Content uploaded by James H Mcauley
Author content
All content in this area was uploaded by James H Mcauley
Content may be subject to copyright.
Available via license: CC BY 2.0
Content may be subject to copyright.
RESEARC H ARTIC L E Open Access
The effectiveness of the McKenzie method in
addition to first-line care for acute low back pain:
a randomized controlled trial
Luciana AC Machado
1,2
, Chris G Maher
1*
, Rob D Herbert
1
, Helen Clare
3
, James H McAuley
1,4
Abstract
Background: Low back pain is a highly prevalent and disabling condition worldwide. Clinical guidelines for the
management of patients with acute low back pain recommend first-line treatment consisting of advice,
reassurance and simple analgesics. Exercise is also commonly prescribed to these patients. The primary aim of this
study was to evaluate the short-term effect of adding the McKenzie method to the first-line care of patients with
acute low back pain.
Methods: A multi-centre rando mized controlled trial with a 3-month follow-up was conducted between
September 2005 and June 2008. Patients seeking care for acute non-specific low back pain from primary care
medical practices were screened. Eligible participants were assigned to receive a treatment programme based on
the McKenzie method and first-line care (advice, reassurance and time-contingent acetaminophen) or first-line care
alone, for 3 weeks. Primary outcome meas ures included pain (0-10 Numeric Rating Scale) over the first seven days,
pain at 1 week, pain at 3 weeks and global perceived effect (-5 to 5 scale) at 3 weeks. Treatment effects were
estimated using linear mixed models.
Results: One hundred and forty-eight participants were randomized into study groups, of whom 138 (93%)
completed the last follow-up. The addition of the McKenzie method to first-line care produced statistically
significant but small reductions in pain when compared to first-line care alone: mean of -0.4 points (95%
confidence interval, -0.8 to -0.1) at 1 week, -0.7 points (95% confidence interval, -1.2 to -0.1) at 3 weeks, and -0.3
points (95% confidence interval, -0.5 to -0.0) over the first 7 days. Patients receiving the McKenzie method did not
show additional effects on global perceived effect, disability, function or on the risk of persistent symptoms. These
patients sought less additional health care than those receiving only first-line care (P = 0.002).
Conclusions: When added to the currently recommended first-line care of acute low back pain, a treatment
programme based on the McKenzie method does not produce appreciable additional short-term improvements in
pain, disability, function or global perceived effect. However, the McKenzie method seems to reduce health
utilization although it does not reduce patient’s risk of developing persistent symptoms.
Trial Registration: Australian New Zealand Clinical Trials Registry: ACTRN12605000032651
Background
Current c linical guidelines [1-3] recommend the provi-
sion of advice, reassurance and simple analgesics as
first-line treatment for patients with acute low back pain
consulting a primary care physician. Although not
recommended in most gui delines, exercises are also
commonly prescribed for this population. In t he USA
over two-thirds of physicians include exercises in their
initial care recommendations for patients with acute low
back pain [4].
The effectiveness of exercises for the treatment of
patients with acute low back pain is disputed. While
many trials have concluded that exercises are ineffective
for this population [5-7], these negative findings have
been challenged on the grounds that the therapy was
* Correspondence: cmaher@george.org.au
1
The George Institute for International Health, PO Box M201 Missenden Rd
Sydney, NSW 2050, Australia
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
© 2010 Machado et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution , and
reproduction in any medium, provided the original work is properly cited.
not appropriately administered [8]. A common criticism
is that patients are asked to exercise in a way that does
not take consideration of symptom response. Propo-
nents of exercise therapy argue that better results would
be obtained if exercises were customized to a patient’s
clinical presentation [8-11]. This treatment rationale
forms the basis of the McKenzie method [12,13], which
consists of a system of classification and classification-
based treatment that is commonly used to treat low
back pain in many countries and par ticularly in the
USA [14] and Europe [15-17].
Classification in the McKenzie method follows a com-
prehensive cl inical e xaminat ion inc luding exami nation
of posture and range of movement, together with the
assessment of patient’s symptomatic response to differ-
ent loading strategies applied to the spine (Figure 1).
Findings from this examination determine the classifica-
tion of low back pain into one of three syndromes:
derangement syndrome; dysf unction syndrome; or pos-
tural s yndrome [12]. The core component of treatment
in the McKenzie method i s exercise, which consists of
sustained postures or repeated movements similar to the
loading strategies used for the assessment . This method
also includes other components such as education and
postural training.
With the McKenzie a pproac h exercise is not used to
strengthen the back muscles, but to pr omote rapid
symptom relief. A key principle is to teach the patient
simple strategies to self-manage their pain. For example,
one pattern of symptomatic response observed during
the execution of the different loading strat egies is
known as ‘centralization’ and describes the phenomenon
by which pain referred from the spine is progressively
abolished [12], and this ca n be observed as early as at
thefirstvisittothetherapist[18].Foramoredetailed
account of the f ull clini cal picture of each of the sy n-
dromes and their respective treatment program, see
McKenzie’s textbooks [12,13].
Thescientificevidencetosupporttheuseofthe
McKenzie method is still scarce, particularly in pri mary
medical care. Only one high-quality randomized con-
trolled trial, conducted by Cherkin et al. [19] , has evalu-
ated the effectiveness of the McKenzie method for
patients with low back pain seeking care from primary
care physicians. Cherkin’s trial compared the McKenzie
treatment to an educational booklet and found that the
former was only marginally more effective for patients
with predominantly acute low back pain who were
refractory to the physician’s care after 1 w eek [19]. No
previous trial h as evaluated the effectiveness of a treat-
ment program based on the McKenzie method for
patients with acute low back pain when they first pre-
sented to a primary care physician. We conducted a
randomized controlled trial in which patients with acute
low back pain first presenting to a primary care physi-
cian were randomized either to receive first-line care
alone or first-line care and a treatment program based
on the McKenzie method.
Methods
This study was a multi-centre randomized controlled
trial conducted between September 2005 and June 2008.
The study was approved by the University of Sydney
Human Research Ethics Committee and prospectively
registe red with the Australian and New Zealand Cli nical
Trials Registry (ACTRN12605000032651) [20].
Patients
Thirty-one primary care physicians in 27 medical practices
located in a socio-economically diverse region in Sydney,
Australia, screened for eligibility consecutive patients seek-
ing care for low back pain. The screening for eligible
patients occurred between September 2005 and December
2007. To be eligible for inclusion, patients had to be 18 to
80 years old, present with a new episode of acute non-spe-
cific low back pain and be able and willing to visit one of
the trial physical therapists for commencement of the
McKenzie treatment program within 48 h of presentation
to the physician. A new episode of acute non-specific low
back pain was defined as pain in the area between the
12th rib and buttock crease (with or w ithout leg pain) of
less than 6 weeks duration, preceded by a period of at
least 1 month without low back pain in which the patient
did not consult a health care practitioner [21]. Patients
were excluded if they had any of the following: nerve root
compromise; ‘red flags’ for serious spinal pathology (for
example, infection, fracture); spinal surgery in the past 6
months; pregnancy; severe cardiovascular or metabolic
disease; or the inability to read and understand English.
Written consent was obtained from all those who agre ed
to participate.
Randomization
A statistician not involved in recruitment, data collec-
tion or treatment, developed a computer-generated ran-
domization sequence (randomly permuted blocks of 4, 6
and 8) that was placed in sequentially numbered, sealed
opaque envelopes [22]. All patients who satisfied the
eligibility criteria and volunteered to participate were
randomized to the First-line Care Group (recommended
first-line care alone) or the McKenzie Group (McKenzie
method in addition to the first-line care). It was not fea-
sible to blind participants or th erapists to the treatment
allocation. Participan ts were inf ormed that they would
receive the best available care according to the current
scientific knowledg e and that they woul d have a 50%
chance of receiving an additional exercise programme of
unknown efficacy.
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 2 of 10
Intervention
During the intervention period (3 w eeks), participants
were asked not to seek treatment for their back com-
plaints other than that provided in the trial.
First-Line Care
Participants in the First Line Care Group received first-
line care a s per guidel ine recommendations [1-3]. As
not all physicians were familiar with these recommenda-
tions, the physicians underwent individual training ses-
sions with members of the research team (LM and JM)
prior to the start of their participation in the study. The
first-line care consisted of the provision of advice to
remain active and to avoid bed rest, reassurance of the
favourable prognosis of acute low back pain and instruc-
tions to take acetaminophen (paracetamol) on a time-
contingent basis. Non-steroidal anti-inflammatory drugs
(NSAIDs) were not prescribed during the ensuing 3
weeks. However, participants already on a course of
NSAIDs when first visit ing the pr imary car e physician
were allowed to continue use of this medication. Partici-
pants were instructed to follow the physician’sadvice
for the next 3 weeks and, if necessary, to retu rn for fol-
low-up visits during this period. Although there was no
limit to th e number of follow-up visits, phys icians were
instructed to restrict treatment to a dvice and simple
analgesics.
McKenzie method
In addition to the first-line care, participants in the
McKenzie Group were immediately referred to a physi-
cal therapist and started a treatme nt programme bas ed
Figure 1 Dynamic loading strategies applied to the spine in the McKenzie method: (A) flexion in standing; (B) extension in standing;
(C) flexion in lying; (D) extension in lying; (E) side glide in standing; (F) therapist-assisted side glide in standing. Reproduced with
permission of Spinal Publications NZ Ltd.
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 3 of 10
on the McKenzie method within 48 h of their consul-
tation with the physician. Treatment was provided by
15 physi cal t herap ists wh o ha d co mpleted at least 100
h of postgraduate training and had achieved the status
of accredited (cr edentialed) McKenzie therapists. In
order to ensure an optimal implementation of the
treatment, physical therapists also attended a trai ning
session with a senior educator from the McKenzie
Institute International (HC) prior to the commence-
ment o f the study.
Physical therapists were instructed to follow exclu-
sively the treatment principles described in McKenzie’s
textbooks [12,13] and not to u se other treatment mod-
alities. After testing the participants’ pain response to a
comprehensive physical examination, therapists initially
classified each patient into one of the three McKenzie
syndromes (derangement, dysfu nction, or postural) and
an individualized treatment programme m atching this
classification was then provided. (For a brief descrip-
tion of syndromes and their matching treatment pro-
gram see [20].) For most participants, the guiding
treatment principle was to encourage directions of
movement and postures that produced centralization
of pain.
The number of treatment sessions was at the discre-
tion of the physical therapist, with a maximum of six
sessions over 3 weeks. In addition to the scheduled
treatment sessio ns, participants were encouraged to per-
form the prescribed exercises at home an d to follow the
therapist’s postural advice at all times. A copy of the
Treat Your Own Back book [23] was provided to all par-
ticipants, as usual in tre atment with the McKenzie
method. Some participants also received a lumbar sup-
port (original McKenzie lumbar ro ll) at the thera pist’ s
discretion. The McKenzie method was provided at no
cost to participants.
Outcomes
Our primary focus was on the short-term treatment
effects because a treatment programme based on the
McKenzie method is promoted as providing rapid symp-
tom improvement in patients with low back pain
[24-26]. Primary and secondary outcome measures were
nominated apriori[20] and were self-assessed by
participants.
Primary outcome measures
• Pain at 1 week, mean pain over the first 7 days and
pain at 3 weeks: participants were asked to rate the
average pain over the past 24 h on a 0-10 Numeric
Rating Scale [27].
• Global perceived effect at 3 weeks: participants
were asked to rate global perceived effect on a -5 to
5scale,anchoredat‘vastly worse’ an d ‘completely
recovered’ [28].
Secondary outcome measures
• Disability at 1 and 3 weeks: participants were asked
to measure any disability due to low back pain on the
0-24 Roland Morris Disability Questionnaire [29].
• Function at 1 and 3 weeks: participants were asked
to measure function on a 0-10 Patient Specific Func-
tional Scale [30].
• Glob al perc eived ef fect at 1 week: same as globa l
perceived effect at 3 weeks.
• Persistent low back pain at 3 months: participants
were asked ‘During the past 3 months have you ever
been completely free of low back pain? By thi s I
mean no low back pain at all, and would this pain-
free period have lasted for a whole month’.Those
answering ‘ No’ were co nsidered to have persistent
low back pain.
Adherence
Physical therapists assessed adherence with home exer-
cises and postural correction for participants in the
McKenzie Group at each visit. In order to be considered
as being adherent, participants had to report they were
maintaining the postural correction and performing the
prescribed exercises at home on at least 50% of the con-
sultations where adherence was assessed.
Assessment procedure
Participants received an assessment booklet in which
self-assessed outcomes were recorded. W ithin 24 h of
the consultatio n with the primar y care physician, and
prior to randomizatio n, a member of the research team
contacted participants by telephone in order to provide
instructions on how to complete the booklet and to col-
lect demographic data. In order to reduce the potential
for missing data due to participants mi splacing outco me
diaries, answers from the outcome diaries were tran-
scribed by a researcher who wa s blinded to allocation
during a telephone fo llow-up at 1 and 3 weeks (LM). At
3 months, the same blinded researcher contacted all
participants to enquire about persistent low back pain
and to collect information on additional hea lth care ser-
vices with respect to low back pain accessed after the 3-
week treatment period.
Statistical analysis
The sample size of 148 participants, determined a priori,
provided better than an 80% power to detect a differ-
ence of 1 unit in pain scores [27] between treatment
groups with an a level of 0.05, assuming a standard
deviation [SD] o f 2 units [31] and allowing for a loss to
follow-up of up to 15%. This sample size also allowed
for the detection of a difference of 1.2 units on the -5 to
5 global perceived effect scale (SD 2.4). Data were dou-
ble-entered and analysed by intention-to-treat.
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 4 of 10
Treatment effects were estimated using linear mixed
models (ran dom intercept and fixed coefficients) which
incorporated treatment, time and the interaction
between treatment and time. Outcomes were linearly
related to the log of time (r
2
of mean outcomes versus
log time ≥ 0.97) so time was entered into the models as
the log of time. Analyses were c onducted with the
‘xtmixed’ procedure in Stata v9. Estimates of effects at
specific time points (1 and 3 weeks) and the mean effect
on pain over the first 7 days were obtained from the
regression model using the ‘lincom’ routine. Relative
risks w ere calculated from the ratio of the proportions
of participants in the t wo groups with persistent symp-
toms and the use of other health care treatments.
Results
Between September 2005 and December 2007, primary
care physicians s creened 260 c onsecutive patients for
eligibility. A total of 148 patients were randomized (Fig-
ure 2). One pat ient randomized to each gr oup was mis-
diagnosed. Both were subsequently ruled ineligible to
participate in the trial and, thus, were considered legiti-
mate post-randomization exclusions [32] (on e had pain
from kidney stones and the other had pain in the thor-
acic spine on ly). The two groups’ demographic and clin-
ical characteristics were similar at baseline (Table 1).
Of the participants allocated to the McKenzie Group,
94% were initially classified into the derangement syn-
drome and 6% were classified into the dysfunction syn-
drome. Lumbar rolls were prescribed to 93% of
participants. Participants received a median of four (range
1-6) sessions with the physical therapist over a 3-week per-
iod, with a median of two (range 1-3) sessions over the
first week. Data on exercise adherence over the first week
were available for 56 participants in the McKenzie Group
and data on exercise adherence over the 3-week treatment
period were available for 50 participants in the McKenzie
Group. Adherenc e rates (proportions) were 66% over the
first week and 74% over the treatment period. The group
for which adheren ce data was available had similar base-
line scores for pain, disability and function to the group
for whom data was not available ( data not shown). The
maximum number of participants lost to follow-up at any
time-point was eight (5%).
Table 2 shows the mean outcomes by group and Fig-
ure 3 shows reductions in pain over ti me in the groups.
The additional effects of t he McKenzie method on pain
(that is, the adjusted difference in outcomes between the
McKenzieGroupandtheFirst-lineCareGroup)were
statistically significant but smaller than our pre-specified
threshold for between-group clinical importance of 1
unit (P = 0.02; Table 2). The addition of the McKenzie
260 consecutive patients
assessed for eligibility
112 patients excluded
57 did not meet inclusion criteria*
older than 80 years old (5)
no pain on lower back (7)
pain of more than 6 weeks (31)
recurrent episode of pain (27)
not able/willing to visit PT (12)
22 met exclusion criteria*
nerve root compromise (3)
suspected serious spine pathology (8)
spinal surgery in past 6 months (1)
pregnant (2)
contra-indication to exercise (3)
unable to understand English (6)
33 refused to
p
artici
p
ate
148 patients randomized
2 patients excluded after
randomization (ineligible)
Completed follow-up
assessment
70 at 1 week
70 at 3 weeks
70 at 3 months
Completed follow-
up assessment
68 at 1 week
69 at 3 weeks
68 at 3 months
73 patients assigned
to first line care &
McKenzie method
73 patients assigned
to first-line care
alone
Figure 2 Flow of participants through the trial. *Some patients presented more than one exclusion criteria.
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 5 of 10
method reduced pain by a mean of 0.4 points on a 0-10
pain scale at 1 week (95% confidence interval [CI], -0.8
to -0.1) and by a mean of 0.7 points at 3 weeks (95% CI,
-1.2 to -0.1). The average pain experienced over the first
7 days was also slightly lower in the McKenzie Group
(mean effect, -0.3; 95% CI, -0.5 to -0.0). For all other
outcomes, the additional effects of the McKenzie
method were near zero at all time points and not statis-
tically significant (Table 2).
Thirty-seven (53%) subjects in the McKenzie Group and
32 (47%) in the First-line Care Group developed persistent
low back pain; this difference was not statistically signifi-
cant (relative risk, 1.1; 95% CI, 0.8 to 1.6; P =0.49).Sub-
jects under treatment with the McKenzie method were less
likely to seek additional health care for their back com-
plaints after the 3-week treatment period; five participants
(7%) in the McKenzie Group and 18 participants (26%) in
theFirst-lineCareGroupsought additional health care
(relative risk, 0.27; 95% CI, 0.1 to 0.7; P = 0.002). According
to the participants’ repo rts, the most commonly sought
additional treatments were other forms of physical therapy
(32%), NSAIDs (18%) and acupuncture (14%).
Discussion
Findings of our study are in line with current clinical
practice guidelines, which support the view that most
patients with acute non-specific low back pain will exhi-
bit a speedy recovery. For instance, participants from
both the McKenzie Group and the First -line Care
Group showed large short-term improvements in o ut-
comes from baseline (for example, there were 53% and
59% reductions in pain at 1 week follow-up, respec-
tively). However, our trial showed that, in patients with
acute low back pain receiving recommended first-line
care, the addition of a treatment programme based on
the McKenzie method did not have clinically meaningful
effects on pain, disability, function, global perceived
effect or risk of developing persistent symptoms.
We considered the additional effect of the McKenzie
method on pain (that is, adjusted between-group differ-
ences in pain of 0.7 points or less) to be small. For
instance, thi s effect w as less than that conside red by a
panel of experts to be a minimal important change [33].
We acknowledge that cut-offs used to detect minimal
important changes are generally developed to assist
Table 1 Characteristics of participants at baseline.
McKenzie (N = 73) First-line Care Group (N = 73)
Age 47.5 ± 14.4 45.9 ± 14.9
Sex (female) 38 (52%) 35 (48%)
Duration of current low back pain episode
Less than 2 weeks 48 (66%) 49 (67%)
From 2 to 6 weeks 25 (34%) 24 (33%)
Pain radiating to the leg 33 (45%) 36 (50%)
Pain is movement-dependant* 61 (84%) 58 (80%)
Previous episode of low back pain 54 (74%) 49 (67%)
Participating in moderate exercise
†
44 (60%) 46 (63%)
Taking medication (any types) 54 (74%) 52 (71%)
Using non-steroidal anti-inflammatory drugs 28 (38%) 22 (30%)
Days off work or school due to low back pain 0.7 ± 1.4 0.8 ± 1.2
Compensation case (worker’s compensation) 3 (4%) 1 (1%)
General health status
Excellent 11 (15%) 16 (22%)
Very good 35 (48%) 37 (51%)
Good 18 (25%) 17 (23%)
Fair 6 (8%) 2 (3%)
Poor 3 (4%) 1 (1%)
Pain
‡
6.6 ± 1.8 6.3 ± 1.9
Disability
§
13.7 ± 5.5 13.5 ± 5.3
Function
¶
3.7 ± 1.6 3.4 ± 1.8
Data are means ± standard deviations or frequencies (%).
*Positive answer to question ‘Does your pain change in intensity or location depending on any position or movement?’.
†
Moderate exercise was any type of exercise performed for at least 30 min, three times per week or more.
‡
Numerical rating scale, scored from 0 (no pain) to 10 (worst pain possible).
§
Roland-Morris Disability Questionnaire, scored from 0 (no disability) to 24 (high disability).
¶
Patient Specific Functional Scale, scored from 0 (unable to perform activity) to 10 (able to perform activity at pre-injury level).
NSAIDs, non-steroidal anti-inflammatory drugs.
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 6 of 10
Time (days)
127654321
Pain (0-10 scale)
0
2
4
6
8
10
McKenzie Group
First-line Care Group
Figure 3 Pain scores in the McKenzie and First-line Care groups. Value s are unadjusted means and standard errors. For clarity, data for the
two groups have been slightly offset on the time axis.
Table 2 Mean outcomes in treatment groups and effects of the addition of the McKenzie method to the
recommended first-line care.
Outcome N** Unadjusted mean outcome (SE) Adjusted mean outcome (SE) Treatment effect* (95% CI) P value
McKenzie First-line Care McKenzie First-line Care McKenzie - First-line Care
Pain
†
0.02
1 week 70/69 3.5 (0.3) 3.7 (0.3) 3.6 (0.2) 4.0 (0.2) -0.4 (-0.8 to -0.1)
3 weeks 70/68 2.0 (0.2) 2.3 (0.3) 1.8 (0.2) 2.5 (0.2) -0.7 (-1.2 to -0.1)
Mean pain over first 7 days 70/69 4.9 (0.2) 4.9 (0.2) 4.7 (0.2) 5.0 (0.2) -0.3 (-0.5 to -0.0)
Global perceived effect
‡
1 week 70/68 2.6 (0.2) 2.1 (0.2) 2.6 (0.2) 2.1 (0.2) 0.5 (-0.0 to 1.1) 0.07
3 weeks 70/69 3.6 (0.1) 3.3 (0.2) 3.6 (0.1) 3.3 (0.2) 0.3 (-0.3 to 0.8) 0.33
Disability
§
0.74
1 week 70/68 8.4 (0.7) 9.0 (0.8) 8.0 (0.5) 8.2 (0.5) -0.2 (-1.5 to 1.0)
3 weeks 70/69 4.6 (0.7) 4.5 (0.7) 4.8 (0.7) 5.1 (0.7) -0.3 (-2.3 to 1.6)
Function
¶
0.90
1 week 70/68 6.2 (0.3) 5.8 (0.3) 6.2 (0.2) 6.2 (0.2) 0.0 (-0.4 to 0.5)
3 weeks 70/69 7.9 (0.2) 7.7 (0.3) 7.8 (0.2) 7.7 (0.3) 0.0 (-0.7 to 0.8)
*Treatment effects are model-based adjusted differences in outcomes between groups. For a global perceived effect, which was measured only at two time
points after randomization, adjusted means and treatment effects are the same as unadjusted means and treatment effects. Primary outcomes are highlighted in
italics. Effectiveness of the addition of the McKenzie method is indicated by negative effects for pain and disability and by positive effects of global perceived
effect and function. **Number of participants in McKenzie/First-line Care groups for whom data were available.
†
Numerical rating scale, scored from 0 (no pain) to 10 (worst pain possible).
‡
Global perceived effect scale scored from -5 (much worse) to 5 (completely recovered).
§
Roland-Morris Disability Questionnaire, scored from 0 (no disability) to
24 (high disability).
¶
Patient Specific Functional Scale, scored from 0 (unable to perform activity) to 10 (able to perform activity at pre-injury level).
SE, standard error.
CI, confidence interval.
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 7 of 10
interpretat ion of within-group differences. Nevertheless,
we still bel ieve the magni tude o f the additional effect of
the McKenzie method on pain was trivial in our trial
and would not be considered worthwhile by most health
practit ioners respons ible for m anaging pati ents with
acute low b ack pain in primary care and, possibly, also
by the patients. In making this judgement we also took
into account the fact that participants and therapists
were not (and could not be) blinded to treatment alloca-
tion. Lack of blinding is likely to e xaggerate treatment
effects on subjective outcomes such as pain [34]. Finally,
estimates of the effects of treatments in the present trial
were very precise, as indicated by the narrow confidence
intervals (it is unlikely that improvements seen after the
addition of the McKenzie method to first-line care
would be larger than 12% even in the best-case
scenario).
Some trials provide a description of the results by
computing the proportion of subjects who improve wit h
each treatment. We chose not to follow this approach in
the curre nt trial because statistical power is reduced
when continuous outcomes are dichotomized and
because the choice of cut-offs for improvement can
influence the results. As an il lustration for p ain
improvement it was seen that at 1 week, with a cut-off
of two points or less for improvement, the McKenzie
Group had more participa nts who i mproved (54.4% ver-
sus 45.6%), when the cut-off for improvement was raised
to three or four points the First-line Care Group had
more participants who improved (53.7% versus 46.3%);
and for a cut-o ff of five or six points there was a similar
number of pa rticipants improved in both groups (50%).
This patt ern demonstrates the limitations of this
approach.
TheadditionoftheMcKenziemethodtofirst-line
care may be worthwhile under the health sector per-
spective. This i s because we found t hat participants in
the First-line Care Group sought more additional care
than participants in the McKenzie Group. Unfortu-
nately, this assumption cannot be c onfirmed from the
results of the present trial since a proper cost-effective-
ness a nalysis would need to be conducted and our trial
was n ot designed to include such analysis (our main
focus was on treatment effectiveness). It is possible that
the i mbalance found in the usage of additional care
could have diluted the estimates of the effect of the
McKenzie method at 3 months. However, in our view,
the greater use of additional care by those in the First-
line Care Group does not s ignificantly threaten the
internal validity of our study because our primary inter-
est was in the effects of the McKenzie method in the
first 3 weeks. Moreover, the available evidence from ran-
domized trials suggests t hat the effe cts of other treat-
ments for acute low back pain are modest at best [1,7]
and so any additional care is unlikely to have substa n-
tially distorted e stimates of effects of the McKenzi e
method at 3 months.
The quality of complex interventions, such as exercise,
probably influences the size of the effect observed in
randomized controlled trials [35]. In our trial the
McKenzie method was applied by therapists who had
undergone many hours of po stgraduate training pro-
vided by the McKenzie Institute International and were
considered to have expertise in the provision of this
therapy [36]. Thus, it is doubtful that other therapists
would have been likely to d eliver the treatment with
greater fidelity to McKenzie’s principles. Likewise,
because physicians in this trial were educated about cur-
rent guideline-based recommendations, i t is possible
that the first-line care they provided differs fr om that
typically provided by primary care physicians. Data from
recent surveys show that many physicia ns ar e not famil-
iar with the recommendations of clinical guidelines or
do not usually foll ow guideline recom mendations
[37,38].
Although previous randomized controlled trials
[19,39,40] have attempted to investigate the effects of a
treatment programme based on the McKenzie method
for low back pain in primary medical care, only the trial
conducted by Cherk in et al. [19] has implemented this
therapy according to its original principles (that is, the
programme was individualized and p roperly customized
to the patient’s clinical presentation) [12,41]. Cherkin et
al. [19] found that an individualized and customized
treatment programme based on the McKenzie method
reduced patient’s ‘bothersomeness of symptoms’ by 0.8
points on a 0-10 bothersomeness scale at 4 weeks, com-
pared to an educational booklet. This effect was not sta-
tistically significant at the conventional 0.05 level (P =
0.06). Apart from a greater satisfaction with treatment
among participants receiving the McKenzie method, the
authors did not find any greater effect of this therapy on
any other outcome at any time-point when compared to
an educational booklet [19]. Our trial and the trial of
Cherkin et al. [19] found remark ably similar estimates of
the effects of the McKenzie method. Both trials support
the conclusion that a treatment programme based on th e
McKenzie method does not provide an appreciable bene-
fit for patients already rec eiving re latively simple, low-
risk and inexpensive treatments, such as advice, reassur-
ance and simple analgesics or an educational booklet.
Our trial has a number of strengths. The investigation
of the additional effects of the McKenzie method in
patients already receiving first-line care from the pr i-
mary care physician is not common in the low back
pain literature. However, such a study design provides
information that is extremely relevant and helpful to
physicians in their clinical decision making as current
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 8 of 10
practice guidelines recommend that patients are only
referred for add it ional treatment if they do not succeed
with previous first-line care. In addition, the treatment
in our tri al was delivered by hi ghly-trained therapists,
patient’s adherence to the treatment programme was
satisfactory and we had a very low rate of loss to follow-
up. We acknowledge that it would have been preferable
to select only one primary outcome measure in order to
minimize the potential for a type I error (when we regis-
tered t he trial we specified four primary outcome mea-
sures). However, this would be a concern if we were
claiming that the statistically significant pain outcomes
were clinically important, whereas our view is that these
effects are trivially small. One limit ation was the lack of
therapist and patient blinding th at could have been pos-
sible with the use of a placebo. However, given the nat-
ure of the programme of the McKenzie therapy,
therapist bli nding was not feasible and, in the design
stage, the re was no consensus among the research team
on a placebo that was both clearly inert and credible to
the patient. Another limitatio n was that adherence to
protocol by either the physical therapists or primary
care physicians were not evaluated. Finally, the lack of a
proper cost-effectiveness analysis and the short-term fol-
low-up limited the scope of our conclusions.
Conclusions
A treatment pr ogramme based o n the McKenzie method
does not produce appreciable improvements in pain, dis-
ability, function, global perceived effect or risk of devel-
oping persistent symptoms in patients with acute low
back pain receiving recomm ended first-line care. Patients
with acute low back pain receiving only the recom-
mended first-line care seek more additional health care
than patients receiving the McKenzie method.
Abbreviations
CI: confidence interval; NSAIDs: non-steroidal anti-inflammatory drugs; SD:
standard deviation.
Acknowledgements
We thank Dr Anurina Das for her valuable assistance with data collection.
We also thank the participating health practitioners and Backcare Products
Australia. This work was supported by a research and development grant
from the University of Sydney, Australia. Dr Machado is a research fellow
supported by Fundação de Amparo à Pesquisa do Estado de Minas Gerais
(FAPEMIG), Brazil. Dr Machado’s PhD was supported by a scholarsh ip from
the Australian Government. Professor Maher and Associate Professor Herbert
are senior research fellows funded by the National Health & Medical
Research Council (NHMRC), Australia.
Author details
1
The George Institute for International Health, PO Box M201 Missenden Rd
Sydney, NSW 2050, Australia.
2
Escola de Educação Física, Fisioterapia e
Terapia Ocupacional, Universidade Federal de Minas Gerais, Av Antônio
Carlos 6627, Pampulha, Belo Horizonte, MG 31270-901, Brazil.
3
Focus on
Backs 1/124 Shirley Road, Crows Nest, NSW 2065, Sydney, Australia.
4
Faculty
of Health Sciences, The University of Sydney, Sydney, Australia.
Authors’ contributions
LACM, CGM, RDH, JHM and HC contributed to the study conception and
design. LACM and JHM were responsible for the data acquisition. LACM,
CGM and RDH performed the analysis and interpretation of data. LACM and
CGM were responsible for drafting the manuscript. All authors performed a
critical revision of the manuscript for important intellectual content. LACM,
JHM and HC provided administrative and technical support. LACM was the
study coordinator. All authors read and approved the final manuscript.
Competing interests
HC is the D irector of Education for the McKenzie Institute International.
There are no other potential competing interests to report.
Received: 3 August 2009
Accepted: 26 January 2010 Published: 26 January 2010
References
1. Chou R, Qaseem A, Snow V, Casey D, Cross T Jr, Shekelle P, Owens D, for
the Clinical Efficacy Assessment Subcommittee of the American College of
Physicians and the American College of Physicians/American Pain Society
Low Back Pain Guidelines Panel: Diagnosis and treatment of low back
pain: a joint clinical practice guideline from the American College of
Physicians and the American Pain Society. Ann Inter Med 2007,
147:478-491.
2. Van Tulder M, Becker A, Bekkering T, Breen A, Gil del Real M, Hutchinson A,
Koes B, Laerum E, Malmivaara A, on behalf of the COST B13 working group
on guidelines for the management of acute low back pain in primary care:
European guidelines for the management of acute nonspecific low back
pain in primary care. Eur Spine J 2006, 15:S169-S191.
3. NHMRC: Evidence-based management of acute musculoskeletal pain
Canberra: National Health and Medical Research Council 2003.
4. Webster B, Courtney T, Huang Y-H, Matz S, Christiani D: Survey of acute
low back pain management by specialty group and practice experience.
J Occup Environ Med 2006, 48:723-732.
5. van Tulder M, Malmivaara A, Esmail R, Koes B: Exercise therapy for low
back pain: a systematic review within the framework of the Cochrane
Collaboration Back Review Group. Spine 2000, 25:2784-2796.
6. Hayden J, van Tulder M, Malmivaara A, Koes B: Meta-analysis: exercise
therapy for nonspecific low back pain. Ann Inter Med 2005, 142:765-775.
7. Machado L, Kamper S, Herbert R, Maher C, McAuley J: Analgesic effects of
treatments for non-specific low back pain: a meta-analysis of placebo-
controlled randomized trials. Rheumatology 2009, 48:520-527.
8. Machado L, de Souza M, Ferreira P, Ferreira M: The McKenzie method for
low back pain: a systematic review of the literature with a meta-analysis
approach. Spine 2006, 31:E254-E262.
9. McKenzie R: (Letters). Spine 2001, 26:1829-1831.
10. Fritz JM, Delitto A, Erhard RE: Comparison of classification-based physical
therapy with therapy based on clinical practice guidelines for patients
with acute low back pain: a randomized clinical trial. Spine 2003,
28:1363-1372.
11. Brennan G, Fritz J, Hunter S, Thackeray A, Delitto A, Erhard R: Identifying
subgroups of patients with acute/subacute “nonspecific” low back pain:
results of a randomized clinical trial. Spine 2006, 31:623-631.
12. McKenzie R, May S: The lumbar spine: Mechanical Diagnosis & Therapy, vol 1
New Zealand: Spinal Publications 2003.
13. McKenzie R, May S: The lumbar spine: Mechanical Diagnosis & Therapy, vol 2
New Zealand: Spinal Publications 2003.
14. Battié MC, Cherkin DC, Dunn R, Ciol MA, Wheeler KJ: Managing low back
pain: attitudes and treatment preferences of physical therapists. Phys
Ther 1994, 74:219-226.
15. Gracey J, McDonough S, Baxter D: Physiotherapy management of low
back pain: a survey of current practice in Northern Ireland. Spine 2002,
27
:406-411.
16. Foster N, Thompson K, Baxter D, Allen J: Management of nonspecific low
back pain by physiotherapists in Britain and Ireland: a descriptive
questionnaire of current clinical practice. Spine 1999, 24:1332-1342.
17. Hamm L, Mikkelsen B, Kuhr J, Støvring H, Munck A, Kragstrup J: Danish
physiotherapists’ management of low back pain. Advances in Physiother
2003, 5:109-113.
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 9 of 10
18. Long A: The centralization phenomenon: its usefulness as a predictor of
outcome in conservative treatment of chronic low back pain (a pilot
study). Spine 1995, 20:2513-2521.
19. Cherkin D, Deyo R, Battie M, Street J, Barlow W: A comparison of physical
therapy, chiropractic manipulation, and provision of an educational
booklet for the treatment of patients with low back pain. N Engl J Med
1998, 339 :1021-1029.
20. Machado L, Maher C, Herbert R, Clare H, McAuley J: The McKenzie method
for the management of acute non-specific low back pain: design of a
randomised controlled trial [ACTRN012605000032651]. BMC Musculoskelet
Disord 2005, 6:50.
21. de Vet H, Heymans M, Dunn K, Pope D, Beek A, Macfarlane G, Bouter L,
Croft P: Episodes of low back pain: a proposal for uniform definitions to
be used in research. Spine 2002, 27:2409-2416.
22. Pocock S: Clinical trials. A practical approach Chichester: John Wiley & Sons,
1 1984.
23. McKenzie R: Treat your own back. 6th (revised) ed New Zealand: Spinal
Publications 2006.
24. Delitto A, Cibulka M, Erhard R, Bowling R, Tenhula J: Evidence for use of an
extension-mobilization category in acute low back syndrome: a
prescriptive validation pilot study. Phys Ther 1993, 73:216-228.
25. Schenk R, Jozefczyk C, Kopf A: A randomized trial comparing
interventions in patients with lumbar posterior derangement. J Manual
Manip Ther 2003, 11:95-102.
26. Werneke M, Hart D, Cook D: A descriptive study of the centralization
phenomenon: a prospective analysis. Spine 1999, 24:676-683.
27. Farrar J, Young J Jr, LaMoreaux L, Werth J, Poole R: Clinical importance of
changes in chronic pain intensity measured on an 11-point numerical
pain rating scale. Pain 2001, 94:149-158.
28. Pengel L, Refshauge K, Maher C: Responsiveness of pain, disability, and
physical impairment outcomes in patients with low back pain. Spine
2004, 29:879-883.
29. Roland M, Morris R: A study of the natural history of back pain. Part I:
development of a reliable and sensitive measure of disability in low-
back pain. Spine 1983, 8:141-144.
30. Stratford P, Gill C, Westaway M, Binkley J: Assessing disability and change
on individual patients: a report of a patient specific measure. Physiother
Can 1995, 47:258-263.
31. Hancock MJ, Maher CG, Latimer J, McLachlan A, Cooper C, O Day R,
Spindler M, McAuley J: Assessment of diclofenac or spinal manipulative
therapy, or both, in addition to recommended first-line treatment for
acute low back pain: a randomised controlled trial. Lancet 2007,
370:1638-1643.
32. Fergusson D, Aaron S, Guyatt G, Hebert P: Post-randomisation exclusions:
the intention to treat principle and excluding patients from analysis.
BMJ 2002, 325:652-654.
33. Ostelo R, Deyo R, Stratford P, Waddell G, Croft P, Von Korff M, Bouter L, de
Vet H: Interpreting change scores for pain and functional status in low
back pain. Spine 2008, 33:90-94.
34. Wood L, Egger M, Gluud L, Schulz K, Jüni P, Altman D, Gluud C, Martin R,
Wood A, Sterne J: Empirical evidence of bias in treatment effect
estimates in controlled trials with different interventions and outcomes:
meta-epidemiological study. BMJ 2008, 336:601-605.
35. Herbert R, Bo K: Analysis of quality of interventions in systematic reviews.
BMJ 2005, 331:507-509.
36. Clare H, Adams R, Maher C: Reliability of the McKenzie spinal pain
classification using patient assessment forms. Physiotherapy 2004,
90:114-119.
37. Fullen B, Maher T, Bury G, Tynan A, Daly L, Hurley D: Adherence of Irish
general practitioners to European guidelines for acute low back pain: a
prospective pilot study. Eur J Pain 2007, 11:614-623.
38. Overmeer T, Linton S, Holmquist L, Eriksson M, Engfeldt P: Do evidence-
based guidelines have an impact in primary care? A cross-sectional
study of Swedish physicians and physiotherapists. Spine 2004, 30:146-151.
39. Underwood M, Morgan J: The use of a back class teaching extension
exercises in the treatment of acute low back pain in primary care. Fam
Pract 1998, 15:9-15.
40. Malmivaara A, Hakkinen U, Aro T, Heinrichs M, Koskenniemi L, Kuosma E,
Lappi S, Paloheimo R, Servo C, Vaaranen V, Hernberg S: The treatment of
acute low back pain: bed rest, exercises, or ordinary activity?. N Engl J
Med 1995, 332:351-355.
41. Donelson R: (Letters). Spine 2001, 26:1827-1829.
Pre-publication history
The pre-publication history for this paper can be accessed here:http://www.
biomedcentral.com/1741-7015/8/10/prepub
doi:10.1186/1741-7015-8-10
Cite this article as: Machado et al.: The effectiveness of the McKenzie
method in addition to first-line care for acute low back pain: a
randomized controlled trial. BMC Medicine 2010 8:10.
Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit
Machado et al. BMC Medicine 2010, 8:10
http://www.biomedcentral.com/1741-7015/8/10
Page 10 of 10
