improvements in pain patients. The intention to treat was not sification approaches that tailor treat-
and disability compared to other in- met for 4 studies, and it was not clear ments based on clinical characteristics
terventions and exercise alone, but how participants who dropped out rather than pathoanatomical diagnoses,
had similar outcomes compared to the were accounted for statistically.1,20,22,38 such as treatment-based classification
combination of manual therapy and ex- Also, some studies included only the and movement system impairments.13,27
ercise. The SMD values represented a
small treatment effect for the compari- A
son of MDT to other interventions for Study Weight SMD IV, Random (95% CI)
pain (SMD, –0.33) and disability (SMD, Miller et al38 9.6% –0.63 (–1.38, 0.12)
–0.28); therefore, despite statistical sig- Halliday et al22 14.5% –0.04 (–0.54, 0.46)
nificance, the clinical significance of the Paatelma et al45 16.7% –0.53 (–0.94, –0.13)
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
difference may be less meaningful. Other Long et al31 18.8% –0.81 (–1.13, –0.49)
symptom-matched approaches have also Garcia et al17 18.8% –0.09 (–0.41, 0.23)
demonstrated similar findings in patients Petersen et al47 21.6% –0.04 (–0.25, 0.17)
with chronic LBP.2,56 Total 100.0% –0.33 (–0.63, –0.03)
Although effective in treating chronic
–2 –1 0 1 2
LBP, MDT might not be any better than Favors MDT Favors other interventions
combined manual therapy plus exercise. Heterogeneity: τ = 0.10, χ = 19.81, df = 5 (P = .001), I = 75%.
2 2 2
It has been shown in treatment-based Test for overall effect: z = 2.19 (P = .03).
classification that patients who may ben-
B
efit from specific exercise may also benefit
Journal of Orthopaedic & Sports Physical Therapy®
T
insignificant improvements in outcome here is moderate- to high- type of intervention being compared to
measures for both the classification-spe- quality evidence that MDT is not MDT, and the effect sizes were generally
cific and the non–classification-specific superior to other rehabilitation considered small to moderate, which
groups.24,61 However, this current review interventions for reducing pain and means clinical significance needs to be
did find a significant difference between disability in patients with acute LBP. determined. Although some evidence
Downloaded from www.jospt.org at Grand Valley State University on April 12, 2018. For personal use only. No other uses without permission.
patient-matched treatment and generic In patients with chronic LBP, there is supported the use of MDT for assessing
exercise for disability in the short term moderate- to high-quality evidence that and treating LBP, therapists should be
for chronic LBP, albeit moderate. MDT is superior to other rehabilitation careful when using this approach exclu-
sively, because other treatments have
A
shown similar effectiveness, and a pa-
tient’s values and preferences should be
considered. t
Study Weight SMD IV, Random (95% CI)
Moncelon and Otero40 2.3% –0.11 (–1.16, 0.94)
Miller et al38 4.4% –0.64 (–1.39, 0.11)
KEY POINTS
Halliday et al22 9.1% –0.32 (–0.82, 0.18)
FINDINGS: For reducing pain and dis-
Paatelma et al 45
13.4% 0.00 (–0.40, 0.40)
ability in patients with acute low back
Garcia et al17 18.5% –0.36 (–0.68, –0.03)
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
–2 –1 0 1 2 REFERENCES
Favors MDT Favors exercise
Heterogeneity: τ = 0.00, χ = 1.18, df = 3 (P = .76), I = 0%.
2 2 2
1. Bonnet F, Monnet S, Otero J. Short-term effects
Test for overall effect: z = 4.39 (P<.0001).
of treatment according to “directional prefer-
FIGURE 5. Forest plot of the effectiveness of MDT for improving disability in patients with chronic low back pain in ence” of low back pain: a randomized controlled
comparison to (A) other physical therapy interventions, (B) a combination of manual therapy with exercise, and (C) trial. Kinésithér Rev. 2011;11:51-59. https://doi.
exercise. The other physical therapy interventions included either a combination of manual therapy with exercise org/10.1016/S1779-0123(11)75100-2
or exercise alone. Abbreviations: CI, confidence interval; IV, independent variable; MDT, Mechanical Diagnosis and 2. Browder DA, Childs JD, Cleland JA, Fritz JM. Ef-
Therapy; SMD, standardized mean difference. fectiveness of an extension-oriented treatment
back pain. N Engl J Med. 1998;339:1021-1029. for patients with acute low back pain: a ran- 27. Karayannis NV, Jull GA, Hodges PW. Physiother-
https://doi.org/10.1056/NEJM199810083391502 domized clinical trial. Spine (Phila Pa 1976). apy movement based classification approaches
4. Chou R, Qaseem A, Snow V, et al. Diagnosis 2003;28:1363-1371; discussion 1372. https://doi. to low back pain: comparison of subgroups
and treatment of low back pain: a joint clinical org/10.1097/01.BRS.0000067115.61673.FF through review and developer/expert survey. BMC
practice guideline from the American College 16. F urlan AD, Pennick V, Bombardier C, van Tulder Musculoskelet Disord. 2012;13:24. https://doi.
of Physicians and the American Pain Society. M. 2009 updated method guidelines for system- org/10.1186/1471-2474-13-24
Ann Intern Med. 2007;147:478-491. https://doi. atic reviews in the Cochrane Back Review Group. 28. Kilpikoski S, Airaksinen O, Kankaanpaa
org/10.7326/0003-4819-147-7-200710020-00006 Spine (Phila Pa 1976). 2009;34:1929-1941. M, Leminen P, Videman T, Alen M. Interex-
5. Delitto A, Cibulka MT, Erhard RE, Bowling RW, https://doi.org/10.1097/BRS.0b013e3181b1c99f aminer reliability of low back pain assess-
Tenhula JA. Evidence for use of an extension- 17. G arcia AN, Costa LC, da Silva TM, et al. Ef- ment using the McKenzie method. Spine
mobilization category in acute low back syn- fectiveness of back school versus McKenzie (Phila Pa 1976). 2002;27:E207-E214. https://doi.
drome: a prescriptive validation pilot study. Phys exercises in patients with chronic nonspecific low org/10.1097/00007632-200204150-00016
Ther. 1993;73:216-222. https://doi.org/10.1093/ back pain: a randomized controlled trial. Phys 29. K
oes BW, van Tulder MW, Ostelo R, Kim
ptj/73.4.216 Ther. 2013;93:729-747. https://doi.org/10.2522/ Burton A, Waddell G. Clinical guidelines for
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
6. Delitto A, George SZ, Van Dillen LR, et al. Low back ptj.20120414 the management of low back pain in pri-
pain. J Orthop Sports Phys Ther. 2012;42:A1-A57. 18. G arcia AN, Costa LC, Hancock MJ, et al. McKen- mary care: an international comparison. Spine
https://doi.org/10.2519/jospt.2012.0301 zie Method of Mechanical Diagnosis and Therapy (Phila Pa 1976). 2001;26:2504-2513. https://doi.
7. de Morton NA. The PEDro scale is a valid mea- was slightly more effective than placebo for pain, org/10.1097/00007632-200111150-00022
sure of the methodological quality of clinical but not for disability, in patients with chronic 30. Krause P, Forderreuther S, Straube A. TMS motor
trials: a demographic study. Aust J Physiother. non-specific low back pain: a randomised pla- cortical brain mapping in patients with complex
2009;55:129-133. https://doi.org/10.1016/ cebo controlled trial with short and longer term regional pain syndrome type I. Clin Neurophysiol.
S0004-9514(09)70043-1 follow-up. Br J Sports Med. In press. https://doi. 2006;117:169-176. https://doi.org/10.1016/j.
8. DerSimonian R, Laird N. Meta-analysis in clinical org/10.1136/bjsports-2016-097327 clinph.2005.09.012
trials. Control Clin Trials. 1986;7:177-188. https:// 19. G iesecke T, Gracely RH, Grant MA, et al. Evidence 31. Long A, Donelson R, Fung T. Does it matter
doi.org/10.1016/0197-2456(86)90046-2 of augmented central pain processing in idio- which exercise? A randomized control trial of
9. Dettori JR, Bullock SH, Sutlive TG, Franklin RJ, pathic chronic low back pain. Arthritis Rheum. exercise for low back pain. Spine (Phila Pa 1976).
Journal of Orthopaedic & Sports Physical Therapy®
Patience T. The effects of spinal flexion and exten- 2004;50:613-623. https://doi.org/10.1002/ 2004;29:2593-2602. https://doi.org/10.1097/01.
sion exercises and their associated postures in art.20063 brs.0000146464.23007.2a
patients with acute low back pain. Spine (Phila 20. G illan MG, Ross JC, McLean IP, Porter RW. The 32. Machado LA, de Souza M, Ferreira PH, Ferreira
Pa 1976). 1995;20:2303-2312. natural history of trunk list, its associated dis- ML. The McKenzie method for low back pain:
10. Deutscher D, Werneke MW, Gottlieb D, Fritz JM, ability and the influence of McKenzie manage- a systematic review of the literature with a
Resnik L. Physical therapists’ level of McKenzie ment. Eur Spine J. 1998;7:480-483. https://doi. meta-analysis approach. Spine (Phila Pa 1976).
education, functional outcomes, and utiliza- org/10.1007/s005860050111 2006;31:E254-E262. https://doi.org/10.1097/01.
tion in patients with low back pain. J Orthop 21. H agg O, Fritzell P, Nordwall A. The clinical brs.0000214884.18502.93
Sports Phys Ther. 2014;44:925-936. https://doi. importance of changes in outcome scores 33. Machado LA, Maher CG, Herbert RD, Clare
org/10.2519/jospt.2014.5272 after treatment for chronic low back pain. Eur H, McAuley JH. The effectiveness of the McK-
11. Elnaggar IM, Nordin M, Sheikhzadeh A, Spine J. 2003;12:12-20. https://doi.org/10.1007/ enzie method in addition to first-line care
Parnianpour M, Kahanovitz N. Effects of s00586-002-0464-0 for acute low back pain: a randomized con-
spinal flexion and extension exercises on 22. H alliday MH, Pappas E, Hancock MJ, et al. A ran- trolled trial. BMC Med. 2010;8:10. https://doi.
low-back pain and spinal mobility in chronic domized controlled trial comparing the McKenzie org/10.1186/1741-7015-8-10
mechanical low-back pain patients. Spine method to motor control exercises in people with 34. Maher CG, Sherrington C, Herbert RD, Moseley
(Phila Pa 1976). 1991;16:967-972. https://doi. chronic low back pain and a directional prefer- AM, Elkins M. Reliability of the PEDro scale
org/10.1097/00007632-199108000-00018 ence. J Orthop Sports Phys Ther. 2016;46:514- for rating quality of randomized controlled tri-
12. Erhard RE, Delitto A, Cibulka MT. Relative ef- 522. https://doi.org/10.2519/jospt.2016.6379 als. Phys Ther. 2003;83:713-721. https://doi.
fectiveness of an extension program and a com- 23. H efford C. McKenzie classification of mechanical org/10.1093/ptj/83.8.713
bined program of manipulation and flexion and spinal pain: profile of syndromes and directions 35. Malmivaara A, Häkkinen U, Aro T, et al. The
extension exercises in patients with acute low of preference. Man Ther. 2008;13:75-81. https:// treatment of acute low back pain—bed rest,
back syndrome. Phys Ther. 1994;74:1093-1100. doi.org/10.1016/j.math.2006.08.005 exercises, or ordinary activity? N Engl J Med.
https://doi.org/10.1093/ptj/74.12.1093 24. H enry SM, Van Dillen LR, Ouellette-Morton 1995;332:351-355. https://doi.org/10.1056/
13. Fritz J. Disentangling classification systems from RH, et al. Outcomes are not different for NEJM199502093320602
their individual categories and the category- patient-matched versus nonmatched treat- 36. McKenzie R, May S. The Lumbar Spine: Mechani-
specific criteria: an essential consideration ment in subjects with chronic recurrent low cal Diagnosis and Therapy. 2nd ed. Wellington,
to evaluate clinical utility. J Man Manip Ther. back pain: a randomized clinical trial. Spine J. New Zealand: Spinal Publications; 2003.
2010;18:205-208. https://doi.org/10.1179/10669 2014;14:2799-2810. https://doi.org/10.1016/j. 37. Menezes Costa LC, Maher CG, Hancock MJ,
8110X12804993427162 spinee.2014.03.024 McAuley JH, Herbert RD, Costa LO. The prognosis
14. Fritz JM, Brennan GP. Preliminary examination 25. H oy D, Bain C, Williams G, et al. A systematic of acute and persistent low-back pain: a meta-
ferred reporting items for systematic reviews ing patients with mechanical low-back pain. J ment of acute low back pain in primary care.
and meta-analyses: the PRISMA statement. Orthop Sports Phys Ther. 2000;30:368-383; Fam Pract. 1998;15:9-15. https://doi.org/10.1093/
Ann Intern Med. 2009;151:264-269. https://doi. discussion 384-389. https://doi.org/10.2519/ fampra/15.1.9
org/10.1371/journal.pmed.1000097 jospt.2000.30.7.368 61. Van Dillen LR, Norton BJ, Sahrmann SA, et al.
40. Moncelon S, Otero J. The McKenzie Method of 50. Rosedale R, Rastogi R, May S, et al. Efficacy of ex- Efficacy of classification-specific treatment and
Mechanical Diagnosis and Therapy in chronic low ercise intervention as determined by the McKen- adherence on outcomes in people with chronic
back pain with directional preference. Kinésithér zie system of Mechanical Diagnosis and Therapy low back pain. A one-year follow-up, prospec-
Rev. 2015;15:31-37. https://doi.org/10.1016/j. for knee osteoarthritis: a randomized controlled tive, randomized, controlled clinical trial. Man
kine.2014.11.086 trial. J Orthop Sports Phys Ther. 2014;44:173-181. Ther. 2016;24:52-64. https://doi.org/10.1016/j.
41. Murtezani A, Govori V, Meka VS, Ibraimi Z, Rrecaj https://doi.org/10.2519/jospt.2014.4791 math.2016.04.003
S, Gashi S. A comparison of McKenzie therapy with 51. Roussel NA, Nijs J, Meeus M, Mylius V, Fayt C, 62. Vos T, Allen C, Arora M, et al. Global, regional,
electrophysical agents for the treatment of work re- Oostendorp R. Central sensitization and altered and national incidence, prevalence, and years
lated low back pain: a randomized controlled trial. central pain processing in chronic low back pain: lived with disability for 310 diseases and inju-
J Back Musculoskelet Rehabil. 2015;28:247-253. fact or myth? Clin J Pain. 2013;29:625-638. ries, 1990-2015: a systematic analysis for the
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
clinical practice: criteria for the classification patients with back pain who meet a clinical pre- org/10.1097/00007632-199904010-00012
of central sensitization pain. Pain Physician. diction rule for spinal manipulation. J Man Manip 65. Werneke MW, Deutscher D, Hart DL, et al. McKen-
2014;17:447-457. Ther. 2012;20:43-49. https://doi.org/10.1179/204 zie lumbar classification: inter-rater agreement
45. Paatelma M, Kilpikoski S, Simonen R, Heinonen 2618611Y.0000000017 by physical therapists with different levels of
A, Alen M, Videman T. Orthopaedic manual 55. Schenk R, Jozefczyk C, Kopf A. A random- formal McKenzie postgraduate training. Spine
therapy, McKenzie method or advice only for ized trial comparing interventions in patients (Phila Pa 1976). 2014;39:E182-E190. https://doi.
low back pain in working adults: a random- with lumbar posterior derangement. J Man org/10.1097/BRS.0000000000000117
ized controlled trial with one year follow-up. J Manip Ther. 2013;11:95-102. https://doi. 66. Werneke MW, Hart DL, Resnik L, Stratford PW,
Rehabil Med. 2008;40:858-863. https://doi. org/10.1179/106698103790826455 Reyes A. Centralization: prevalence and effect
org/10.2340/16501977-0262 56. Sheeran L, van Deursen R, Caterson B, Sparkes on treatment outcomes using a standardized
46. Petersen T, Kryger P, Ekdahl C, Olsen S, Ja- V. Classification-guided versus generalized operational definition and measurement method.
cobsen S. The effect of McKenzie therapy as postural intervention in subgroups of nonspecific J Orthop Sports Phys Ther. 2008;38:116-125.
compared with that of intensive strengthen- chronic low back pain: a pragmatic random- https://doi.org/10.2519/jospt.2008.2596
ing training for the treatment of patients ized controlled study. Spine (Phila Pa 1976). 67. Whitehead A. Meta-Analysis of Controlled Clinical
with subacute or chronic low back pain: a 2013;38:1613-1625. https://doi.org/10.1097/ Trials. Chichester, UK: John Wiley; 2002.
randomized controlled trial. Spine (Phila BRS.0b013e31829e049b 68. Yamada K, Matsudaira K, Imano H, Kitamura
Pa 1976). 2002;27:1702-1709. https://doi. 57. Stankovic R, Johnell O. Conservative treatment A, Iso H. Influence of work-related psychosocial
org/10.1097/00007632-200208150-00004 of acute low-back pain. A prospective random- factors on the prevalence of chronic pain and
47. Petersen T, Larsen K, Nordsteen J, Olsen S, ized trial: McKenzie method of treatment versus quality of life in patients with chronic pain. BMJ
Fournier G, Jacobsen S. The McKenzie method patient education in “mini back school”. Spine Open. 2016;6:e010356. https://doi.org/10.1136/
compared with manipulation when used adjunc- (Phila Pa 1976). 1990;15:120-123. bmjopen-2015-010356
tive to information and advice in low back pain 58. Stanton TR, Fritz JM, Hancock MJ, et al. Evalua-
patients presenting with centralization or periph- tion of a treatment-based classification algorithm
@ MORE INFORMATION
eralization: a randomized controlled trial. Spine for low back pain: a cross-sectional study. Phys
(Phila Pa 1976). 2011;36:1999-2010. https://doi. Ther. 2011;91:496-509. https://doi.org/10.2522/
org/10.1097/BRS.0b013e318201ee8e ptj.20100272 WWW.JOSPT.ORG
1. McKenzie therap*.mp. 13. Low Back Pain/ 34. randomized controlled trial.pt.
2. McKenzie method*.mp. 14. (low* back adj2 pain*).mp. 35. controlled clinical trial.pt.
3. McKenzie treatment*.mp. 15. lumbar pain.mp. 36. RCT.ti,ab.
4. McKenzie exerci*.mp. 16. lumbar strain.ti,ab. 37. random*.ti,ab.
5. centralization.mp. 17. lumbar sprain.ti,ab. 38. placebo.ti,ab.
6. extension exercise*.mp. 18. Back Pain/ 39. trial.ti,ab.
7. flexion exercise*.mp. 19. (backache* or back ache*).ti,ab. 40. groups.ti,ab.
8. “mechanical diagnosis and therapy”.mp. 20. discogenic pain.ti,ab. 41. or/34-40
9. MDT.mp. 21. dorsalgia.ti,ab.
Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
29. Spondylolysis/
30. spondylolysis.ti,ab.
31. Spondylolysthesis.ti,ab.
32. lumbago.ti,ab.
33. or/13-32