Psychological response in spinal manipulation (PRISM): a systematic review of psychological outcomes in randomised controlled trials.
ABSTRACT The most important risk factors for back and neck pain are psychosocial. Nevertheless, systematic reviews of spinal manipulation have concentrated on pain and spine related disability, and ignored psychological outcomes.
To assess whether spinal manipulation was effective in improving psychological outcome.
Systematic review of randomised controlled trials (RCTs).
RCTs were identified by searching Medline, CINAHL, Embase, CENTRAL, AMED, PsycINFO until November 2005. Trials reporting psychological outcomes including the mental health components of generic outcomes were extracted, and combined where appropriate in meta-analyses.
One hundred and twenty nine RCTs of spinal manipulation were identified; 12 had adequately reported psychological outcomes. Six trials with a verbal intervention comparator were combined in a meta-analysis, and found a mean benefit from spinal manipulation equivalent to 0.34 of the population standard deviation (S.D.) [95% confidence interval (CI) 0.23-0.45] at 1-5 months; 0.27 of the S.D. [95% CI 0.14-0.40] at 6-12 months. Eight trials with a physical treatment comparator were combined in a meta-analysis and found a mean benefit of 0.13 of the S.D. [95% CI 0.01-0.24] in favour of manipulation at 1-5 months; 0.11 of the S.D. [95% CI -0.02 to 0.25] at 6-12 months.
There was some evidence that spinal manipulation improved psychological outcomes compared with verbal interventions.
- [Show abstract] [Hide abstract]
ABSTRACT: Shoulder impingement syndrome (SIS) is a complex, multi-factorial problem that is treated with a variety of different conservative options. One conservative option that has shown effectiveness is manual therapy to the thoracic spine. Another option, manual therapy to the cervical spine, has been studied only once with good results, evaluating short-term outcomes, in a small sample size. The purpose of this study was to investigate the benefit of neck manual therapy for patients with SIS. The study was a randomised, single blinded, clinical trial where both groups received pragmatic, evidence-based treatment to the shoulder and one group received neck manual therapy. Subjects with neck pain were excluded from the study. Comparative pain, disability, rate of recovery and patient acceptable symptom state (PASS) measures were analyzed on the 68 subjects seen over an average of 56.1 days (standard deviation (SD)=55.4). Eighty-six percent of the sample reported an acceptable change on the PASS at discharge. There were no between-groups differences in those who did or did not receive neck manual therapy; however, both groups demonstrated significant within-groups improvements. On average both groups improved 59.7% (SD=25.1) for pain and 53.5% (SD=40.2) for the Quick Disabilities of the Shoulder and Hand Questionnaire (QuickDASH) from baseline. This study found no value when neck manual therapy was added to the treatment of SIS. Reasons may include the lack of therapeutic dosage provided for the manual therapy approach or the lack of benefit to treating the neck in subjects with SIS who do not have concomitant neck problems.Manual therapy 06/2013; · 2.32 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: The aim of this study was to identify the minimal clinically important threshold score for rate of recovery (RoR) using the Patient Acceptable Symptom State (PASS) as an anchor measure, and to measure the association of the RoR with the raw and percentage change scores of the QuickDASH and the Numeric Pain Rating Scale (NPRS). Threshold measures for RoR and the PASS were examined using a Receiver Operating Characteristic (ROC) Curve. Participants were dichotomized to those who stated "yes" and those who stated "no" on the PASS. Pearson correlation coefficients were performed to determine the relationship between the NPRS and QuickDASH raw and percentage change scores and the RoR at discharge. ROC curve statistics suggest that a cut-point of greater than 87% on the self-report RoR (SN=0.62, SP=0.73; +LR=2.26: -LR=0.56) corresponded to the patient considering their state acceptable. No significant associations existed between either the raw NPRS or the raw QuickDASH, and the RoR scores although percentage change scores were associated with the RoR (QuickDASH r=0.29; p=0.02; NPRS = r=0.30; p=0.02). Patients reporting greater than or equal to 87% on RoR are likely to have met the Patient Acceptable Symptom State. Although statistically significant, the lack of strong association between RoR and change scores for the NPRS and QuickDASH suggests that the measures capture different constructs. Level 2C.International journal of sports physical therapy. 12/2013; 8(6):784-92.
- [Show abstract] [Hide abstract]
ABSTRACT: Low back pain (LBP) is common and costly and few treatments have been shown to be markedly superior to any other. Effort has been focused on stratifying patients to better target treatment. Recently the STarT Back Screening Tool (SBT) has been developed for use in primary care to enable sub grouping of patients based on modifiable baseline characteristics and has been shown to be associated with differential outcomes. In the UK the SBT is being recommended to assist in care decisions for those presenting to general practitioners with LBP. In the light of growing recommendation for widespread use of this tool, generalisability to other LBP populations is important. However, studies to date have focused only on patients attending physiotherapy whereas LBP patients seeking other treatment have not been investigated. This study aims to investigate the utility of the SBT to predict outcomes in LBP patients presenting for chiropractic management. A total of 404 patients undergoing chiropractic care were asked to complete the SBT before initial treatment. Clinical outcomes were collected at 14, 30 and 90 days following this initial consultation. The clinical course was described comparing SBT categories and logistic regression analysis performed to examine the tool's prognostic utility. Although the high-risk categories had greater pain at baseline this difference rapidly faded, with both change in composite outcome scores and pain scores being statistically insignificant between the risk groups at 30 and 90 days follow up. In addition, both univariate and adjusted analysis showed no prognostic utility of the SBT categorisations to differentiate clinical outcomes between risk groups. Whilst the SBT appears useful in some back pain populations it does not appear to differentiate outcomes in LBP patients seeking chiropractic care.Chiropractic & manual therapies. 06/2012; 20(1):17.
Complementary Therapies in Medicine (2007) 15, 271—283
Psychological response in spinal manipulation
(PRISM): A systematic review of psychological
outcomes in randomised controlled trials
Nefyn H. Williamsa,∗, Maggie Hendrya, Ruth Lewisa,
Ian Russellb, Alex Westmorelandc, Clare Wilkinsona
aDepartment of General Practice, Cardiff University, Centre for Health Services Research/North Wales
Clinical School, Gwenfro Building, Wrecsam Technology Park, Wrecsam LL13 7YP, United Kingdom
bInstitute of Medical and Social Care Research, University of Wales Bangor, United Kingdom
cWales Office of Research and Development for Health and Social Care, Cardiff, United Kingdom
Available online 8 March 2007
Nevertheless, systematic reviews of spinal manipulation have concentrated on pain
and spine related disability, and ignored psychological outcomes.
Objective: To assess whether spinal manipulation was effective in improving psycho-
Design: Systematic review of randomised controlled trials (RCTs).
Methods: RCTs were identified by searching Medline, CINAHL, Embase, CENTRAL,
AMED, PsycINFO until November 2005. Trials reporting psychological outcomes
including the mental health components of generic outcomes were extracted, and
combined where appropriate in meta-analyses.
Results: One hundred and twenty nine RCTs of spinal manipulation were identified;
12 had adequately reported psychological outcomes. Six trials with a verbal inter-
vention comparator were combined in a meta-analysis, and found a mean benefit
from spinal manipulation equivalent to 0.34 of the population standard deviation
(S.D.) [95% confidence interval (CI) 0.23—0.45] at 1—5 months; 0.27 of the S.D. [95%
CI 0.14—0.40] at 6—12 months. Eight trials with a physical treatment comparator
were combined in a meta-analysis and found a mean benefit of 0.13 of the S.D. [95%
CI 0.01—0.24] in favour of manipulation at 1—5 months; 0.11 of the S.D. [95% CI
−0.02 to 0.25] at 6—12 months.
Conclusions: There was some evidence that spinal manipulation improved psycho-
logical outcomes compared with verbal interventions.
© 2007 Elsevier Ltd. All rights reserved.
∗Corresponding author. Tel.: +44 1978 726090;
fax: +44 1978 727431.
E-mail address: firstname.lastname@example.org (N.H. Williams).
Spinal pain is a common cause of disability and
work loss,1,2and one of the main reasons for
0965-2299/$ — see front matter © 2007 Elsevier Ltd. All rights reserved.
272N.H. Williams et al.
consulting general practitioners (GPs),3and prac-
titioners of complementary medicine, particularly
osteopaths and chiropractors.4A Cochrane review
of spinal manipulation for low back pain5found
that spinal manipulation was more effective than
placebo or therapies judged to be ineffective or
harmful, such as traction, corset or bed rest; but
had no advantage over GP care, analgesics, phys-
ical therapy, exercises, or back school, in terms
of pain and spine related disability. The Cochrane
review for neck pain6found evidence that spinal
manipulation in combination with exercise was
more effective than either alone, in terms of
pain, spine related disability, and global perceived
effect. These Cochrane reviews only considered
a narrow selection of outcomes, which did not
include psychological measures. Psychosocial fac-
tors are more important in the presentation and
development of neck and back pain and associated
disability than biological factors,7so any psycho-
logical benefit of manipulation may have important
therapeutic implications. Many RCTs not only use
a condition-specific outcome measure for neck or
low back pain, but also generic scales with a psy-
chological component, such as the SF-368or SF-12,9
or psychological instruments as secondary outcome
measures. The Cochrane reviews can also be crit-
icised because the quality criteria scores used10
discriminated against pragmatic trials in favour of
To assess whether randomised controlled trials of
spinal manipulation show evidence of effectiveness
in improving psychological outcome.
Inclusion criteria for studies to review
• Type of study. Published randomised controlled
• Type of participant. Adults with neck pain, back
pain or headache believed to originate from the
spine and associated structures.
• Type of intervention. Studies using manipula-
tion or mobilisation techniques. Manipulation is
a localised force of high velocity and low ampli-
tude directed at a spinal segment. Mobilisation
uses low velocity passive movement techniques,
within the spine’s normal range of movement.
• Type of outcome. Psychological outcomes includ-
ing: back pain beliefs; fear-avoidance beliefs;
self-efficacy; depression and anxiety; generic
outcome measures with a psychological com-
ponent. Social outcomes were not included, as
there is already an extensive literature on occu-
• Language. One trial written in German was trans-
Search strategy for identification of studies
Embase, CENTRAL, AMED, and PsycINFO were
using a search strategy adapted from previous
systematic reviews5,6that used a combination
of free text words and indexed terms covering:
randomized controlled trials, back pain, neck
pain, sciatica, whiplash injuries, spine, spinal
diseases, intervertebral disc and musculoskeletal
manipulations (Appendix A). Reference lists from
previous Cochrane systematic reviews5,6were also
screened, in addition to citation tracking and hand
searching of key journals. Study selection was
performed by NW, and checked by an independent
reviewer (MH or AW), with discrepancies resolved
by consensus. It was not feasible to mask the
names of authors or journals, as reviewers were
already familiar with the literature. Copies of the
full papers of all RCTs of spinal manipulation were
obtained before deciding whether they reported
psychological outcomes. Multiple publications of
the same study were identified and collated.
Quality assessment, data extraction and
A quality scoring system was used, which had
been developed from the list of quality criteria
used by the ‘Back Review Group’ of the Cochrane
collaboration,10but which did not discriminate
against pragmatic trials. The score ranges from
0 to 16, and has evidence of construct validity11
(Box 1). Trials were scored independently by two
reviewers (NW, MH) and any discrepancies resolved
by consensus. We derived change scores of psy-
chological outcomes for each intervention, and
calculated standardised mean differences (SMD)
using baseline standard deviations (S.D.). If neces-
sary, we extracted the data needed for calculations
from graphs. If the baseline S.D. was not reported
we used the population or the follow-up S.D., if
known. All analyses were conducted using Rev-
Man.12Clinical homogeneity, or the differences
between studies, were assessed by consideration
of the type and delivery of the intervention, the
Systematic review of psychological outcomes in spinal manipulation273
Box 1: Criteria for the pragmatic trial quality
P1Adequacy of randomisation
Score 0Non-random number generation or
Score 1Random numbers from table or
computer, but no safeguards against
repeated sequences of numbers
Score 2Truly unpredictable sequence of
P2Concealment of allocation
Score 0No allocation concealment or no
Score 1Sequentially numbered envelopes
Score 2Secure from all clinicians and
research assistants determining
eligibility. Remote telephone
P3Adequacy of sample size
Score 0Less than 50 patients per arm of the
Score 1Between 50 and 100 patients per arm
Score 2At least 100 patients per arm
P4 Suitability of outcome measurement
Score 0Visual analogue score, or ad hoc
collection of questions
Score 1Generic measure only
Score 2Appropriate condition-specific
P5Blinding of outcome assessment
Score 0Outcomes assessed by unblinded
clinician or researcher, or no detail
Score 1Patient completed outcome
measure, but no detail describing
blinding of data coding by researcher
Score 2Outcome assessment by blinded
clinician or researcher, outcome
measures labelled with coded
P6Adequacy of patient follow-up
Score 0Less than 1-month follow-up, or
more than 40% drop out rate, or drop
out rate not described
Score 1At least 3 months follow-up with less
than 40% drop out rate, or at least
1-month follow-up with less than
25% drop out rate
Score 2At least 3 month follow-up with less
than 25% drop out rate
P7Appropriateness of analysis
Score 0No intention to treat analysis
Score 1Insufficient information to determine
whether intention to treat
Score 2Intention to treat analysis
P8Blinding of data analysis
Score 0Analysis not blinded, or no
information about coding of data
Outcome assessment data coded,
but uncertainty as to whether the
code is broken before or after the
Outcome assessment data coded,
and the code is broken after the
statistical analysis is completed
setting, study population, and length of follow-up.
The different psychological outcome measures used
were assessed for compatibility. Results were com-
bined in meta-analyses when trials were deemed
to be sufficiently homogenous for pooling to make
sense. A statistical test for heterogeneity was
used to guide whether a fixed effects or random
effects model was used for pooling the trial results.
In a sensitivity analysis, meta-analyses were re-
calculated: after excluding trials with a quality
score of eight or less; after excluding trials which
examined regions other than the lower back; after
using the alternative model for data pooling; for
groups of trials comparing manipulation with indi-
vidual talking or physical treatments; for groups of
trials including or excluding chronic pain.
Identification of trials
The search strategy identified 2466 potentially rel-
evant titles and abstracts, which were screened for
potential inclusion. After reviewing the full text of
254 articles, 226 articles concerning 129 separate
trials of spinal manipulation were identified. Out of
these, 35 trials had used psychological instruments,
or generic or pain outcome measures with psycho-
logical sub-scales (Fig. 1). However, 22 trials did not
report usable data; 10 only used psychological out-
comes at baseline and 12 only reported total scores
of generic or pain outcome measures, and not their
psychological sub-scales. Thirteen RCTs reported
psychological outcomes at follow-up (Table 1); 12
in sufficient detail to calculate SMDs. Most of these
trials have been published in the last four years.
Trial homogeneity and quality
The 13 included trials were heterogeneous in terms
of the region of spine treated, profession of manip-
ulator used, the nature of the control group,
and psychological outcome reported (Table 1).
274N.H. Williams et al.
Figure 1 Systematic review flowchart.
Regions of the spine treated included the lower
back (n=8), the neck (n=4) and all of the spine
(n=1). Osteopaths, chiropractors, physiothera-
pists or physicians performed spinal manipulation,
mostly using a combination of mobilisation and high
velocity thrust techniques. A variety of psycholog-
ical outcomes were reported including depression
scores, fear-avoidance beliefs, back beliefs, gen-
eral health questionnaire, mental score or mental
health domains of the SF-36 or SF-12 generic
outcome measures. Control interventions could
be grouped into two categories: verbal interven-
tions including primary care, advice and education;
physical treatments including electrotherapy, sham
manipulation, muscle relaxants, cervical collar, and
surgery. The null hypothesis that the treatment
effect was the same in both groups of studies was
not rejected, so fixed effects models were used.
Most trials were high quality (Table 2).
Spinal manipulation versus verbal
Three trials used primary care as the comparator.
One of these compared manipulation for neck or
back pain with usual primary care in the United
Kingdom (UK)13,14; two compared manipulation for
low back pain with optimal primary care in the
UK15,16and Sweden.17—21Optimal care differed
from usual care because of consistent advice and
encouragement to remain active. The Swedish trial
single depression item in favour of spinal manipula-
tion. A t-test of the difference between the groups
Systematic review of psychological outcomes in spinal manipulation
Characteristics of included studies
N.H. Williams et al.
Table 1 (Continued)
Systematic review of psychological outcomes in spinal manipulation277
Quality scores of RCTs of spinal manipulation reporting psychological outcomes
was statistically significant (p=0.016), from which
the S.D. of the 4-month score was calculated. The
two higher quality trials from the UK13—16found an
improvement in the mental component score of the
SF-36 or SF-12 in favour of spinal manipulation. The
UK BEAM trial15,16also found a relative improve-
ment in back beliefs but not fear avoidance beliefs
in favour of spinal manipulation.
Two trials for low back pain used advice as a com-
parator. One trial from Finland compared muscle-
energy technique manipulation with an educational
booklet and advice from a physician27—30and found
a statistically significant improvement in depres-
sion score at 5 months in favour of manipulation.
The other trial from the UK found a small, but sta-
tistically significant improvement in SF-36 mental
component summary score at 2 months in favour of
manipulation by a physiotherapist compared with
advice on remaining active.31
One trial in the USA did not find a statistically
significant difference in Zung depression score after
4 weeks between chiropractic manipulation and a
back education programme.23,24
The six trials that could be included in a meta-
analysis were pooled using a fixed effects model.
There was a benefit in favour of manipulation with
SMD 0.34 [95%confidence interval (CI) 0.23—0.45]
at 1—5 months; 0.27 [95% CI 0.14—0.40] at 6—12
months (Fig. 2).
Spinal manipulation versus other physical
Two trials from the UK used an exercise programme
as a comparator. One trial compared manipula-
tion for neck pain with exercise advice and a
home exercise programme32and did not find a
statistically significant difference in SF-12 men-
tal score. A baseline S.D. was not reported, so
the known population S.D. of 10 was used to
calculate the SMD.9The other trial compared
manipulation for low back pain with group exercise
classes15,16and found a small statistically signifi-
cant difference in SF-36 mental score in favour of
manipulation at 3 months, a relative improvement
in fear avoidance beliefs in favour of exercise, but
similar changes in back beliefs between the two
Two trials from the UK using electrotherapy as
a comparator, did not find a statistically signifi-
cant difference in SF-36 mental health domain or
SF-12 mental score. One trial for acute low back
pain compared manipulation with interferential
therapy33,34; the other trial for neck pain compared
manipulation with pulsed short-wave diathermy.32
This latter trial did not report a baseline S.D., so the
278N.H. Williams et al.
Figure 2 Meta-analysis of trials of spinal manipulation compared with verbal interventions.
known population S.D. of 10 was used to calculate
Three low back pain trials from the USA used
a placebo-like intervention, sham manipulation,
as a comparator and did not find a statistically
significant difference after 4 weeks in SF-36 men-
tal health domain, or Zung depression score.22—26
Unpublished results were obtained from the authors
of one of these trials.22
One poor quality trial from the UK for neck
pain found an improvement in general health
questionnaire score at 6 months favouring mobil-
isation to acupuncture, both performed by a
One trial from the USA for subacute low back
pain did not find a statistically significant differ-
ence in Zung depression score between chiropractic
manipulation and muscle relaxant medication.25,26
One poor quality trial from Sweden compared
surgery with physiotherapy including gentle mobil-
isation for cervical radiculopathy35—38and found
that there was no difference in mood state,
although change scores were not reported in a for-
mat that could be used to calculate SMD.
Two poor quality trials used a cervical collar
as a comparator. One trial from Germany found
a lower score for depression and anxiety items in
a group receiving mobilisation, manipulation and
post-isometric relaxation compared to a group with
a cervical collar.39Although no baseline scores
were reported and it was not possible to calculate
change scores, it was assumed that both groups had
Figure 3 Meta-analysis of trials of spinal manipulation compared with other physical treatments.
Systematic review of psychological outcomes in spinal manipulation279
interventions, excluding neck pain trials, and comparing trials including and excluding chronic pain
Sensitivity analysis combining trials comparing spinal manipulation with individual verbal and physical
ComparatorFollow-up periodNumber of
Electrotherapy 6—8 weeks2 0.04
Sham manipulation1 month30.14
Low back pain only
High quality trials only
Random effects model
Random effects model
comparable scores at baseline, and that the SMD
calculated from the post-treatment scores should
be a good approximation. The other trial from Swe-
den compared a collar with physiotherapy including
gentle mobilisation for cervical radiculopathy.35—38
There was no difference in mood state, although
change scores were not reported.
It was possible to calculate SMDs from eight
of these trials, which were combined in a meta-
analysis using a fixed effects model. The SMD was
only 0.13 in favour of spinal manipulation at 1—5
months [95% CI 0.01—0.24] and 0.11 at 6—12 months
[95% CI −0.02 to 0.25] (Fig. 3).
There was minimal difference in SMD when either
neck pain or poor quality trials were excluded from
the meta-analyses, or when a random effects model
was used. Separate meta-analyses were performed
on groups of trials that used primary care, advice,
exercise, electrotherapy and sham manipulation as
comparators. There were small and medium sized
SMDs in favour of manipulation compared to pri-
mary care and advice, but no significant difference
compared to other physical treatments. As psy-
chological factors tend to be more prevalent in
chronic pain conditions, trials including and exclud-
ing chronic pain were compared (Table 3).
There was a small improvement in psychologi-
cal outcomes after spinal manipulation compared
with verbal interventions; a smaller improve-
ment compared with physical treatments, although
95% confidence intervals included zero at 6—
280N.H. Williams et al.
Strengths and weaknesses
This was the first systematic review of spinal manip-
ulation to examine psychological outcomes. Only
13 out of 129 trials reported any form of psycho-
logical outcome, and it was possible to calculate
SMD in 12 of these. Meta-analyses were performed
on two broad groups of RCTs, so must be inter-
preted with caution. When combining RCTs for
such a meta-analysis, there is a balance between
inappropriate lumping together of heterogeneous
RCTs,41and over cautious avoidance of combin-
ing RCT results with the loss of information that
data synthesis may provide to inform future sys-
tematic reviews.42These two groups of RCTs were
not statistically heterogeneous, and the sensitivity
analysis found similar results when pairs of trials of
individual treatments were combined, when neck
pain or poorer quality trials were excluded, or when
random effects models were used.
Most of the included trials reported either a
psychological component of a generic outcome, or
a single psychological outcome usually measuring
mood. The UK BEAM trial15,16reported both the
mental score of the SF-36 and two measures of
benefit of spinal manipulation has more to do with
improving back beliefs than fear avoidance beliefs.
It also had a much larger sample size than any of
the other trials.
The quality scoring system used was designed
to avoid discrimination against pragmatic trials;
however it might in turn discriminate against
explanatory trials. Three of the included trials were
explanatory, or had an explanatory component
as they compared spinal manipulation with sham
manipulation.22—26Nonetheless, all three scored
highly for quality (quality score 11 or 12/16). Sim-
ilar to other scoring systems, this one did not
differentiate between poor reporting and poor
trial quality43; however only four trials scored
poorly,19—21,35—40one of which did not report suffi-
cient detail to calculate the SMD, so was excluded
from the meta-analysis.35—38
Meaning of the study
What is the nature of the psychological benefit from
spinal manipulation compared to verbal interven-
tions? It could be due to a difference between the
characteristic effects of the compared treatments;
or their incidental (placebo) effects.44The latter
is supported by the finding that when the inciden-
tal (placebo) effects were equalised between the
interventions in three explanatory trials,22—26there
was no difference in treatment effect, but the lack
of statistical difference could also be due to small
sample size. Increased incidental (placebo) effects
could occur because of increased contact between
therapist and patient, or if patients’ preference for
treatment results in resentful demoralisation when
randomised to the less favoured option. If the back
or neck pain is secondary to a psychological dis-
turbance such as depression, the recognition that
there is something wrong or the acknowledgement
of the need for help may be beneficial. There may
be a difference in the caring effect of the thera-
pist due to a difference in empathy, understanding
of spinal pain, or explanation to the patient. It has
been argued that the use of explanatory trials in
complex interventions such as spinal manipulation
leads to false negative results, as the characteristic
and incidental effects are intertwined and diffi-
cult to disentangle, even when using placebo-like
treatments such as sham manipulation.45From this
perspective, the psychological effects of manipula-
tion are not just incidental effects, but contribute
to its characteristic treatment effect by reduc-
ing distressing symptoms such as pain and fear. A
recent study of prognostic factors following spinal
manipulation46found that the presence of fear
avoidance beliefs resulted in a worse outcome.
Interestingly, the UK BEAM trial15,16found that fear
avoidance beliefs were less likely to improve fol-
lowing spinal manipulation than after exercise.
Is spinal manipulation the best method of
improving psychological health in patients with
back and neck pain? Psychological interventions
have been examined in other systematic reviews.
Cognitive behavioural therapy (CBT) was more
effective than waiting list control for short-term
pain relief, but there was no difference in effec-
tiveness when behavioural components were added
to other treatments such as physiotherapy, back
education or medical treatment.47The average
effect size of CBT across all domains was approx-
imately 0.5, so is larger than that found for spinal
manipulation in this review.48Multidisciplinary,
biopsychosocial rehabilitation has a strong psycho-
logical component. Intensive rehabilitation with a
functional restoration approach improved pain and
function compared to other treatments, but less
intensive rehabilitation did not.49
Implications for research and clinical
Future systematic reviews should examine psy-
chological outcomes as well as pain and physical
function; future trials should consider the inclusion
of psychological outcomes, and should adequately
report their results. The psychological effect of
Systematic review of psychological outcomes in spinal manipulation281
spinal manipulation could be compared with that of
psychological treatments such as CBT, particularly
investigated further in an explanatory RCT compar-
ing the effects of spinal manipulation with a sham
or placebo-like control using a range of psychologi-
cal outcomes; although it is uncertain how feasible
it is to equalise the incidental (placebo) effects in
such trials. The clinical implications are that phys-
ical treatments, such as spinal manipulation have
psychological benefits, and in the words of the Chi-
nese sage Chen Jen; ‘‘When you treat disease, first
treat the mind.’’50
Declaration of interest
NHW is a registered osteopath and provides osteo-
pathic treatment as part of his work as a general
This systematic review was supported by the Stone
Forum, which receives a grant from Merck Sharpe
and Dohm. Thanks to John Licciardone for providing
additional unpublished results from his trial.
Appendix A. Medline literature search
strategies for randomised controlled
trials of spinal manipulation
randomized controlled trial.pt.
controlled clinical trial.pt.
Randomized Controlled Trials/
7 not 8
exp Clinical Trials/
(clin$ adj25 trial$).tw.
((singl$ or doubl$ or trebl$ or tripl$)
adj25 (blind$ or mask$)).tw.
(latin adj square).tw.
19 not 18
20 not 9
exp Evaluation Studies/
(control$ or prospective$ or
28 not 8
29 not (9 or 21)
9 or 21 or 30
exp Back Pain/
(lumbar adj pain).ti,ab.
exp Spinal Diseases/
(disc adj degeneration).ti,ab.
(disc adj prolapse).ti,ab.
(disc adj herniation).ti,ab.
(facet adj joints).ti,ab.
exp neck pain/
(back adj pain).ti,ab.
(spin$ adj25 pain).ti,ab.
(thoracic adj25 pain).ti,ab.
(cervical adj pain).ti,ab.
(lumbar adj trauma).ti,ab.
(lumbar adj pain).ti,ab.
exp cervical vertebrae/
exp musculoskeletal manipulations/
(spin$ adj25 manipulat$).ti,ab.
(orthop$ adj25 manipulat$).ti,ab.
(physiotherap$ adj25 manipulat$).ti,ab.
31 and 71 and 78
282N.H. Williams et al.
1. Cote P, Cassidy D, Carroll L. The Saskatchewan health and
back pain survey. The prevalence of neck pain and related
disability in Saskatchewan adults. Spine 1998;23:1689—98.
2. Clinical Standards Advisory Group [CSAG]. Back pain. Lon-
don: HMSO; 1994.
3. McCormick A, Fleming D, Charlton J. Morbidity statistics
from general practice: fourth national study 1991—1992.
London: Office of Population Censuses and Surveys; 1995.
4. Thomas KJ, Carr J, Westlake L, et al. Use of non-
orthodox and conventional health care in Great Britain. BMJ
5. Assendelft WJJ, Morton SC, Yu EI, et al. Spinal manipula-
tive therapy for low back pain. Cochrane Database Syst Rev
6. Gross AR, Hoving JL, Haines TA, et al. A Cochrane review
of manipulation and mobilisation for mechanical neck dis-
orders. Spine 2004;29:1541—8.
7. Linton SJ. A review of psychological risk factors in back and
neck pain. Spine 2000;25:1148—56.
8. Ware JE, Sherbourne CD. The MOS 36-item short-form
health survey (SF-36). I. Conceptual framework and item
selection. Med Care 1992;30:473—83.
9. Ware JE, Kosinski M, Keller SD. A 12-item short-form health
survey, construction of scales and preliminary tests of reli-
ability and validity. Med Care 1996;34:220—33.
10. van Tulder MW, Assendelft WJJ, Koes BW, et al. Method
guidelines for systematic reviews in the Cochrane col-
laboration back review group for spinal disorders. Spine
11. Williams NH. The effectiveness and efficiency of a primary
care based osteopathy clinic for spinal pain. PhD thesis.
Cardiff University, Wales College of Medicine; 2004.
12. Meta-View Rev-Man [computer programme]. Version 4.2.5.
Cochrane Collaboration; 2004.
13. Williams NH, Wilkinson C, Russell I, et al. Random-
ized Osteopathic MANipulation Study (ROMANS): pragmatic
trial for spinal pain in primary care. Family Pract
14. Williams NH, Edwards RT, Linck P, et al. Cost-utility analy-
sis of osteopathy in primary care: results from a pragmatic
randomized controlled trial. Family Pract 2004;21:643—50.
15. UK BEAM Trial Team. UK Back pain Exercise And Manip-
ulation (UK BEAM) randomised trial: effectiveness of
physical treatments for back pain in primary care. BMJ
16. UK BEAM Trial Team. UK Back pain Exercise And Manipu-
lation (UK BEAM) randomised trial: cost-effectiveness of
physical treatments for back pain in primary care. BMJ
17. Blomberg S, Svardsudd K, Mildenberger F. A controlled mul-
ticenter trial of manual therapy in low back pain. Scand J
Prim Health Care 1992;10:170—8.
18. Blomberg S, Svardsudd K, Tibblin G. Manual therapy with
steroid injections in low back pain. Scand J Prim Health
19. Blomberg S, Tibblin G. A controlled multicenter trial of
manual therapy with steroid injections in low back pain:
functional variables, side effects and complications during
four months follow up. Clin Rehab 1993;7:49—62.
20. Blomberg S, Hallin G, Grann K, et al. Manual therapy with
steroid injections—–a new approach to treatment of low
back pain. A controlled multicenter trial with evaluation
by orthopedic surgeons. Spine 1994;19:569—77.
21. Blomberg S, Svardsudd K, Tibblin G. A randomized study of
manual therapy with steroid injections in low back pain;
telephone interview follow-up of pain, disability, recovery
and drug consumption. Eur Spine J 1994;3:246—54.
ulative treatment for chronic low back pain: a randomized
controlled trial. Spine 2003;28:1355—62.
23. Triano JJ. A randomized controlled clinical trial of manip-
ulation for chronic low back pain patients. J Manipulative
Physiol Ther 1994;17:291.
24. Triano JJ, McGregor M, Hondras MA, et al. Manipulative
therapy versus education in chronic low back pain. Spine
25. Hoiriis KT, Pfleger B, McDuffie FC, et al. A randomized
clinical trial comparing chiropractic to conservative med-
ical care for subacute low back pain. Eur J Chiropract
26. Hoiriis KT, Pfleger B, McDuffie FC, et al. A randomized
clinical trial comparing chiropractic adjustments to mus-
cle relaxants for subacute low back pain. J Manipulative
Physiol Ther 2004;27:388—98.
27. Niemisto L, Lahtinen-Suopanki T, Rissanen P, et al. A
randomized trial of combined manipulation, stabilizing
exercises, and physician consultation compared to physi-
cian consultation alone for chronic low back pain. Spine
28. Niemisto L, Sarna S, Lahtinen-Suopanki T, et al. Predictive
factors for 1-year outcome of chronic low back pain fol-
lowing manipulation, stabilizing exercises, and physician
consultation or physician consultation alone. J Rehab Med
29. Riipinen M, Niemisto L, Lindgren K-A, et al. Psychosocial
differences as predictors for recovery from chronic low
back pain following manipulation, stabilizing exercises and
physician consultation or physician consultation alone. J
Rehab Med 2005;37:152—8.
30. Niemisto L, Rissanen P, Sarna S, et al. Cost-effectiveness of
combined manipulation, stabilising exercises, and physician
consultation compared to physician consultation alone for
chronic low back pain: a prospective randomized trial with
2 year follow up. Spine 2005;30:1101—5.
31. Frost H, Lamb SE, Doll HA, et al. Randomised controlled
trial of physiotherapy compared with advice for low back
pain. BMJ 2004;329:708.
32. Dziedzic K, Hill J, Lewis M, et al. Effectiveness of man-
ual therapy or pulsed shortwave diathermy in addition to
advice and exercise for neck disorders: a pragmatic ran-
domized controlled trial in physical therapy clinics. Arth
33. Hurley DA, McDonough SM, Dempster M, et al. A randomized
clinical trial of manipulative therapy and interferen-
tial therapy for acute low back pain. Spine 2004;29:
34. Hurley DA, McDonough SM, Baxter GD, et al. A descriptive
study of the usage of spinal manipulative therapy tech-
niques within a randomized clinical trial in acute low back
pain. Manual Ther 2005;10:61—7.
35. Persson LCG, Moritz U, Brandt L, et al. Cervical radiculopa-
thy: pain, muscle weakness and sensory loss in patients with
cervical radiculopathy treated with surgery, physiotherapy,
or cervical collar. A prospective controlled study. Eur Spine
36. Persson LCG, Karlberg M, Magnusson M. Effects of differ-
ent treatments on postural performance in patients with
cervical root compression. A randomized prospective study
assessing the importance of the neck in postural control. J
Vestibular Res 1996;6:439—53.
37. Persson LCG, Carlsson CA, Carlsson JY. Long-lasting cer-
vical radicular pain managed with surgery, physiotherapy,
Systematic review of psychological outcomes in spinal manipulation283
or a cervical collar. A prospective randomized study. Spine
38. Persson LCG, Lilja A. Pain, coping, emotional state and
physical function in patients with chronic radicular neck
pain. A comparison between patients treated with surgery,
physiotherapy, or neck collar—–a blinded prospective ran-
domized study. Disability Rehab 2001;23:325—35.
39. Giebel GD, Edelmann M, Huser R. Die distortion der hal-
swirbelsaule: Fruhfunktionalle vs. ruhigstellende behand-
lung. Zentralbl Chir 1997;122:517—21.
40. David J, Modi S, Aluko AA, et al. Chronic neck pain: a com-
parison of acupuncture treatment and physiotherapy. Br J
41. Egger M, Dickersin K, Davey Smith G. Problems and limita-
tions in conducting systematic reviews. In: Egger M, Davey
Smith G, Altman DG, editors. Systematic reviews in health
care: meta-analysis in context. 2nd ed. London: BMJ; 2001.
42. Edwards A, Hood K, Matthews E, et al. The effectiveness
of one-to-one risk communication interventions in health-
care: a systematic review. Med Decision Making 2000;20:
43. Juni P, Altman DG, Egger M. Assessing the quality of ran-
domised controlled trials. In: Egger M, Davey Smith G,
Altman DG, editors. Systematic reviews in health care:
meta-analysis in context. 2nd ed. London: BMJ; 2001. p.
44. Grunbaum A. Explication and implications of the placebo
context. In: White L, Tursky B, Schwartz GE, editors.
Placebo theory, research, and mechanisms. New York:
Guildford Press; 1985. p. 9—36.
45. Paterson C, Dieppe P. Characteristic and incidental
(placebo) effects in complex interventions such as acupunc-
ture. BMJ 2005;330:1202—5.
46. Childs JD, Fritz JM, Flynn TW, et al. A clinical prediction
rule to identify patients with low back pain most likely to
benefit from spinal manipulation: a validation study. Ann
Intern Med 2004:920—8.
47. Ostelo RWJG, van Tulder MW, Vlaeyen JWS, et al.
Behavioural treatment for chronic low-back pain. Cochrane
Database Syst Rev 2005;(1).
48. Morley S, Eccleston C, Williams A. Systematic review
of randomised controlled trials and behavioural ther-
apy for chronic pain in adults, excluding headache. Pain
49. Guzman GJ, Esmail R, Karjalainen K, et al. Multidisciplinary
rehabilitation for chronic low back pain: systematic review.
50. McDonald P. Oxford dictionary of medical quotations.
Oxford: Oxford University Press; 2004.
Available online at www.sciencedirect.com