Available via license: CC BY-NC 4.0
Content may be subject to copyright.
1
GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
Reporting of adverse events associated
with spinal manipulation in randomised
clinical trials: an updated systematic
review
Lindsay M Gorrell ,1 Benjamin T Brown,2 Roger Engel,2 Reidar P Lystad 3
To cite: GorrellLM, BrownBT,
EngelR, etal. Reporting of
adverse events associated
with spinal manipulation in
randomised clinical trials: an
updated systematic
review. BMJ Open
2023;13:e067526. doi:10.1136/
bmjopen-2022-067526
►Prepublication history and
additional supplemental material
for this paper are available
online. To view these les,
please visit the journal online
(http://dx.doi.org/10.1136/
bmjopen-2022-067526).
Received 17 August 2022
Accepted 12 April 2023
1Integrative Spinal Research
Group, Department of
Chiropractic Medicine, University
Hospital Balgrist and University
of Zurich, Zurich, Switzerland
2Department of Chiropractic,
Faculty of Medicine, Health and
Human Sciences, Macquarie
University, Sydney, New South
Wales, Australia
3Australian Institute of Health
Innovation, Faculty of Medicine,
Health and Human Sciences,
Macquarie University, Sydney,
New South Wales, Australia
Correspondence to
Dr Lindsay M Gorrell;
lindsaymary. gorrell@ uzh. ch
Original research
© Author(s) (or their
employer(s)) 2023. Re- use
permitted under CC BY- NC. No
commercial re- use. See rights
and permissions. Published by
BMJ.
ABSTRACT
Objectives To describe if there has been a change in
the reporting of adverse events associated with spinal
manipulation in randomised clinical trials (RCTs) since
2016.
Design A systematic literature review.
Data sources Databases were searched from March
2016 to May 2022: MEDLINE (Ovid), Embase, CINAHL,
ICL, PEDro and Cochrane Library. The following search
terms and their derivatives were adapted for each
platform: spinal manipulation; chiropractic; osteopathy;
physiotherapy; naprapathy; medical manipulation and
clinical trial.
Methods Domains of interest (pertaining to adverse
events) included: completeness and location of reporting;
nomenclature and description; spinal location and
practitioner delivering manipulation; methodological
quality of the studies and details of the publishing journal.
Frequencies and proportions of studies reporting on
each of these domains were calculated. Univariable and
multivariable logistic regression models were tted to
examine the effect of potential predictors on the likelihood
of studies reporting on adverse events.
Results There were 5399 records identied by the
electronic searches, of which 154 (2.9%) were included
in the analysis. Of these, 94 (61.0%) reported on adverse
events with only 23.4% providing an explicit description of
what constituted an adverse event. Reporting of adverse
events in the abstract has increased (n=29, 30.9%) while
reporting in the results section has decreased (n=83,
88.3%) over the past 6 years. Spinal manipulation was
delivered to 7518 participants in the included studies.
No serious adverse events were reported in any of these
studies.
Conclusions While the current level of reporting of
adverse events associated with spinal manipulation in
RCTs has increased since our 2016 publication on the
same topic, the level remains low and inconsistent with
established standards. As such, it is imperative for authors,
journal editors and administrators of clinical trial registries
to ensure there is more balanced reporting of both benets
and harms in RCTs involving spinal manipulation.
INTRODUCTION
The use of high- velocity, low- amplitude
(HVLA) spinal manipulation to treat spinal
pain and dysfunction is recommended
in clinical and best practice guidelines1–4
and is commonly used by several health-
care professions.5–7 Despite this, concerns
remain surrounding adverse events following
the intervention.8 9 Adverse events associ-
ated with spinal manipulation are typically
benign, transient and do not require further
treatment.10 Indeed, some authors classify
increased muscle soreness or stiffness in the
treatment area as an ‘expected outcome of
treatment’ rather than an adverse event.11 At
the other end of the spectrum, catastrophic
events, such as vertebral artery dissection,
have been temporally associated with spinal
manipulation.12 However, such events are
rare, and as a result, are typically reported
in individual case reports or case series with
little to no information regarding the inter-
vention that was delivered.13 Indeed, synthesis
of the current literature suggests that there is
no evidence for cervical spine manipulation
causing cervical artery dissection.14 Addition-
ally, several large population- based studies
have reported that there is no difference in
risk of cervical artery dissection following
visits to a chiropractor compared with those
occurring following a visit to a primary
care provider15 16 or, in those who received
cervical spinal manipulation compared with
matched controls.17 18 Furthermore, recent
STRENGTHS AND LIMITATIONS OF THIS STUDY
⇒This systematic review is reported following the
Preferred Reporting Items for Systematic Reviews
and Meta- Analysis (PRISMA) guidelines.
⇒The search strategy was inclusive of professions
that deliver spinal manipulation.
⇒The search included several databases relevant to
manual therapy.
⇒Due to heterogeneity of reporting of adverse events,
only descriptive statistics were used to describe do-
mains of interest.
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
2GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
biomechanical studies report that head angular displace-
ments and vertebral artery length changes are small
during cervical spine manipulation thrusts19 and that the
vertebral artery does not experience longitudinal force
during cervical spine manipulation.20 Despite this liter-
ature, the serious nature of such events that are tempo-
rally associated with cervical spine manipulation makes
it imperative that the circumstances surrounding such
events are reported transparently.
Randomised clinical trials (RCTs) are the gold stan-
dard study design for measuring effectiveness (benefit/s)
of interventions for the treatment of spinal pain and
dysfunction. However, as the risks of an intervention are
also important to both patients and practitioners, RCTs
should report on not only the efficacy of spinal manip-
ulation, but also any adverse events associated with the
intervention. The Consolidated Standards of Reporting
Trials (CONSORT) statement, first published in 1996
with several updates since, provides the scientific commu-
nity (specifically researchers and journal editors) with a
scaffold to standardise and improve the quality of RCT
reporting.21–23 The CONSORT statement acknowledges
the importance of reporting adverse events alongside
effectiveness data. The 2004 Harms extension docu-
ment24 provides specific recommendations for how and
where these data should be included in scientific manu-
scripts. However, reporting of adverse events in RCTs in
the wider medical literature remains insufficient since the
publication of the 2004 extension,25 a finding that is also
evident in RCTs that involve spinal manipulation.26 Thus,
the objective of this review was to describe if there has
been a change in the reporting of adverse events associ-
ated with spinal manipulation in RCTs since 2016.
METHODOLOGY
This systematic literature review is reported following the
Preferred Reporting Items for Systematic Reviews and
Meta- Analyses guidelines.27
Denitions
Spinal manipulation was defined as a manual procedure
involving an HVLA thrust delivered to a spinal joint with
the intention of moving the joint past its physiological
range of motion but without exceeding the anatomic
limit.28 For the purposes of this review, spinal manipu-
lation delivered using drop- piece- table and mechanical
implements (eg, Activator instrument) were considered
HVLA procedures.29
An adverse event was defined as any unfavourable reac-
tion with a temporal association to spinal manipulation
that resulted in an alteration in a participant’s activities of
daily living,30 31 irrespective of the timing of onset, dura-
tion or severity of the event.32
A serious adverse event was defined as any unfavour-
able sign, symptom or disease temporally associated with
the treatment, whether or not caused by the treatment
that results in death or is life threatening or results in
inpatient hospitalisation or prolongation of existing
hospitalisation for more than 24 hours with a persistent
or significant incapacity or substantial disruption of the
ability to conduct normal life functions.30
To be classified as reporting on adverse events ‘directly’,
a study must have provided explicit description of their
operational definition of an adverse event (eg, ‘In the
current study, an adverse event was defined as a sequelae
of 1- week duration with any symptom perceived as
distressing and unacceptable to the patient that required
further treatment [excerpt from reference 63].’33), and/
or how data on adverse events were measured (eg, ‘Active
and passive surveillance methods were used to collect
information on adverse events.’34), and/or provide a
substantial description of adverse events observed during
data collection.35 36 In contrast, all other studies reporting
on adverse events ‘indirectly’ did not explicitly provide
such information.
Patient and public involvement
No patients were involved in this systematic literature
review.
Eligibility criteria
Consistent with the 2016 review,26 RCTs reporting original
data on spinal manipulation as either the sole interven-
tion, or as the sole intervention in a comparator group,
delivered by any regulated health professional, and
published in English, were eligible for inclusion. Studies
reporting on reviews, other trial designs, trial registra-
tions, protocols, commentaries, editorials and confer-
ence proceedings were excluded. Further exclusion
criteria included retracted articles, secondary analyses,
studies in which the full text was not available in English
and studies where manipulation was only applied to an
area other than the spine. Studies were also excluded if it
was unclear if the intervention being delivered involved
an HVLA manipulation.
Search strategy
The following databases were searched from 1 March
2016 to 12 May 2022: MEDLINE (Ovid), Embase,
CINAHL, ICL, PEDro and Cochrane Library. Reference
lists of included studies were screened to insure all rele-
vant literature was captured. The following search terms
and derivatives were adapted for each platform: spinal
manipulation; chiropractic; osteopathy; physiotherapy;
naprapathy; medical manipulation and clinical trial. An
example of each search strategy is provided in online
supplemental appendix 1.
Study selection process
Records retrieved from the electronic searches were
exported to the Rayyan online platform.37 Duplicate
records, and records included in the 2016 review, were
removed before title and abstract screening. Two authors
(LMG and BTB) independently screened included
studies in a stepwise process, beginning with review of
each title and abstract. Full texts of the studies remaining
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
3
GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
after this step were retrieved and further screened against
the eligibility criteria (LMG and RE). Any disagreements
regarding inclusion were resolved by consensus and if
consensus could not be reached, disagreements were
resolved by a third author (BTB).
Data extraction
Adverse events reporting data were extracted from the
remaining studies by two authors (LMG and RPL).
These data included descriptive information (ie, title,
author, year of publication, country where the data was
collected, journal of publication, spinal region treated
(eg, cervical spine) and type of practitioner delivering
the spinal manipulation (eg, chiropractor)), whether
the study reported on adverse events (ie, reported/
not and if reported, directly/indirectly), location of
reporting within the article, classification of adverse
events reported (eg, mild, moderate, serious and severe),
completeness of adverse events reporting (ie, onset,
duration and number of events reported), number of
participants in the spinal manipulation group/s and
descriptions of any definitions and/or classification
systems used. Other data collated by the lead author
(LMG) included whether the study was published in
a journal that follows the International Committee of
Medical Journal Editors (ICMJE) guidelines via a search
of the ICMJE website38 on 29 May 2022. Additionally, the
most recently published impact factor (year 2020) for
each journal was manually extracted by the lead author
(LMG) from the Clarivate Journal Citations Reports
website39 on 29 May 2022.
Assessment of risk of bias using the Cochrane ROB V.2
assessment tool40 was performed by three authors working
in pairs (LMG and RE, and LMG and BTB) for all included
studies to assess the methodological quality of the publi-
cation. Disagreements were resolved by consensus and
if consensus could not be reached, disagreements were
resolved by a third author (RPL).
Figure 1 Preferred Reporting Items for Systematic Reviews and Meta- Analyses ow diagram.
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
4GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
Data analysis
Data were analysed using descriptive statistics. Frequen-
cies and proportions of studies reporting on each of the
specified domains above were calculated in Microsoft
Excel (V.2102). Continuous variables with highly skewed
distributions (ie, journal impact factor and sample size of
spinal manipulation group) were categorised into tertiles.
Univariable and multivariable logistic regression models
were fitted to examine the effect of potential predictors
on the likelihood of studies reporting on adverse events.
The multivariable logistic regression model was fitted
using backward elimination, whereby the least significant
potential predictors were sequentially eliminated from
the multivariable model until only significant predictors
remained. The observed effects from the univariable and
multivariable logistic regression models were reported as
ORs and adjusted ORs (aORs), respectively, with 95% CIs.
All statistical analyses were performed using the statistical
computing software R V.4.0.3 (The R Foundation for
Statistical Computing, Vienna, Austria).
RESULTS
There were 5399 records initially identified by the elec-
tronic searches (figure 1). A total of 3363 unique records
remained after de- duplication (n=2034) and the removal
of retracted articles (n=2). After title and abstract
screening, full texts of the 452 remaining studies were
screened. Of these, 154 fulfilled the eligibility criteria and
were included in the analysis (see online supplemental
appendix 2). The most common reasons for exclusion
were: the intervention did not consist of HVLA spinal
manipulation (n=163) and/or the study related to a
conference proceeding (n=49).
Comprehensiveness of reporting of adverse events
Of the 154 included studies, 94 (61.0%) reported on
adverse events. Of these 94 studies, 36 (38.3%) directly
reported on adverse events, with studies in which spinal
manipulation was delivered by a chiropractor most
frequently reporting these data (n=17; 47.2%, table 1).
Indirect reporting occurred in 58 studies (61.7%), with
studies in which spinal manipulation was delivered by a
physiotherapist being the most frequent (n=29; 50.0%,
table 1). Of the 60 studies (39.0%) that did not report
on adverse events, studies in which spinal manipula-
tion was delivered by a physiotherapist were the most
frequent (n=28; 46.7%, table 1). A description of what
constituted an adverse event definition and/or the classi-
fication system used was provided in 22 studies (23.4%).
However, most studies did not provide a description and
instead used terms such as ‘adverse event’ (n=70, 74.5%),
‘adverse effect’ (n=22, 23.4%), ‘’side effect’ (n=19, 20.2%)
and ‘harm’ (n=11, 11.7%) without adequate explana-
tion. When mentioned, terms pertaining to classification
systems (predominantly severity) were (number of studies
in which the term was used, %): ‘mild’ (n=20, 21.3%),
‘moderate’ (n=17, 18.1%), ‘serious’ (n=27, 28.7%) and
‘severe’ (n=14, 14.9%). The onset of an adverse event/s
was unclear in 30 (31.9%) studies. Duration of adverse
events were reported heterogeneously, with some studies
providing a time from either baseline or the start of inter-
vention, whereas others provided a temporal descriptor
such as ‘short- term’, ‘temporary’ or ‘transient’. Of the
9 studies providing times, durations were as follows:
<72 hours (n=3, 3.2%), >72 hours (n=2, 2.1%) or mixed
duration (n=4, 4.3%). An evaluation tool was mentioned
in 26 (27.7%) studies.
Number and location of adverse events reporting
No serious adverse events were reported in any of the
154 included studies, representing 7518 participants who
received spinal manipulation. Furthermore, of the 94
studies reporting on adverse events, 63 (67.0%) reported
that no adverse events occurred. Adverse events were
reported in the abstract of 29 (30.9%) and results section
of 83 (88.3%) studies. Furthermore, adverse events were
mentioned in several locations throughout the included
studies: the introduction (n=15, 16.0%), methods (n=56,
59.6%), discussion (n=30, 31.9%), conclusion (n=7,
7.4%) and supplementary materials (n=1, 1.1%).
Descriptors of studies reporting on adverse events
Descriptive statistics are provided in table 2. Of the 94
studies reporting on adverse events, 55 (58.5%) were rated
at a ‘high risk of bias’, 29 (30.9%) as ‘some concerns’ and
10 (10.6%) at a ‘low risk of bias’ (online supplemental
appendix 3). Additionally, 33 (35.1%) were published in
journals stating that they follow the ICMJE recommenda-
tions. For the remaining studies, the median of the most
recently published (2020) impact factor was 2.5 (IQR:
2.1–4.2).
Predictors for the reporting of adverse events
There was very strong evidence that studies with an
impact factor in the upper (aOR: 5.72 (95% CI 2.23 to
15.85); p<0.001) and middle (aOR: 3.52 (95% CI 1.51
to 8.57); p=0.004) tertiles were more likely to report on
Table 1 Comprehensiveness of reporting of AEs by
provider delivering the intervention
Directly
reports on
AE (n=36),
n (%)
Indirectly
reports on
AE (n=58),
n (%)
Does not
report on
AE (n=60),
n (%)
Chiropractor 17 (47.2) 12 (20.7) 7 (11.7)
Medical practitioner 1 (2.8) 4 (6.9) 5 (8.3)
Mixed 7 (19.4) 7 (12.1) 7 (11.7)
Naprapath 0 (0.0) 0 (0.0) 1 (1.7)
Osteopath 4 (11.1) 2 (3.4) 9 (15.0)
Physiotherapist 6 (16.7) 29 (50.0) 28 (46.7)
Unclear 1 (2.8) 4 (6.9) 3 (5.0)
AE, adverse event;
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
5
GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
adverse events than those in the lower tertile when the
model was adjusted for risk of bias, impact factor, spinal
region of manipulation and number of participants
receiving spinal manipulation (table 3). There was also
strong evidence that studies in which a chiropractor deliv-
ered the spinal manipulation were more likely to report
on adverse events (aOR: 4.58 (95% CI 1.14 to 20.24);
p=0.036). Studies in which spinal manipulation was deliv-
ered to more than one region or, it was unclear which
regions the manipulations were delivered, were also
more likely to report on adverse events (aOR: 3.18 (95%
CI 1.16 to 9.05); p=0.027). While not achieving statistical
significance, another factor of note included studies in
which cervical spine manipulation was delivered (aOR:
3.04 (95% CI 0.88 to 11.30); p=0.085).
DISCUSSION
There has been a change in the reporting of adverse
events associated with spinal manipulation in RCTs since
2016. Specifically, the percentage of included studies
reporting adverse events has increased from 38.0%
(2016 study26) to 61.0% (current study). However, the
current review highlights that the reporting of adverse
events in RCTs involving spinal manipulation as an
intervention remains poor and is not consistent with
established standards. Specifically, of the 154 included
studies, just over half (n=94, 61.0%) reported on adverse
events. Furthermore, of these 94 studies, less than half
(38.3%) reported directly on adverse events, with only
23.4% providing an explicit description of what consti-
tuted an adverse event. Further complicating this issue is
the vast heterogeneity of terms (ie, ‘adverse effect’, ‘side
effect’, ‘harm’, etc) used to describe adverse events.
This is disappointing given that there have been many
calls in the literature for the improvement of adverse
events reporting in RCTs, and for the development
and use of standardised definitions and classification
systems.24 26 32 41–46
Table 2 Characteristics of included studies by reporting on AEs
Overall (n=154),
n (%)
Reports on AE (n=94),
n (%)
Does not report on AE
(n=60), n (%)
ICMJE journal
Published in ICJME journal 53 (34.4) 33 (35.1) 20 (33.3)
Risk of bias
Low risk 13 (8.4) 10 (10.6) 3 (5.0)
Some concerns 47 (30.5) 29 (30.9) 18 (30.0)
High risk 94 (61.0) 55 (58.5) 39 (65.0)
Impact factor
Upper tertile 47 (30.5) 36 (38.3) 11 (18.3)
Middle tertile 54 (35.1) 37 (39.4) 17 (28.3)
Lower tertile 53 (34.4) 21 (22.3) 32 (53.3)
Spinal region
Cervical 24 (15.6) 17 (18.1) 7 (11.7)
Thoracic 33 (21.4) 15 (16.0) 18 (30.0)
Lumbopelvic 28 (18.2) 13 (13.8) 15 (25.0)
Mixed/unclear 69 (44.8) 49 (52.1) 20 (33.3)
Type of practitioner
Chiropractor 36 (23.4) 29 (30.9) 7 (11.7)
Osteopath 15 (9.7) 6 (6.4) 9 (15.0)
Physiotherapist 63 (40.9) 35 (37.2) 28 (46.7)
Medical practitioner 9 (5.8) 4 (4.3) 5 (8.3)
Mixed/other/unclear 31 (20.1) 20 (21.2) 11 (18.3)
Sample size spinal manipulation group*
Upper tertile 51 (33.3) 40 (42.6) 11 (18.6)
Middle tertile 50 (32.7) 28 (29.8) 22 (37.3)
Lower tertile 52 (34.0) 26 (27.7) 26 (44.1)
*One study with unclear sample size excluded.
AE, adverse event;
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
6GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
A recent scoping review explores the complexity of the
current literature reporting on adverse events associated
with spinal and peripheral joint manipulation and mobil-
isation.47 Specifically, the authors report that conflicting
opinions regarding facets of adverse event definition
and classification such as: symptom severity and dura-
tion, relatedness to the intervention (eg, time to onset
and treatment provided), action taken to treat the symp-
toms and expectedness, which profession delivered the
intervention and geographical location (with possible
medico- legal constraints and/or different expectations of
reporting/not reporting), are all factors to reflect on when
considering adverse events associated with joint manipula-
tion and mobilisation. In an attempt to address the lack of
standardised definitions and classification systems across
professions that deliver spinal manipulation, the same
authors have conducted an international Delphi study
(manuscript in preparation; protocol paper41) to deter-
mine, by expert consensus, a standardised definition and
severity classification for adverse events associated with
spinal and peripheral joint manipulation and mobilisa-
tion. The development and use of such guidelines would
constitute an important step toward uniform reporting of
adverse events associated with spinal manipulation across
all stakeholder professions and geographical locations.
However, until this work is published, online supple-
mental appendix 2 of the 2004 CONSORT Harms exten-
sion24 provides a checklist of items to include and specific
examples of good reporting when reporting on harms
(adverse events) in RCTs. Furthermore, it appears that
an update to this guideline is emergent.25 It is hoped
that these updated guidelines will ensure that authors
and journal editors alike are both aware of and imple-
ment better harms reporting in the future. We strongly
Table 3 Univariable and multivariable logistic regression
Variable OR 95% CI P value aOR* 95% CI P value
ICMJE journal
Yes 1.08 0.55 to 2.16 0.821 – – –
No† – – – – – –
Risk of bias
Low risk 2.36 0.67 to 11.01 0.213 – – –
Some concerns 1.14 0.56 to 2.37 0.716 – – –
High risk† – – – – – –
Impact factor
Upper tertile 4.99 2.14 to 12.32 <0.001 5.72 2.23 to 15.85 <0.001
Middle tertile 3.32 1.52 to 7.48 0.003 3.52 1.51 to 8.57 0.004
Lower tertile† – – – – – –
Spinal region
Cervical 2.80 0.91 to 9.27 0.080 3.04 0.88 to 11.30 0.085
Thoracic 0.96 0.35 to 2.66 0.939 1.09 0.34 to 3.45 0.887
Lumbopelvic† – – – – – –
Mixed/unclear 2.83 1.15 to 7.11 0.025 3.18 1.16 to 9.05 0.027
Type of practitioner
Chiropractor 6.21 1.71 to 24.85 0.007 4.58 1.14 to 20.24 0.036
Osteopath† – – – – – –
Physiotherapist 1.88 0.60 to 6.19 0.282 1.35 0.37 to 5.18 0.648
Medical practitioner 1.20 0.22 to 6.53 0.831 0.81 0.12 to 5.47 0.829
Mixed/other/unclear 2.72 0.78 to 10.17 0.121 2.26 0.57 to 9.64 0.253
Sample size spinal manipulation group‡
Upper tertile 3.64 1.57 to 8.87 0.003 – – -
Middle tertile 1.27 0.58 to 2.79 0.544 – – -
Lower tertile† – – – – – –
*The nal model was adjusted for impact factor, spinal region of manipulation, and type of practitioner, while ICMJE journal status, risk of bias
and number of participants receiving spinal manipulation were omitted via backward elimination method.
†Reference group.
‡One study with unclear sample size excluded.
aOR, adjusted OR; ICMJE, International Committee of Medical Journal Editors.
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
7
GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
encourage researchers and journal editors alike to read
and use the most recent CONSORT Harms checklist
during all phases of study development, data collection,
manuscript preparation, submission and during the
review process. One important item on this checklist is
that both benefits and harms should be stated in either
the title and/or abstract of a manuscript. This point is
salient as the abstract is the second- most read section of a
scientific manuscript after the title.48 Encouragingly, the
reporting of adverse events in the abstract has doubled
(15.7%–30.9%, 2016–current) when compared with our
previous review of the literature.26 Despite this, the current
reporting on adverse events in the title/abstract of RCTs
using spinal manipulation remains poor, a finding that
is also present in the wider published medical literature
discussing adverse events.49–52 Despite an overall increase
in the number of studies reporting on adverse events
in RCTs involving spinal manipulation (38.0%–61.0%,
201626–current), adverse events reporting in the results
section has decreased (93.6% vs 88.3%) over the past 6
years and remains lower than that in the wider published
literature.50 53 It is unknown why there would be a decrease
in the reporting on adverse events associated with spinal
manipulation in one section of a scientific manuscript
that it could reasonably be expected to be reported.
Furthermore, an important source of information for the
formulation of a considered evidence- based risk- benefit
analysis for the use of spinal manipulation as a treatment
option by both clinician and patient49 52 is transparent
data reporting on both the efficacy and adverse events
occurring in RCTs involving spinal manipulation.
Consistent with the literature,31 32 42 43 47 there was
considerable heterogeneity of nomenclature used to
describe adverse events associated with spinal manipula-
tion. Similar terms were used to indicate an adverse event
in the current (compared with 2016) review: ‘adverse
event’ (2016—73.0%; 2022—74.5% of studies), ‘adverse
effect’ (23.6%; 23.4%), ‘side effect’ (21.3%; 20.2%) and
‘harm’ (16.4%; 11.7%). Additionally, while similar terms
were used to describe classification systems previously
reported (ie, ‘serious’, ‘mild’, ‘moderate’ and ‘severe’),
these terms were rarely defined, which is consistent with
the existing literature.26 52 Additionally, when present,
the reporting of onset and duration of adverse events was
inconsistent, again highlighting that there is an urgent
need for the development of a standardised definition
and classification system for the reporting of adverse
events.41 Furthermore, the responsibility for improved
reporting of adverse events falls not only to authors but
also to custodians of clinical trial registries and journal
editors to ensure that there are provisions in study proto-
cols for the adequate capture of adverse events and also
that these events are adequately reported. that is, using
the most recent CONSORT Harms extension guide-
lines,24 alongside efficacy/effectiveness data.25 46 54
Manuscript reviewers and journal editors must be
aware of the current best practices for the reporting of
harms24 and enforce these guidelines during peer review
processes of both protocol and end- of- study results
papers. However, this may not be as straight- forward as
it appears. Despite this, there is a need for improved
reporting of adverse events in RCTs that include spinal
manipulation as an intervention and a first step would be
for journals to incorporate clear instructions on harms
reporting in their guidelines and instructions to authors.
As a second step, journal editors may facilitate this process
by limiting publication to only those studies that adhere
to the current guidelines for the reporting of harms in
RCTs that include spinal manipulation as an intervention.
Indeed, if this was to occur, authors would need to ‘step-
up’, to use expanded methodologies, reporting and statis-
tical analyses that allow for the capture and reporting of
adverse events data in RCTs that include spinal manip-
ulation as an intervention. Specifically, data on adverse
events should be actively collected as it has been reported
that passive surveillance leads to an under- reporting25 54
and appropriate statistical analysis plans should be used
to analyse the data.49 54 55 As a minimum standard, authors
should explicitly state whether active or passive surveil-
lance systems were used.46 49
RCTs published in journals with a higher impact factor,
in which spinal manipulation was delivered by a chiro-
practor and to multiple/unclear regions, were more likely
to report on adverse events. While it is perhaps intuitive
that better designed studies, that is, those at a lower risk
of bias, could reasonably be published in higher impact
journals, this does not appear to be the case as there
was no influence of risk of bias level in the final model.
This disconnect between the publication of studies with
better methodological quality in higher impact jour-
nals is also seen in the medical literature. Specifically,
a previous study reported that there were methodolog-
ical weaknesses in 184 studies published in 2015–2016
by 4 of the top ranked general medical journals (BMJ,
JAMA, Lancet and NEJM).54 Furthermore, while there is
no obvious reason why studies in which spinal manip-
ulation was delivered by a chiropractor would be more
likely to report on adverse events, possible reasons for this
finding could include that chiropractors are more likely
to deliver cervical spine manipulation in general and/or
that due to perceived ‘risks’ of cervical spine manipula-
tion, other professions choose not to conduct trials inves-
tigating this intervention. This hypothesis is suggested by
the data which shows that while not achieving statistical
significance, studies in which cervical spine manipula-
tion was delivered had approximately three times greater
odds of reporting on adverse events. It is possible that this
result did not achieve statistical significance due to the
relatively small number of studies reporting on manip-
ulation delivered only to the cervical spine. Regarding
the increased likelihood of studies reporting on adverse
events if spinal manipulation was delivered to multiple/
unclear regions, it is possible that this finding is spurious
as there was a larger number of studies (n=49) in this
category compared with studies in which the interven-
tion was delivered to a single region. This hypothesis is
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
8GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
supported by a secondary analysis of our previous review
which reported that the region treated was not a signifi-
cant predictor for reporting on adverse events.56
Due to the methodological design of the review, we are
unable to comment on the incidence of adverse events
associated with spinal manipulation. Furthermore, RCTs
are not necessarily the best research design for collecting
data on serious adverse events as they often have strict
inclusion criteria and may exclude participants who are
at risk of experiencing such events. Additionally, RCTs are
powered to detect intervention effects and thus are likely
to be underpowered for estimating the risk of serious
adverse events. Despite this, the consistent reporting of
the number of spinal manipulations delivered to every
participant in RCTs could allow for the calculation of
accurate incidence rates for all classifications of adverse
events (serious included) and could eventually facilitate
the pooling of data across multiple studies thus allowing
for a better informed risk- benefit assessment of spinal
manipulation.25 46 We acknowledge that the calculation of
accurate incidence rates is not straight- forward. Indeed,
factors such as the use of different spinal manipulation
techniques, how to parse out adverse events attributable
to different interventions (eg, orthopaedic testing, soft
tissue treatment or exercise) and how to amalgamate
reports on different cohorts (eg, neck vs low back pain)
must all be considered. While this task seems insurmount-
able, consistent reporting of the number of spinal manip-
ulations delivered to every participant in RCTs is the first
step towards this goal. To this end, the number of spinal
manipulations delivered was only available in 75 (48.7%)
of the included studies. Coupled with the implementa-
tion of standardised definitions and classification systems
for adverse events associated with spinal manipulation,
reporting on the number of spinal manipulations deliv-
ered in each study could allow for the interdisciplinary
calculation of incidence rates for all classifications across
all healthcare professionals delivering the intervention.
Such an outcome is extremely important in the context
of obtaining informed consent to deliver spinal manip-
ulation. Specifically, in many countries in which spinal
manipulation is delivered, the process of obtaining
informed consent requires the disclosure of all mate-
rial information that a reasonable patient would require
to make an informed decision about whether or not to
receive that intervention.57 In the absence of accurate
incidence rates for the different classifications of adverse
events associated with spinal manipulation, this is a diffi-
cult task for the clinician to perform.
There are several differences between the current
review and our 2016 review.26 Specifically, the current
review included an improved search strategy, including
both an expansion to the number of databases searched
(ie, MEDLINE (Ovid), Embase, CINAHL and ICL were
added) in addition to the inclusion of several search
terms that did not limit the search to spinal manipula-
tion delivered by chiropractors and osteopaths (ie, phys-
iotherapists, naprapaths and medical manipulation were
added). Additionally, the current review reports on anal-
yses that we had previously reported separately in two
manuscripts: the original review26 and a secondary anal-
ysis.56 By reporting these analyses in a single manuscript,
we hope it is clearer for readers to identify that the current
level of reporting of adverse events associated with spinal
manipulation in RCTs is both poor and not consistent
with established standards, and understand the possible
explanations for this observation. By streamlining the
dissemination of this information, we hope to make it
easier for readers to identify areas in which researchers
may improve the reporting of adverse events in this field.
Limitations
There are several limitations to this literature review. First,
the decision to classify the reporting of adverse events as
‘direct’ (explicit description of operational definition of
an adverse event provided and/or how data on adverse
events were measured and/or a substantial descrip-
tion of adverse events observed during data collection
provided) as opposed to ‘indirect’ (no explicit reporting
of such information) was arbitrary. However, this classifi-
cation did not influence whether the study reported on
adverse events or not. As such, we do not feel this factor
had any material influence on our results. Second, as
outlined above, small differences in the methodology
between the current and previous reviews26 56 mean that
it is not possible to directly compare all reported findings
between the two reviews. However, as these differences
occurred due to methodological improvements in the
current review, we do not believe this affected the results
and/or discussion in the current review.
CONCLUSION
While the current level of reporting of adverse events
associated with spinal manipulation in RCTs has increased
since our 2016 publication on the same topic, the level
remains low and inconsistent with established standards.
As such, it is imperative for authors, journal editors and
administrators of clinical trial registries to ensure there
is more balanced reporting of both benefits and harms
of spinal manipulation in RCTs. We strongly recommend
that authors adhere to the most recent CONSORT Harms
checklist when reporting their results and advocate for
the creation of standardised definitions and classifica-
tion systems relating to adverse events in manual therapy.
This will facilitate the future pooling of adverse events
data across all professions sing spinal manipulation and
improve the ability to calculate incidence rates for the
different levels of adverse events.
Twitter Reidar P Lystad @RLystad
Acknowledgements The authors would like to acknowledge Dr Martina Gosteli for
her assistance with the literature search.
Contributors LMG: conceptualisation, screening, risk of bias assessment, data
extraction and curation, formal analysis, methodology, project administration,
visualisation and writing—original draft, review and editing and guarantor. RPL:
data extraction and curation, formal analysis, methodology, visualisation and
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
9
GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
writing—original draft, review and editing. BTB: screening, risk of bias assessment
and writing—review and editing. RE: screening, risk of bias assessment,
methodology and writing—review and editing.
Funding The authors have not declared a specic grant for this research from any
funding agency in the public, commercial or not- for- prot sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in
the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not applicable.
Ethics approval Ethics approval was not required for this systematic literature
review.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request. Data are
available from the corresponding author upon reasonable request.
Supplemental material This content has been supplied by the author(s). It has
not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been
peer- reviewed. Any opinions or recommendations discussed are solely those
of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and
responsibility arising from any reliance placed on the content. Where the content
includes any translated material, BMJ does not warrant the accuracy and reliability
of the translations (including but not limited to local regulations, clinical guidelines,
terminology, drug names and drug dosages), and is not responsible for any error
and/or omissions arising from translation and adaptation or otherwise.
Open access This is an open access article distributed in accordance with the
Creative Commons Attribution Non Commercial (CC BY- NC 4.0) license, which
permits others to distribute, remix, adapt, build upon this work non- commercially,
and license their derivative works on different terms, provided the original work is
properly cited, appropriate credit is given, any changes made indicated, and the use
is non- commercial. See:http://creativecommons.org/licenses/by-nc/4.0/.
ORCID iDs
Lindsay MGorrell http://orcid.org/0000-0002-1937-741X
Reidar PLystad http://orcid.org/0000-0003-0506-0902
REFERENCES
1 Whalen W, Farabaugh RJ, Hawk C, etal. Best- practice
recommendations for chiropractic management of patients with neck
pain. J Manipulative Physiol Ther 2019;42:635–50.
2 Bussières AE, Stewart G, Al- Zoubi F, etal. Spinal manipulative
therapy and other conservative treatments for low back pain: a
guideline from the Canadian chiropractic guideline initiative.
J Manipulative Physiol Ther 2018;41:265–93.
3 Bussières AE, Stewart G, Al- Zoubi F, etal. The treatment of neck
pain- associated disorders and whiplash- associated disorders:
a clinical practice guideline. J Manipulative Physiol Ther
2016;39:523–64.
4 Oliveira CB, Maher CG, Pinto RZ, etal. Clinical practice guidelines
for the management of non- specic low back pain in primary care:
an updated overview. Eur Spine J 2018;27:2791–803.
5 Beliveau PJH, Wong JJ, Sutton DA, etal. The chiropractic profession:
a scoping review of utilization rates, reasons for seeking care, patient
proles, and care provided. Chiropr Man Therap 2017;25.
6 Lin I, Wiles L, Waller R, etal. What does best practice care for
musculoskeletal pain look like? eleven consistent recommendations
from high- quality clinical practice guidelines: systematic review. Br J
Sports Med 2020;54:79–86.
7 National Institute for Health and Care Excellence (NICE). Low back
pain and sciatica in over 16S: assessment and management; 2016.
8 Puentedura EJ, March J, Anders J, etal. Safety of cervical spine
manipulation: are adverse events preventable and are manipulations
being performed appropriately? A review of 134 case reports. J Man
Manip Ther 2012;20:66–74.
9 Biller J, Sacco RL, Albuquerque FC, etal. Cervical arterial
dissections and association with cervical manipulative therapy: a
statement for healthcare professionals from the American heart
association/american stroke association. Stroke 2014;45:3155–74.
10 Funabashi M, Pohlman KA, Goldsworthy R, etal. Beliefs, perceptions
and practices of chiropractors and patients about mitigation
strategies for benign adverse events after spinal manipulation
therapy. Chiropr Man Therap 2020;28.
11 Heneghan NR, Davies SE, Puentedura EJ, etal. Knowledge and pre-
thoracic spinal thrust manipulation examination: a survey of current
practice in the UK. J Man Manip Ther 2018;26:301–9.
12 Albuquerque FC, Hu YC, Dashti SR, etal. Craniocervical arterial
dissections as sequelae of chiropractic manipulation: patterns of
injury and management. JNS 2011;115:1197–205.
13 Ernst E. Deaths after chiropractic: a review of published cases. Int J
Clin Pract 2010;64:1162–5.
14 Church EW, Sieg EP, Zalatimo O, etal. Systematic review and
meta- analysis of chiropractic care and cervical artery dissection: no
evidence for causation. Cureus 2016;8:e498.
15 Cassidy JD, Boyle E, Côté P, etal. Risk of vertebrobasilar stroke and
chiropractic care: results of a population- based case- control and
case- crossover study. Eur Spine J 2008:176–83.
16 Whedon JM, Song Y, Mackenzie TA, etal. Risk of stroke after
chiropractic spinal manipulation in Medicare B beneciaries
aged 66 to 99 years with neck pain. J Manipulative Physiol Ther
2015;38:93–101.
17 Cassidy JD, Boyle E, Côté P, etal. Risk of carotid stroke after
chiropractic care: a population- based case- crossover study. J Stroke
Cerebrovasc Dis 2017;26:842–50.
18 Whedon JM, Petersen CL, Li Z, etal. Association between cervical
artery dissection and spinal manipulative therapy -A Medicare claims
analysis. BMC Geriatr 2022;22:917.
19 Gorrell LM, Kuntze G, Ronsky JL, etal. Kinematics of the head and
associated vertebral artery length changes during high- velocity, low-
amplitude cervical spine manipulation. Chiropr Man Therap 2022;30.
20 Gorrell LM, Sawatsky A, Edwards WB, etal. Vertebral arteries do not
experience tensile force during manual cervical spine manipulation
applied to human cadavers. J Man Manip Ther 2022:1–9.
21 Begg C, Cho M, Eastwood S, etal. Improving the quality of reporting
of randomized controlled trials. The CONSORT statement. JAMA
1996;276:637–9.
22 Moher D, Schulz KF, Altman DG. The CONSORT statement: revised
recommendations for improving the quality of reports of parallel-
group randomised trials. Lancet 2001;357:1191–4.
23 Schulz KF, Altman DG, Moher D, etal. Consort 2010 statement:
updated guidelines for reporting parallel group randomised trials.
BMC Med 2010;8:18.
24 Ioannidis JPA, Evans SJW, Gøtzsche PC, etal. Better reporting
of harms in randomized trials: an extension of the CONSORT
statement. Ann Intern Med 2004;141:781–8.
25 Junqueira DR, Phillips R, Zorzela L, etal. Time to improve the
reporting of harms in randomized controlled trials. J Clin Epidemiol
2021;136:216–20.
26 Gorrell LM, Engel RM, Brown B, etal. The reporting of adverse
events following spinal manipulation in randomized clinical trials- a
systematic review. Spine J 2016;16:1143–51.
27 Page MJ, McKenzie JE, Bossuyt PM, etal. The PRISMA 2020
statement: an updated guideline for reporting systematic reviews.
BMJ 2021;372:71.
28 Herzog W. The biomechanics of spinal manipulation. J Bodyw Mov
Ther 2010;14:280–6.
29 Bergmann T. Chiropractic Technique Principles and Procedures, 3rd
ed. Missouri: USA, 2011.
30 Pohlman KA, O’Beirne M, Thiel H, etal. Development and validation
of providers’ and patients’ measurement instruments to evaluate
adverse events after spinal manipulation therapy. Eur J Int Med
2014;6:451–66.
31 Walker BF, Hebert JJ, Stomski NJ, etal. Outcomes of usual
chiropractic. The OUCH randomized controlled trial of adverse
events. Spine 2013;38:1723–9.
32 Carnes D, Mullinger B, Underwood M. Dening adverse events in
manual therapies: a modied Delphi consensus study. Man Ther
2010;15:2–6.
33 Dunning J, Butts R, Zacharko N, etal. Corrigendum to "spinal
manipulation and perineural electrical dry needling in patients with
cervicogenic headache: a multi- center randomized clinical trial" [the
spine Journal 21/2 (2021) p284- 295]. Spine J 2021;21:1238.
34 Maiers M, Hartvigsen J, Evans R, etal. Short- or long- term treatment
of spinal disability in older adults with manipulation and exercise.
Arthritis Care Res (Hoboken) 2019;71:1516–24.
35 Vining R, Long CR, Minkalis A, etal. Effects of chiropractic care
on strength, balance, and endurance in active- duty U.S. military
personnel with low back pain: a randomized controlled trial. J Altern
Complement Med 2020:592–601.
36 Schulz C, Evans R, Maiers M, etal. Spinal manipulative therapy and
exercise for older adults with chronic low back pain: a randomized
clinical trial. Chiropr Man Therap 2019;27:21.
37 Ouzzani M, Hammady H, Fedorowicz Z, etal. Rayyan- a web and
mobile APP for systematic reviews. Syst Rev 2016;5:210.
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
10 GorrellLM, etal. BMJ Open 2023;13:e067526. doi:10.1136/bmjopen-2022-067526
Open access
38 International Committeee of Medical Journal Editors (ICMJE).
Journals following the ICMJE recommendations; 2016.
39 Clarivate Journal citation reports. Available: https://clarivate.com/
webofsciencegroup/solutions/journal-citation-reports/ [Accessed 29
May 2022].
40 Higgins JP, Savović J, Page MJ, etal. Assessing risk of bias in a
randomized trial. In: Cochrane Handbook for Systematic Reviews of
Interventions. John Wiley & Sons, Ltd, 2019: 205–28.
41 Funabashi M, Pohlman KA, Gorrell LM, etal. Expert consensus
on a standardised denition and severity classication for adverse
events associated with spinal and peripheral joint manipulation and
mobilisation: protocol for an international e- Delphi study. BMJ Open
2021;11:e050219.
42 Carnes D, Mars TS, Mullinger B, etal. Adverse events and manual
therapy: a systematic review. Man Ther 2010;15:355–63.
43 Carlesso LC, Macdermid JC, Santaguida LP. Standardization of
adverse event terminology and reporting in orthopaedic physical
therapy: application to the cervical spine. J Orthop Sports Phys Ther
2010;40:455–63.
44 Carlesso LC, Cairney J, Dolovich L, etal. Dening adverse events
in manual therapy: an exploratory qualitative analysis of the patient
perspective. Man Ther 2011;16:440–6.
45 Carlesso LC, Gross AR, Santaguida PL, etal. Adverse events
associated with the use of cervical manipulation and mobilization for
the treatment of neck pain in adults: a systematic review. Man Ther
2010;15:434–44.
46 Zorzela L, Loke YK, Ioannidis JP, etal. PRISMA harms checklist:
improving harms reporting in systematic reviews. BMJ
2016;352:i157.
47 Funabashi M, Gorrell LM, Pohlman KA, etal. Denition and
classication for adverse events following spinal and peripheral
joint manipulation and mobilization: a scoping review. PLOS ONE
2022;17:e0270671.
48 Pitkin RM. The importance of the Abstract. Obstet Gynecol
1987;70:267.
49 Zorzela L, Golder S, Liu Y, etal. Quality of reporting in systematic
reviews of adverse events: systematic review. BMJ 2014;348:f7668.
50 Komorowski AS, MacKay HJ, Pezo RC. Quality of adverse
event reporting in phase III randomized controlled trials of
breast and colorectal cancer: a systematic review. Cancer Med
2020;9:5035–50.
51 Berwanger O, Ribeiro RA, Finkelsztejn A, etal. The quality of
reporting of trial Abstracts is suboptimal: survey of major general
medical journals. J Clin Epidemiol 2009;62:387–92.
52 Pitrou I, Boutron I, Ahmad N, etal. Reporting of safety results in
published reports of randomized controlled trials. Arch Intern Med
2009;169:1756–61.
53 Nuovo J, Sather C. Reporting adverse events in randomized
controlled trials. Pharmacoepidemiol Drug Saf 2007;16:349–51.
54 Phillips R, Hazell L, Sauzet O, etal. Analysis and reporting of
adverse events in randomised controlled trials: a review. BMJ Open
2019;9:e024537.
55 Phillips R, Sauzet O, Cornelius V. Statistical methods for the analysis
of adverse event data in randomised controlled trials: a scoping
review and taxonomy. BMC Med Res Methodol 2020;20:288.
56 Gorrell LM, Brown B, Lystad RP, etal. Predictive factors for reporting
adverse events following spinal manipulation in randomized clinical
trials- secondary analysis of a systematic review. Musculoskelet Sci
Pract 2017;30:34–41.
57 Winterbottom M, Boon H, Mior S, etal. Informed consent for
chiropractic care: comparing patients’ perceptions to the legal
perspective. Man Ther 2015;20:463–8.
on May 4, 2023 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2022-067526 on 4 May 2023. Downloaded from
