[ RESEARCH REPORT ]

EMILIO J. PUENTEDURA, PT, DPT, PhD1 • JOSHUA A. CLELAND, PT, DPT, PhD2 • MERRILL R. LANDERS, PT, DPT, PhD3
PAUL MINTKEN, PT, DPT4 • ADRIAAN LOUW, PT, MSc5 • CÉSAR FERNÁNDEZ-DE-LAS-PEÑAS, PT, MSc, PhD6

Development of a Clinical Prediction

Rule to Identify Patients With Neck
Pain Likely to Benefit From Thrust Joint
Manipulation to the Cervical Spine

N
eck pain is
! STUDY DESIGN: Prospective cohort/predictive used to determine the most parsimonious set of
a common
validity study. variables for prediction of treatment success. Vari-
complaint
! OBJECTIVE: To determine the predictive valid-
ables retained in the regression model were used to
SUPPLEMENTAL
ity of selected clinical examination items and to
develop a multivariate clinical prediction rule. in the adult VIDEO ONLINE

develop a clinical prediction rule to determine which
patients with neck pain may benefit from cervical
thrust joint manipulation (TJM) and exercise.
! BACKGROUND: TJM to the cervical spine has
been shown to be effective in patients presenting
with a primary report of neck pain. It would be
useful for clinicians to have a decision-making tool,
such as a clinical prediction rule, that could ac-
curately identify which subgroup of patients would
respond positively to cervical TJM.
! METHODS: Consecutive patients who presented
to physical therapy with a primary complaint of neck
pain completed a series of self-report measures,
then received a detailed standardized history and
physical examination. After the clinical examination,
all patients received a standardized treatment regi-
men consisting of cervical TJM and range-of-motion
exercise. Depending on response to treatment,
patients were treated for 1 or 2 sessions over ap-
proximately 1 week. At the end of their participation
in the study, patients were classified as having
experienced a successful outcome based on a score
of +5 (“quite a bit better”) or higher on the global
rating of change scale. Sensitivity, specificity, and
positive and negative likelihood ratios were calcu-
lated for all potential predictor variables. Univariate
techniques and stepwise logistic regression were
! RESULTS: Eighty-two patients were included
in data analysis, of whom 32 (39%) achieved a
successful outcome. A clinical prediction rule with
4 attributes (symptom duration less than 38 days,
positive expectation that manipulation will help,
side-to-side difference in cervical rotation range of
motion of 10° or greater, and pain with posteroante-
rior spring testing of the middle cervical spine) was
identified. If 3 or more of the 4 attributes (positive
likelihood ratio of 13.5) were present, the probability
of experiencing a successful outcome improved
from 39% to 90%.
! CONCLUSION: The clinical prediction rule may
improve decision making by providing the ability
to a priori identify patients with neck pain who are
likely to benefit from cervical TJM and range-of-
motion exercise. However, this is only the first step
in the process of developing and testing a clinical
prediction rule, as future studies are necessary to
validate the results and should include long-term
follow-up and a comparison group.
! LEVEL OF EVIDENCE: Prognosis, level 2b.
J Orthop Sports Phys Ther 2012;42(7):577-592,
Epub 14 May 2012. doi:10.2519/jospt.2012.4243
! KEY WORDS: clinical decision rule, clinical
prediction guide, manual therapy, mobilization,
prognosis
general population,
with typical 12-month prevalence
estimates of between 30% and
50%.27 Though chronic neck pain is less
common, with 12-month prevalence es-
timates ranging from 2% to 11%, it often
results in prolonged disability and sub-
stantial negative economic impact.15,27
Studies evaluating the cost-of-illness
and cost-effectiveness of therapeutic
interventions for neck pain have con-
cluded that clinical research should fo-
cus on developing and investigating the
most effective treatments for acute neck
pain, to prevent patients from developing
chronic pain and disability.41 To this end,
clinical practice guidelines have recom-
mended the use of thrust joint manipula-
tion (TJM) and nonthrust mobilization in
this patient population.7 Although there
is strong evidence for such an approach,31
physical therapists may be reluctant to
utilize TJM in the cervical spine because
of real or perceived risks associated with

1
Assistant Professor, Department of Physical Therapy, School of Allied Health Sciences, University of Nevada, Las Vegas, Las Vegas, NV. 2Professor, Department of Physical
Therapy, Franklin Pierce University, Concord, NH. 3Associate Professor, Department of Physical Therapy, School of Allied Health Sciences, University of Nevada, Las Vegas, Las
Vegas, NV. 4Associate Professor, Physical Medicine and Rehabilitation, Physical Therapy Program, School of Medicine, University of Colorado, Aurora, CO. 5Senior Instructor,
International Spine and Pain Institute, Story City, IA. 6Professor, Department of Physical Therapy, Occupational Therapy, Rehabilitation and Physical Medicine, Universidad Rey
Juan Carlos, Alcorcón, Madrid, Spain. The study was approved by the University of Nevada, Las Vegas Office for Research Integrity - Human Subjects. This clinical study received
funding from the American Academy of Orthopaedic Manual Physical Therapists, 2008 Cardon Research Grant. Address correspondence to Dr Emilio J. Puentedura, Assistant
Professor, University of Nevada, Las Vegas, School of Allied Health Sciences, Department of Physical Therapy, 4505 Maryland Parkway, Box 453029, Las Vegas, NV 89154-3029.
E-mail: louie.puentedura@unlv.edu ! Copyright ©2012 Journal of Orthopaedic & Sports Physical Therapy

journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 | 577

!"#$%&'()*+),(-./0*,,&&&1%% 23"$3"$4"&&&56176$"&'8
 [ RESEARCH REPORT
these techniques.16
Clinicians should certainly consider
the potential benefits of TJM in the cer-
vical spine relative to any potential risks,
especially cervical artery dissection.
These risks are well documented, and,
as a result, premanipulative screening
tools have primarily focused on identi-
fying patients who may experience ver-
tebrobasilar insufficiency or may have
cervical arterial dysfunction.6,40 The use
of these screening tools to rule out risk
of arterial dissection associated with TJM
is controversial.57,64 The lack of evidence
for premanipulative screening has caused
some authors to suggest that identifying
patients for whom there may be risks as-
sociated with TJM in the cervical spine
is virtually impossible and that perhaps
the potential benefits may not outweigh
the inherent risks.16,23,24 This may explain
why physical therapists will report uti-
lizing TJM in the thoracic spine more
frequently than in the cervical spine in
patients with neck pain,1 despite evidence
that many of these potential negative out-
comes may be prevented through careful
examination.56
Recent evidence suggests that TJM to
the thoracic spine may improve pain and
disability in patients with neck pain.13
Treating the thoracic spine in these pa-
tients has demonstrated benefits and
involves arguably less risk but may not
completely address the patient’s present-
ing symptoms and mobility impairments.
Our experience and recent evidence54
suggest that the addition of TJM to the
cervical spine to a comprehensive man-
agement package improves clinical out-
comes. Puentedura et al54 completed a
randomized controlled trial comparing
TJM to the cervical spine versus the tho-
racic spine in patients with neck pain
who met the criteria for a proposed clini-
cal prediction rule intended to identify
individuals with neck pain who would
benefit from TJM to the thoracic spine.12
Both groups received the same exercise
program, and the results indicated that
patients with acute neck pain (less than
30 days in duration) who received TJM
578 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy
!"#$%&'()*+),(-./0*,,&&&1%7
to their cervical spine had greater im-
provements in neck disability (P!.001)
and pain (P!.003) at all follow-up times
than those who received TJM to the tho-
racic spine.54
A clinical prediction rule for patients
with neck pain who responded favorably
to cervical TJM was developed in 2006
by Tseng et al67; however, the study only
addressed immediate (within-session)
results, and the threshold for “success”
might have been too low, as the pretest
probability of success was 60%. With a
pretest probability of success of 60%,
and a positive likelihood ratio (+LR) of
5.33 (95% confidence interval [CI]: 1.72,
16.54) for at least 4 of 6 predictors pres-
ent, the authors arrived at an 89% post-
test probability of success. A +LR of 5.33
represents only a small to moderate shift
in probability,36 and the wide CI also
detracts from the clinical utility of the
rule. In addition, by only assessing for an
immediate response, the study failed to
provide information on longer-term out-
comes that would have been more indica-
tive of a true change in condition.
The purpose of this study was to de-
velop a clinical prediction rule using a
more rigorous design (ie, higher thresh-
old for success, longer-term outcome) to
identify patients with neck pain likely to
benefit from TJM to their cervical spine,
based on a reference standard of patient-
reported improvement. Developing this
clinical prediction rule should be seen
as only the first step in creating a tool to
guide the use of TJM to the cervical spine
in patients with mechanical neck pain.
METHODS
T
here are many rules of thumb in
the statistical literature regarding
adequate sample sizes for logis-
tic regression.17,51,52 Most suggest that a
minimum of 10 cases of data should be
collected for each predictor in the model;
others suggest 15 cases of data per pre-
dictor.17,51,52 These suggested sample sizes
are thought to oversimplify the issue
considerably,17 as the required sample
]
size depends on the size of the effects be-
ing measured and the power required to
detect those effects. Furthermore, these
rules of thumb suggest the number of
variables expected to remain in the model
and not the number that can be entered
into the logistic regression model. For
these reasons, it was determined that a
sufficient number of patients had to be
recruited to enable sufficient statistical
assessment of a broad-enough number
of predictors.
The recruitment strategy was a
sample of convenience using a prospec-
tive cohort of patients with mechanical
neck pain. To get an appropriate num-
ber of patients, consecutive patients
with a primary complaint of neck pain
were screened for eligibility criteria at
outpatient physical therapy clinics in 4
different locations (Las Vegas, NV; Au-
rora, CO; Overland Park, KS; and Ma-
drid, Spain). For patients to be eligible
to participate, they had to be between
18 and 60 years of age, have a primary
report of neck pain with or without uni-
lateral upper extremity symptoms, and
have a baseline Neck Disability Index
(NDI) score of 10 points (out of 50) or
greater. Exclusion criteria were as fol-
lows: any medical red flags suggesting
that the etiology of symptoms might be
nonmusculoskeletal; diagnosis of cervi-
cal spinal stenosis (as identified in the
patient's medical intake form); bilateral
upper extremity symptoms; evidence of
central nervous system involvement; his-
tory of whiplash injury within 6 weeks of
the examination; pending legal action re-
garding the neck pain; 2 or more positive
neurologic signs consistent with nerve
root compression (changes in sensation,
myotomal weakness, or decreased deep
tendon reflexes); or any history of cervi-
cal spine surgery, rheumatoid arthritis,
osteoporosis, osteopenia, or ankylosing
spondylitis. The study was approved
by the University of Nevada, Las Vegas
Office for Research Integrity - Human
Subjects, and all study participants pro-
vided informed consent prior to their
participation.
23"$3"$4"&&&56176$5&'8

FIGURE 1. Cervical spine thrust manipulation used in
this study. The therapist used his manipulative hand
to localize the motion segment targeted and used
both hands to perform a high-velocity, low-amplitude
thrust into rotation, which was directed up toward the
patient’s contralateral eye (ONLINE VIDEO).
Therapists
Five physical therapists participated
in the examination and treatment of
patients in this study. All therapists
participated in a standardized train-
ing program, which included studying
a manual of standard procedures with
operational definitions and video clips
demonstrating each examination and
treatment procedure. All therapists at-
tended a 1-hour training session with the
principal investigator and were required
to successfully demonstrate the examina-
tion and treatment techniques to ensure
that all study procedures were performed
in a standardized fashion. All 5 partici-
pating therapists were male, with a mean
(range) of 43.6 (36-54) years of age and
17.9 (9-31) years of experience.
Examination Procedures
All patients underwent a standardized
history and physical examination. De-
mographic information collected in-
cluded age, sex, mechanism of injury (if
any), location and nature of symptoms,
and number of days since onset of symp-
toms. Self-report measures included the
NDI,13 the numeric pain rating scale
(NPRS),26,27,34 the Fear-Avoidance Be-
liefs Questionnaire,73,77 and the Tampa
Scale of Kinesiophobia.47,63 The historical
examination included a body diagram to
assess the distribution of symptoms, fol-
low-up questions regarding aggravating
journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 |
!"#$%&'()*+),(-./0*,,&&&1%9
FIGURE 2. Following the thrust joint manipulation
to the cervical spine, each patient performed the
3-finger exercise for cervical rotation, as used in a
previous study. With this exercise, the cervical spine
is brought in flexion to the point where there is a
3-finger space between the manubrium and chin.
Rotation of the cervical spine is then performed in
both directions, while maintaining the same amount
of neck flexion.53
and relieving factors, 24-hour behavior of
presenting symptoms, any prior history
of neck pain, and patients’ expectations
about what would help their current epi-
sode of neck pain.
The physical examination followed
the same protocol as that described in
detail by Cleland et al12 and began with a
neurological screening examination fol-
lowed by postural assessment. Cervical
range of motion (ROM) was measured,
and symptom response assessed, fol-
lowed by assessments of muscle length
and strength. The amount of motion and
symptom response were recorded for seg-
mental passive intervertebral mobility
testing of the cervical spine and passive
accessory intervertebral mobility testing
of the cervical and thoracic spines (C2-
T9). Reliability analysis of the clinical
examination procedures was not under-
taken, as these values are well document-
ed in the literature.11,12
The physical examination was then
concluded with a number of special tests
typically performed in the examination
of patients with neck pain, including
the Spurling test,66 the neck distraction
test,74 and the upper-limb neurodynamic
test for the median nerve.55 Patients were
screened for any signs of vertebrobasilar
insufficiency, such as nystagmus, gait dis-
turbances, and Horner syndrome, as well
as for upper cervical spine ligamentous
laxity through the Sharp-Purser test, alar
ligament stress test, and transverse liga-
ment test.8 If any patients had a positive
finding on any of these final screening
tests, they were excluded from the study.
Treatment
All patients participating in the study
received a standardized treatment, re-
gardless of the results of the clinical ex-
amination. This was necessary because
treatment outcome was to serve as the ref-
erence criterion.35 Each patient received a
supine TJM to the cervical spine directed
to an appropriate level between C3 and
C7 (FIGURE 1, ONLINE VIDEO). Therapists par-
ticipating in the study were allowed to
use discretion in selecting the level they
felt was more restricted, then performed
an upslope (rotation) technique at that
level to each side of the cervical spine.
The TJM technique used in this study has
been described in detail elsewhere.54 Im-
mediately after performing the TJM, the
treating therapist recorded whether there
was an audible cavitation. If the therapist
and patient did not hear or feel a cavita-
tion, the therapist repeated the TJM to an
allowable maximum of 2 thrusts to each
side. Following the TJM techniques, all
579
23"$3"$4"&&&56176$!&'8
 [ RESEARCH REPORT ]
maximize the likelihood that the clinical
138 patients with Presented with: outcome could be attributable to mean-
mechanical neck š Contraindications, ingful improvements in symptoms due to
Not eligible, n = 42
pain screened for n = 20 the treatment as opposed to the passage
eligibility criteria š History of whiplash of time.
within 6 wk, Patients whose scores on the GROC
n = 10
did not exceed the cutoff of +5 or greater
Declined to š Signs of nerve root
Eligible, n = 96 compression, n = 8 at the second session again received bilat-
participate,
n = 14 š Prior surgery to eral cervical TJM as described for the first
cervical spine, n = 3 treatment, and were instructed to con-
š Inability to comply tinue their cervical ROM exercises and
Agreed to with treatment usual activity. At the start of the third ses-
participate, signed schedule, n = 1 sion, patients again completed the GROC
informed consent, and were categorized as having achieved
and received success, or not, based on the previously
treatment 1, described threshold for success. At this
n = 82
point, the patient’s participation in the
study was complete, and further treat-
ment was administered at the discretion
Follow-up 48 h
Dropouts, n = 0 of the patient’s treating therapist. A flow
later, n = 82
diagram showing subject recruitment,
including reasons for ineligibility and
dropout, is provided in FIGURE 3.
Successful outcome
Yes, n = 20
on visit 2
Adverse Events
No All patients were asked to report any ad-
Total in success group, verse events or treatment side effects at
Underwent second
n = 32 each follow-up. We defined an adverse
treatment, n = 62
event as a sequela with moderate to se-
vere symptoms that were serious, dis-
tressing, and unacceptable to the patient
Successful outcome
Yes, n = 12 and required the patient to withdraw
on visit 3
from further participation in the study.
No We defined treatment side effects as short
Total in nonsuccess term (resolved by the time they returned
group, n = 50 for follow-up), mild in nature, nonseri-
ous, and reversible consequences of the
treatment, such as increased neck pain,
FIGURE 3. Flow diagram of patient recruitment and retention.
headache, and fatigue.

patients were instructed in gentle active
ROM exercise (10 repetitions performed
3-4 times daily) (FIGURE 2) and advised to
maintain usual activity within the limits
of pain. The patient was scheduled for a
follow-up visit 2 to 4 days after the base-
line exam.
The global rating of change (GROC)
was used as the reference criterion for
establishing a successful outcome follow-
ing treatment. The GROC is a 15-point
scale with the anchors at –7 (a very great
deal worse), 0 (about the same), and +7
(a very great deal better). It was deter-
mined a priori that patients who rated
their perceived recovery on the GROC as
+7 (a very great deal better), +6 (a great
deal better), or +5 or higher (quite a bit
better) at the second session would be
categorized as a success and their partici-
pation in the study would be complete.
This high threshold for determining a
successful outcome was based on previ-
ous similar studies,12,13 with the intent to
Data Analysis
Patients participating in the study were
dichotomized as having experienced or
not experienced a successful outcome
based on the previously described cut-
off for the GROC. Mean NDI and NPRS
change scores and 95% CIs were calculat-
ed for both groups and analyzed using an
independent t test to ascertain difference
between the groups. The success group
and the nonsuccess group were then
compared on each of the variables from

580 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy

!"#$%&'()*+),(-./0*,,&&&17$ 23"$3"$4"&&&56176$1&'8
 the history and physical examination us-
ing independent-samples t tests for con- Demographics, Baseline Self-Report
tinuous variables and chi-square tests for TABLE 1 Variables, and Baseline Characteristics
categorical variables. We decided to only of Patients in the Study*
retain variables with a significance level
of P<.05 to minimize the number of pre- Variable All Subjects (n = 82) Success (n = 32) Nonsuccess (n = 50) P Value
dictor variables entered into the regres- Age, y 38.3 ! 14.7 37.3 !13.9 38.9 ! 15.4 .627†
sion model. Female, n (%) 48 (59%) 21 (66%) 27 (54%) .297‡
All potential predictor variables found Body mass index, kg/m2 24.6 ! 5.3 23.7 ! 5.6 25.2 ! 5.1 .228†
to be statistically significant in the com- Duration of symptoms, d 293.2 ! 814.5 35.3 ! 26.7 458.3 ! 1012.3 .005†
parison between the success and nonsuc- NPRS (0-10) 4.1 ! 1.7 4.8 ! 1.6 3.7 ! 1.5 .003†
cess groups were further analyzed using NDI (0-50) 15.3 ! 4.9 14.6 ! 4.7 15.7 ! 5.0 .335†
receiver-operator-characteristic (ROC) FABQPA (0-24) 7.7 ! 5.1 6.6 ! 5.5 8.4 ! 4.8 .106†
curves, with the dichotomous outcome FABQW (0-42) 8.8 ! 8.4 6.5 ! 6.7 10.3 ! 9.1 .031†
of success. Sensitivity and specificity val- TSK-11 (11-44) 22.4 ! 6.1 20.1 ! 6.4 23.3 ! 5.8 .092†
ues were calculated for each level of the Unilateral neck symptoms, n (%) 31 (38%) 17 (53%) 14 (28%) .022‡
variable and plotted on a ROC curve, at Symptoms distal to the shoulder, n (%) 40 (49%) 9 (28%) 31 (62%) .003‡
which time a cutoff point was determined Mode of onset: gradual, n (%) 44 (54%) 12 (38%) 32 (64%) .102‡
for each variable. The point on the curve Prior history of neck pain, n (%) 41 (50%) 13 (41%) 28 (56%) .174‡

that was nearest the upper left corner Symptoms aggravated by:

represented the value with the best di- Turning right, n (%) 46 (56%) 20 (63%) 26 (52%) .186‡

agnostic accuracy, and this point was Turning left, n (%) 43 (52%) 18 (56%) 25 (50%) .010‡

selected as the cutoff defining a positive Looking up, n (%) 39 (48%) 20 (63%) 19 (38%) .021‡

test within this study. If 2 points were Looking down, n (%) 39 (48%) 16 (50%) 23 (46%) .049‡

equidistant, then the point at which the
curve made a marked turn after the steep
initial section was determined to be the
cut-point.
Using these ROC curve–generated
cutoff points, each score was dichoto-
mized as above or below the cut-point
and the following were calculated: sensi-
tivity, specificity, and positive and nega-
tive LRs. These dichotomized potential
predictor variables were then entered
into a backward (LR) stepwise logistic
regression model to determine the most
parsimonious set of variables for pre-
diction of treatment success. Backward
entry was used because it is generally
considered the preferred method for ex-
ploratory analyses.17 Variables retained
in the regression model were then used
to develop a multivariate clinical predic-
tion rule for classifying patients likely to
benefit from cervical TJM for this study
sample.
Variables that demonstrated a signifi-
cant (P<.05) univariate relationship with
the GROC reference criterion were then
assessed for multicollinearity. Variables
that were highly correlated contributed
Driving, n (%) 56 (68%)
Computer use, n (%) 63 (77%)
Abbreviations: FABQPA, Fear-Avoidance Beliefs Questionnaire physical activity subscale; FABQW,
Fear-Avoidance Beliefs Questionnaire work subscale; NDI, Neck Disability Index; NPRS, numeric
pain rating scale; TSK-11, Tampa Scale of Kinesiophobia.
*Data are mean ! SD unless otherwise indicated.
†
Independent-samples t tests.
‡
Chi-square tests.
essentially the same information and did
not have a statistically significant impact
on predicting the dependent variable.
Individual R2 value and variance infla-
tion factor were assessed for each of the
variables. Variables that exhibited high R2
and variance inflation factor values were
determined to be collinear and would,
therefore, affect model fit by multicol-
linearity. This issue was resolved by either
combining a pair of correlated variables
or removing 1 of the 2 correlated vari-
ables. As an example, “cervical rotation
range to the right” was highly correlated
with “cervical rotation range to the left.”
Therefore, a new variable named “side-
to-side difference in rotation ROM” was
created. Because “side-to-side difference
in rotation ROM” was highly correlated
with “side-to-side difference in lateral
23 (72%) 33 (66%) .577‡
29 (91%) 34 (68%) .018‡
flexion ROM,” it made clinical sense to
remove “side-to-side difference in lateral
flexion ROM” and to keep “side-to-side
difference in rotation ROM,” cervical
rotation differences being easier for cli-
nicians to assess using the clinical predic-
tion rule.
After combining or eliminating cor-
related variables, 9 variables were left to
consider for entry into logistic regression.
Although logistic regression is capable of
handling both continuous and categori-
cal data, the remaining 9 variables were
converted to dichotomous (yes/no and
present/not present) attributes. This was
done on the basis that the clinical predic-
tion rule would be easier to administer
clinically if therapists simply had to de-
termine whether the predictor attribute
was present (positive) in a particular

journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 | 581

!"#$%&'()*+),(-./0*,,&&&174 23"$3"$4"&&&56176$2&'8
 [ RESEARCH REPORT ]
patient rather than the extent to which
it was present. Cutoff values for continu- Continuous Variables From
TABLE 2
ous and ordinal data were determined by the Baseline Clinical Examination*
ROC curves, and resulted in symptom
duration of less than 38 days, current Variable All Subjects (n = 82) Success (n = 32) Nonsuccess (n = 50) P Value
pain rating of greater than 5, side-to-side Cervical flexion, deg† 54.6 ! 10.6 53.8 ! 12.0 55.2 ! 9.7 .549
difference in cervical rotation ROM of Cervical extension, deg† 50.9 ! 15.1 46.7 ! 16.5 53.6 ! 13.7 .044
10° or greater, and deep neck flexor en- Cervical lateral flexion right, deg† 38.4 ! 9.3 36.3 ! 9.6 39.8 ! 8.9 .101
durance of less than 18 seconds. These 4 Cervical lateral flexion left, deg† 37.8 ! 9.1 35.7 ! 10.7 39.2 ! 7.7 .087
dichotomized attributes were then com- Side-to-side difference in lateral 7.2 ! 8.0 13.3 ! 9.3 3.3 ! 3.5 <.001
bined with 5 additional attributes (con- flexion, deg†

taining nominal data) that exhibited a Cervical rotation right, deg‡ 69.0 ! 12.1 65.8 ! 13.1 71.0 ! 11.0 .053

significance level less than .05, resulting Cervical rotation left, deg‡ 68.0 ! 15.4 63.8 ! 16.9 70.8 ! 13.9 .042

in 9 potential predictor attributes (inde- Side-to-side difference in rotation, deg‡ 10.8 ! 11.7 19.2 ! 12.1 5.4 ! 7.6 <.001

pendent variables) to be entered into the Deep neck flexor endurance, s 20.0 ! 8.6 16.4 ! 6.6 22.3 ! 9.0 .002
*Data are mean ! SD.
logistic regression. †
Indicates measurement with gravity inclinometer.
‡
Indicates measurement with standard dual-arm goniometer.
RESULTS

B
etween October 2009 and May
Categorical Variables From the History
2011, 82 patients were recruited TABLE 3
Portion of the Baseline Clinical Examination
for the study. All patients enrolled
in the study completed all requirements
All Subjects Success Nonsuccess
(there were no dropouts). Subject demo- Variable (n = 82) (n = 32) (n = 50) P Value (!2)
graphics and initial baseline variables Neck pain is bothersome (yes), n 71 31 40 .029
from the patient history and self-report Headache is bothersome (yes), n 35 16 19 .284
measures for patients, as well as the suc- Arm and hand pain is bothersome (yes), n 17 3 14 .042
cess and nonsuccess groups, are in TABLE Numbness tingling is bothersome (yes), n 17 3 14 .042
1. From the baseline clinical examina- Weakness arm/hand is bothersome (yes), n 7 3 4 .828
tion variables for all patients and both Positive expectation that:
groups, continuous data are in TABLE 2 Medications will help (yes), n 47 16 31 .284
and categorical data in TABLES 3 through Rest will help (yes), n 54 21 33 .972
6. Thirty-two patients (39%) were cat- Surgery will help (yes), n 22 5 17 .067
egorized as having achieved a successful Modalities will help (yes), n 69 25 44 .232
outcome and 50 (61%) as a nonsuccessful Massage will help (yes), n 73 27 46 .281
one. Twenty of the 32 patients (62.5%) Manipulation will help (yes), n 52 29 23 <.001
reported success after the initial treat- Traction will help (yes), n 68 27 41 .780
ment, and the remaining 12 after 2 ses- Aerobic exercise will help (yes), n 71 29 42 .391
sions. The mean ! SD number of days Range-of-motion exercise will help (yes), n 73 30 43 .273
between visits 1 and 2 was 2.8 ! 0.8 and Strengthening exercise will help (yes), n 69 28 41 .506
2.7 ! 0.9 (P = .27) for the success and
nonsuccess groups, respectively. The
mean ! SD number of days between vis- (FIGURE 4) and significantly greater im- such as increased neck pain, headache,
its 1 and 3 was 5.7 ! 0.9 and 5.3 ! 1.0 provements (P<.001) in disability (NDI and fatigue. The presence of these in-
(P = .60) for the success and nonsuccess change score, 9.6; 95% CI: 8.2, 11.0) over creased symptoms led them to report
groups, respectively. Analysis of NPRS the nonsuccess group (NDI change score, worsening (negative scores) on the GROC
and NDI change scores revealed that the 5.4; 95% CI: 4.2, 6.7) (FIGURE 5). (FIGURE 6). However, by the final session,
success group experienced significantly No patients experienced an adverse only 5 patients reported continued wors-
greater improvements (P<.001) in pain event that required them to withdraw ening on the GROC (FIGURE 7).
(NPRS change score, 3.5; 95% CI: 3.0, from further participation in the study. The accuracy statistics and 95% CIs
4.1) compared to the nonsuccess group After the first treatment session, 10 pa- for all 9 potential predictor attributes
(NPRS change score, 1.1; 95% CI: 0.7, 1.5) tients reported treatment side effects entered into the regression are in TABLE

582 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy

!"#$%&'()*+),(-./0*,,&&&17" 23"$3"$4"&&&56176$%&'8
 (95% CI: 4.5, 41.1) and the posttest prob-
Categorical Variables From the Postural ability of success increased to 90% (95%
TABLE 4
Portion of the Baseline Clinical Examination CI: 74%, 96%). The posttest probability
of achieving a successful outcome if 2 or
All Subjects Success Nonsuccess more of the 4 attributes were present was
Variable (n = 82) (n = 32) (n = 50) P Value (!2) 68% (95% CI: 58%, 76%). If the criterion
Presence of: for being positive on the rule was reduced
Forward head posture (yes), n 69 27 42 .964 to only 1 or more of the 4 attributes being
Lateral shift to 1 side (yes), n 33 20 13 .001 present, the +LR decreased to 1.2 (95%
Excessive cervicothoracic junction kyphosis (yes), n 47 17 30 .539 CI: 1.0, 1.4) and the posttest probability
Excessive straightening T3-T5 kyphosis (yes), n 18 12 6 .007 of success was only 43% (95% CI: 40%,
Excessive upper cervical extension (yes), n 10 0 10 .019* 46%).
Rotated to 1 side (yes), n 28 21 7 <.001
Excessive thoracic kyphosis (yes), n 17 1 16 .004* DISCUSSION
Shoulders not level (yes), n 19 5 14 .195

*Continuity correction secondary to cells having expected count less than 5.
Categorical Variables From the
TABLE 5 Active Range-of-Motion Portion of
the Baseline Clinical Examination
All Subjects Success Nonsuccess
Variable (n = 82) (n = 32) (n = 50) P Value (!2)
Increased symptoms with:
Cervical flexion, n 28 11 17 .232
Cervical extension, n 34 19 15 .017
Cervical rotation to the right, n 33 19 14 .001
Cervical rotation to the left, n 31 15 16 .007
Cervical lateral flexion to the right, n 30 16 14 .022
Cervical lateral flexion to the left, n 30 14 16 .004
Cervical quadrant* to the right, n 25 14 11 .098
Cervical quadrant* to the left, n 24 12 12 .184
*Quadrant is the combined extension, sidebending, and rotation to 1 side.
7. The +LR ranged from 1.75 to 6.77, with efit from cervical TJM. All 16 patients
the strongest predictor being “side-to- who met all 4 criteria were in the success
side difference in cervical rotation ROM group. Twenty-six of the 29 patients who
of 10° or greater.” Of the 9 predictor at- were positive on at least 3 of the criteria
tributes entered into the regression, the were in the success group. Of the 53 pa-
following 4 attributes were retained in tients who exhibited 2 or fewer of the 4
the final model: symptom duration less criteria, 47 were in the nonsuccess group
than 38 days, a positive expectation that (TABLE 8). Accuracy statistics were calcu-
manipulation will help, side-to-side dif- lated for each level of the clinical predic-
ference in cervical rotation ROM of 10° tion rule (TABLE 9). The pretest probability
or greater, and pain with posteroanterior for the likelihood of success with TJM for
spring testing of the middle cervical spine this study was 39%. That is, 39% of the
(!2 = 71.6, df = 4, P<.001, Nagelkerke R2 patients exhibited treatment success re-
= 0.790). A combination of these 4 at- gardless of their clinical profile. However,
tributes was considered to be the most taking into account the model of their
accurate predictor to identify those pa- clinical profile, if the patient exhibited at
tients with neck pain who will likely ben- least 3 of 4 attributes, the +LR was 13.5
A
s previous clinical prediction
rule derivation studies12,18,26,48 have
done, the present study focused on
the +LR to identify patients with neck
pain who would experience a successful
outcome with cervical TJM. The +LR is
considered the most useful statistic to
determine a change in probability that a
patient will likely have a successful out-
come based on positive test findings.35
We have developed a clinical prediction
rule that maximizes the +LR (13.5) for
identifying patients with neck pain who
might respond rapidly and positively to a
TJM technique applied to the midcervi-
cal region. Based on our results, a patient
with neck pain who exhibited at least 3
of 4 criteria and was treated with TJM
to the cervical spine showed an increase
from a pretest probability of success of
39% to a posttest probability of success
of 90%. This represents a significant shift
in probability and a potentially powerful
tool to guide clinical decision making in
the use of cervical TJM for treating pa-
tients with nonradiating neck pain.
Due to concerns regarding the poten-
tial risks associated with cervical TJM,
adverse events were monitored during
the course of the study. There were no se-
rious adverse events encountered during
the study that required the withdrawal of
patients from the study. A small percent-
age of patients (12%) reported worsening
of their symptoms after the first treat-
ment session (FIGURE 6), and 1 patient re-
ported feeling “a very great deal worse”

journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 | 583

!"#$%&'()*+),(-./0*,,&&&175 23"$3"$4"&&&56176$7&'8
 [ RESEARCH REPORT ]
6
Categorical Variables From the
4.78
TABLE 6 Passive Range-of-Motion Portion 5

NPRS Score (0-10)

3.69
of the Baseline Clinical Examination 4

3 2.57

All Subjects Success Nonsuccess 2 1.24

Variable (n = 82) (n = 32) (n = 50) P Value (!2) 1
Presence of muscle tightness in: 0
Latissimus dorsi, n 39 12 27 .144 Pre Post

Pectoralis minor, n 51 20 31 .964

Success Nonsuccess
Pectoralis major, n 36 10 26 .065
Upper trapezii, n 55 21 34 .823
FIGURE 4. Baseline and final scores on the NPRS
Anterior/middle scalene, n 54 22 32 .658 for the success and nonsuccess groups. The
Suboccipitals, n 48 26 22 .001 change score (3.5 ! 1.5) for the success group was
Levator scapulae, n 43 18 25 .580 significantly greater (P<.001) than the change score
Symptoms elicited with: (1.1 ! 1.4) for the nonsuccess group. Abbreviation:
NPRS, numeric pain rating scale.
Spring testing (PA) on C2, n 32 19 13 .003
Spring testing (PA) on C3, n 50 28 22 <.001
Spring testing (PA) on C4, n 48 25 23 .004 18
15.70
Spring testing (PA) on C5, n 32 14 18 .124 16
14
Spring testing (PA) on C6, n 22 6 16 .195 14.63
12 10.30

NDI Score (0-50)

Spring testing (PA) on C7, n 23 5 18 .108
10
Spring testing (PA) on T1, n 19 4 15 .067 8
Spring testing (PA) on T2, n 17 6 11 .723 6 5.03

Spring testing (PA) on T3, n 14 5 9 .780 4

2
Spring testing (PA) on T4, n 16 4 12 .200
0
Abbreviation: PA, posteroanterior. Pre Post

(GROC, –7). Despite this, all 10 patients
underwent a repeat cervical TJM dur-
ing their second session, of whom only 5
(6%) continued to report worsening on
the GROC at the third visit (FIGURE 7). It is
important to remember that all patients
participating in this study were screened
for contraindications and precautions
for TJM to the cervical spine and were
excluded from participation if they ex-
hibited any such contraindications. This
underscores the importance and value of
the thorough history and physical exami-
nation prior to intervention.
The absence of the combination of
predictor attributes derived from this
study may also be helpful in identifying
patients with neck pain who should re-
ceive forms of treatment other than TJM.
All 27 patients in the study who exhibited
only 1 of the predictor attributes and all
8 patients who had none of the attributes
failed to achieve success. Given the lower
likelihood of success associated with the
absence of a combination of these attri-
butes, therapists should carefully weigh
the benefit-risk ratio in deciding whether
to include cervical TJM as a part of the
treatment for such patients and perhaps
consider more appropriate treatment
options.
The success rate of 39% for the en-
tire heterogeneous group of patients in
this study was lower than the previously
reported rate of successful outcomes
for manipulating the thoracic spine in
patients with neck pain (54%).12 It was
also lower than the values obtained for
the percentage of successful outcomes
of manipulating the lumbopelvic region
in patients with mechanical low back
pain (45%).18 This may be explained by
the possibility that patients with neck
pain who are more likely to benefit from
cervical TJM represent a small subset of
the heterogeneous group of patients with
neck pain. Fritz and Brennan20 conducted
a preliminary examination of a classifica-
Success Nonsuccess
FIGURE 5. Baseline and final scores for the NDI for
the success and nonsuccess groups. The change
score (9.6 ! 4.0) for the success group was
significantly greater (P<.001) than the change score
(5.4 ! 4.4) for the nonsuccess group. Abbreviation:
NDI, Neck Disability Index.
tion system for patients with neck pain
proposed by Childs et al,10 examining case
data for 274 consecutive patients receiv-
ing physical therapy. They found that the
most common classification subgroup
was centralization (34.7%), followed
closely by the exercise-and-conditioning
subgroup (32.8%). The mobility sub-
group (those presumed to respond more
favorably to cervical TJM and exercise)
represented only 17.5% of the patients.
The remaining subgroups of headache
and pain control accounted for 15% of the
274 patients. The 39% who were found to
have successful outcomes from the TJM
in our study, which is considerably larger

584 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy

!"#$%&'()*+),(-./0*,,&&&17! 23"$3"$4"&&&56176$9&'8
 than the subgroup reported by Fritz and
18
Brennan,20 may be explained in 1 of 2 17
16
ways: the subgroup for mobility may 16
actually be larger than that reported by Success, 20
14
Fritz and Brennan20 or subgroups other
than just the mobility subgroup may ben- 12
efit from cervical TJM.

Number of Patients
10
10
9
Attributes Identified in the Clinical
8
Prediction Rule 7
Worse, 10 6 6
From the broad list of potential vari- 6
ables for inclusion in our clinical predic-
4
tion rule for patients with nonradicular 3
neck pain, only 4 predictor attributes 2 2 2
2
1 1
were selected for the final model: symp- 0 0
tom duration less than 38 days, a posi- 0

tive expectation that manipulation will −7 −6 −5 −4 −3 −2 −1 0 +1 +2 +3 +4 +5 +6 +7

help, side-to-side difference in cervical

rotation ROM of 10° or greater, and pain
with posteroanterior spring testing of the
middle cervical spine. Because these 4
attributes represent an important aspect
of clinical decision for success in cervical
TJM, we will discuss each in turn.
Rating on the GROC
FIGURE 6. Frequency of the GROC scores at the second session. Blue bars represent patients who were classified
as having experienced success with the cervical spine thrust joint manipulation. All other colors represent
nonsuccess, with green bars representing perceived worsening, the black bar representing no perceived change,
and orange bars representing some perceived improvement but not to the predetermined cutoff for success.
Abbreviation: GROC, global rating of change.

Duration of Symptoms
Success, 32
In previous studies, shorter duration of 22
symptoms was the predictor with the
20 19
highest +LR (6.4) in identifying patients
with neck pain who were likely to have 18

a successful outcome with thoracic ma-

16
nipulation,12 and it was also the predictor
14
with the highest +LR in identifying pa- 14
13
tients with low back pain who were likely 12
Number of Patients

12
11
to benefit from lumbar spine manipula-
tion (4.39).18 Furthermore, several stud- 10

ies have shown that patients who receive 8

physical therapy care soon after the onset 6
6
of their musculoskeletal condition have a Worse, 5
better prognosis compared to those who 4
receive care later.54,75,79 But, although we 2
2 1
also found symptom duration to be a 1 1 1 1
0 0 0
predictor with a high +LR (6.0), unlike 0
previous studies, it was not the variable −7 −6 −5 −4 −3 −2 −1 0 +1 +2 +3 +4 +5 +6 +7

with the highest +LR.12,18

Rating on the GROC

Difference in Cervical Rotation ROM

The proposed mechanisms of spinal TJM
are thought to involve biomechanical,
neurophysiological, and placebo effects.
Thus, the remaining 3 predictor attri-
butes appear to have some face validity.
FIGURE 7. Frequency of the GROC scores at the final session. Blue bars represent patients who were classified
as having experienced success with the cervical spine thrust joint manipulation. All other colors represent
nonsuccess, with green bars representing perceived worsening, the black bar representing no perceived change,
and orange bars representing some perceived improvement but not to the predetermined cutoff for success.
Abbreviation: GROC, global rating of change.

journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 | 585

!"#$%&'()*+),(-./0*,,&&&171 23"$3"$4"&&&561764$&'8
 [ RESEARCH REPORT ]
Accuracy Statistics (95% Confidence Intervals)
TABLE 7
for Individual Predictor Attributes in the Study*

Variable Sensitivity Specificity Positive Likelihood Ratio Posttest Probability

Symptom duration <38 d 0.72 (0.53, 0.86) 0.88 (0.75, 0.95) 6.0 (2.7, 13.1) 79% (64%, 89%)
Positive expectation that manipulation will help 0.91 (0.74, 0.98) 0.54 (0.39, 0.68) 2.0 (1.4, 2.7) 56% (48%, 63%)
Current pain rating (NPRS, 0-10) >5 0.50 (0.32, 0.68) 0.82 (0.68, 0.91) 2.8 (1.4, 5.5) 64% (47%, 78%)
No symptoms beyond point of shoulder 0.71 (0.53, 0.86) 0.62 (0.47, 0.75) 1.9 (1.3, 2.9) 54% (44%, 64%)
Postural rotation to 1 side 0.66 (0.47, 0.81) 0.86 (0.73, 0.94) 4.7 (2.3, 9.7) 75% (59%, 86%)
Worse with cervical extension 0.59 (0.41, 0.78) 0.70 (0.55, 0.82) 2.0 (1.2, 3.3) 56% (43%, 68%)
Side-to-side difference in cervical rotation ROM at least 10° 0.81 (0.63, 0.92) 0.88 (0.75, 0.95) 6.8 (3.1, 14.6) 81% (67%, 91%)
Deep neck flexor endurance <18 s 0.69 (0.50, 0.83) 0.72 (0.57, 0.83) 2.5 (1.5, 4.1) 61% (49%, 72%)
Pain with spring (PA) testing of the middle cervical spine 0.88 (0.70, 0.96) 0.50 (0.36, 0.64) 1.8 (1.3, 2.4) 53% (45%, 60%)
Abbreviations: NPRS, numeric pain rating scale; PA, posteroanterior; ROM, range of motion.
*Pretest probability of success, 39%.

receiving the manipulative thrust.22 The

The 4 Attributes Forming the Clinical arthrokinematics of the facet joints in
TABLE 8 Prediction Rule and the Number of the cervical spine have been thoroughly
Patients in Each Group at Each Level* researched over many years and appear
to be well accepted.14,21,37 Lateral flexion
Number of Predictor Attributes Present Success Group Nonsuccess Group Percent Success and rotation are coupled motions in the
4 16 0 16/16 = 100 cervical spine. In the upper cervical seg-
3 10 3 10/13 = 76.9 ments (0-C1 and C1-C2), lateral flexion
2 6 12 6/18 = 33.3 is coupled with contralateral rotation
1 0 27 0/27 = 0 and rotation is coupled with contralat-
0 0 8 0/8 = 0 eral lateral flexion.33,34,44 The opposite
*Attributes in the CPR: symptom duration of less than 38 days, positive expectation that manipula- has been observed in the lower cervical
tion will help, side-to-side difference in cervical rotation range of motion of 10° or greater, pain with
spring (posteroanterior) testing of the middle cervical spine.
spine (C3-C7), where coupling motions
are ipsilateral.3,53 Furthermore, Bogduk
and Mercer3 and Mercer and Bogduk46
have suggested that the notion of lateral
Combination of Predictor
flexion and rotation in the cervical spine
TABLE 9 Attributes and Associated Accuracy
might be an artificial construct, arguing
Statistics (95% Confidence Intervals)
that this movement should be viewed as
a gliding that occurs in the plane of the
Number of Predictor Positive Probability
Attributes Present Sensitivity Specificity Likelihood Ratio of Success, %*
facet joints.
4 0.50 (0.31, 0.68) 1.00 (0.91, 1.00) Infinite (3.2, infinite) 100% (67%, 100%)
In our study, patients who exhibited
3+ 0.81 (0.63, 0.92) 0.94 (0.82, 0.98) 13.5 (4.5, 41.1) 90% (74%, 96%)
decreased ROM in rotation to 1 side also
2+ 1.00 (0.87, 1.00) 0.70 (0.55, 0.82) 3.3 (2.1, 4.9) 68% (58%, 76%)
demonstrated decreased ROM in lat-
1+ 1.00 (0.87, 1.00) 0.16 (0.08, 0.30) 1.2 (1.0, 1.4) 43% (40%, 46%)
eral flexion to the same side. We found
a side-to-side difference in cervical rota-
*The probability of success is calculated using the positive likelihood ratios and assumes a pretest
probability of 39%. Predictor attributes included: symptom duration of less than 38 days, a positive tion ROM to be highly correlated with a
expectation that manipulation will help, side-to-side difference in cervical rotation range of motion of side-to-side difference in cervical lateral
10° or greater, pain with spring (posteroanterior) testing of the middle cervical spine.
flexion ROM (r = 0.80, P<.001), and,
therefore, eliminated 1 or the other when
We found that a side-to-side difference in used a rotation (upslope) cervical TJM finalizing the test variables to enter into
cervical rotation ROM of 10° or greater technique, which is purported to have our prediction model. We chose cervi-
(FIGURE 8, ONLINE VIDEO) was the predic- a biomechanical effect of “opening” or cal rotation ROM difference over lateral
tor that demonstrated the highest +LR distracting the articular surfaces of the flexion ROM difference on the clinical
(6.8). Of interest is that in this study we particular zygapophyseal (facet) joint grounds that it is arguably easier for

586 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy

!"#$%&'()*+),(-./0*,,&&&172 23"$3"$4"&&&5617644&'8
 FIGURE 9. Pain with posteroanterior spring testing
of the middle cervical spine. This was performed
with the patient in comfortable prone lying, and the
therapist used the pads of both thumbs to apply a
FIGURE 8. Side-to-side difference in cervical rotation ROM was assessed with the patient seated in a chair with posterior-to-anterior force on each of the spinous
back support to stabilize the thoracic spine. The therapist used a universal dual-arm goniometer with the axis processes from C3 to C7. Patients were asked to
over the vertex, the stationary arm perpendicular to the shoulders and the movable arm in line with the patient’s report any aggravation or reproduction of their
nose. If the difference in angular ROM between sides was 10° or greater, it was considered positive for the predictor familiar neck pain (ONLINE VIDEO).
attribute (ONLINE VIDEO). Abbreviation: ROM, range of motion.
subgroup of patients presenting with
clinicians to measure. Furthermore, the neck pain may suffer from facet-related chronic muscular or centrally sensitized
cutoff points for the ROC curve favored pain.69 Several studies report the pres- pain. Patients with chronic pain or cen-
rotation (10° difference or greater), as it ence of synovial folds,32,68,76 meniscoids,19 tral sensitization have been shown to
exceeded the standard error of measure- and fat pads65 within facet joints that may report widespread pain and difficulty
ment of 5.5°, whereas lateral flexion (4.5° become entrapped during aberrant mo- accurately locating their pain.59,61,78 In
difference or greater) did not exceed the tions of the cervical spine and lead to the cases in which a patient’s experience of
standard error of measurement of 7°.11 onset of neck pain. It seems plausible that neck pain has extended beyond a local
We found that patients with neck pain if such entrapment did occur, motions mechanism (eg, facet joint arthropathy),
who exhibited a difference in cervical ro- that would require the affected joints to it would be reasonable to conclude that a
tation ROM to either side were limited close down against the entrapped tissues local treatment technique, such as cervi-
toward the painful side or, in the case would provoke symptoms. The possibil- cal TJM, would not be as likely to provide
of patients with bilateral neck pain, to- ity of nociceptive tissues being entrapped immediate benefit.
ward the more painful side. It could be within facet joints in patients with acute
argued that rotation toward the painful neck pain would lend support to the idea Positive Expectation That Manipulation
side may cause the facet joints on that that a cervical TJM technique may allow Will Help
side to close down (articular surfaces for release of the impinged nociceptive Unlike the other attributes discussed so
would approximate) due to the coupled tissues with a subsequent amelioration far, which are theoretically and logically
same-side lateral flexion. Rotation away of pain. classified as treatment-effect mediators
from the painful side may cause the facet (factors that change as a consequence
joints on the painful side to open (articu- Pain With Posteroanterior Spring Testing of the treatment and thereby influence
lar surfaces would be distracted) due to of the Middle Cervical Spine outcome), a positive expectation of ma-
the coupled same-side lateral flexion. The Although the reliability of manual palpa- nipulation is classified as a treatment
therapists who participated in this study tion of the cervical spine has been called moderator because it is a baseline char-
were instructed to apply the rotation into question,60,70 palpation for spinal acteristic that may influence the outcome
(upslope) cervical TJM technique in the tenderness has been shown to be highly of treatment.50 The finding that a positive
nonpainful (or least painful) direction reliable.29 It seems reasonable that a test expectation for manipulation was predic-
first, then to apply the same technique to assessing the presence of local symptoms tive of success is consistent with the fact
the opposite side. It is biomechanically (pain) in the cervical spinal segments to that expectation of benefit (placebo) has
plausible that the cervical TJM technique receive a cervical TJM technique would been shown to have a robust effect on
might have provided an opening (distrac- be identified as a predictor of success pain.71,72 These nonspecific effects need
tion) force to facet joints that were pain- (FIGURE 9, ONLINE VIDEO). In contrast, the to be considered more carefully in all fu-
ful with closing down. A recent review population of patients with neck pain ture research of treatment interventions.
article concluded that more than 50% of who do not report pain with palpation We sought to ensure that the interven-
patients presenting to a pain clinic with in spinal segments may be a different tion (TJM) was delivered in a standard-

journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 | 587

!"#$%&'()*+),(-./0*,,&&&17% 23"$3"$4"&&&561764"&'8
 [ RESEARCH REPORT
ized manner, so as not to confound our
results by affecting patient expectations
about their clinical outcome. All 5 ther-
apists who collected data for this study
gave participants a standardized narra-
tive about the cervical TJM intervention
to be used and limited their discussion
to the information provided in the in-
formed-consent form. The questionnaire
used in this study to assess for treatment
expectations is provided in the APPENDIX.
Perceived Recovery Versus Perceived
Disability
As in a previous clinical prediction rule
derivation study involving patients with
neck pain,12 we chose to use perceived
recovery using the GROC to determine
success rather than a perceived level of
disability using the NDI. This was based
on the fact that the GROC is considered
to be a valid reference standard for iden-
tifying clinically important change.5,30 We
also chose to use perceived recovery as
the reference criterion, because the NDI
has been criticized for not adequately
capturing low levels of disability and for
not being responsive to small but clini-
cally important changes in patients with
low levels of initial disability.28 The mean
change score for the NDI in patients who
reported success was 8.7 (lower-bound
estimate of 95% CI, 4.0) and in patients
who reported nonsuccess was 6.2 (lower-
bound estimate of 95% CI, 1.7), meaning
that both change scores exceeded the rec-
ommended minimal detectable change of
5 points, but only the success group ex-
ceeded the minimal clinically important
difference of 7 points.43 However, with
standard deviations of 4.71 and 4.46, re-
spectively, there were some patients in
the success group who failed to exceed
the minimal detectable change and mini-
mal clinically important difference, and,
equally, some patients in the nonsuccess
group who did exceed the minimal de-
tectable change and minimal clinically
important difference for the NDI. Ad-
ditionally, measures of success ratings
based on the patient’s perceived recovery
have previously been used in clinical trials
588 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy
!"#$%&'()*+),(-./0*,,&&&177
with physical therapy interventions.28 Fi-
nally, this study only examined the short-
term effects of cervical TJM in patients
with neck pain, and within that short
time frame the perception of recovery
may be more appropriate to use than the
perceived level of disability remaining. In
studies that follow patients over a longer
period (eg, 4 weeks, 6 months), perceived
disability may be more appropriate.54
Jaeschke et al36 have proposed that
a score of +4 on the GROC provides in-
creased confidence that improvement or
recovery has occurred. Similar to a pre-
vious clinical prediction rule derivation
study,12 we chose to use a higher thresh-
old for defining success (+5 through +7)
to provide a greater degree of distinction
between patients who improved dramati-
cally and those who might have improved
over time due to the natural history of
their disorder. In this study, we identi-
fied potential predictor attributes of pa-
tients with neck pain who might respond
rapidly to cervical TJM; however, this is
only the first in a 3-step process for de-
veloping and testing a clinical prediction
rule,45 and future studies are necessary to
examine the validity of our results.
Limitations
Cointerventions such as medication and
self-directed exercise are potentially con-
founding variables. We sought to control
for these confounding variables by direct-
ing patients in the study to continue with
their present medication and activities,
while not initiating the use of any new
medication or activity while they were
participating in the study. We collected
data on medication usage and found no
statistical difference in medication usage
(over the counter or prescription) be-
tween the 2 groups.
Although data for this study were col-
lected from 4 geographic locations, there
was an unequal distribution of patients
among the 4 sites. Data were collected
from 42 patients (51%) in Madrid, Spain,
26 patients (32%) in Las Vegas, NV, 9 pa-
tients (11%) in Aurora, CO, and, finally,
5 patients (6%) in Overland Park, KS.
]
Rates of success reported by patients at
each location were as follows: Madrid,
38.1%; Las Vegas, 30.8%; Aurora, 66.7%;
and Overland Park, 40.0%.
A further limitation applicable to this
study was that the cervical TJM tech-
nique used might not have been spe-
cific to the targeted vertebral segments.
Evidence suggests that both thrust and
nonthrust manipulation techniques may
not be joint specific4,58; however, this was
adequately addressed by not claiming
that the cervical TJM technique used in
this study was specific to any segment. It
could also be argued that the results of
this study could not be specifically cor-
related with the cervical TJM technique,
as gentle cervical ROM exercises were
also incorporated in the intervention.
The cervical ROM exercises were includ-
ed to more closely approximate clinical
practice, where therapists are likely to
use more than a single intervention. Ad-
ditionally, the incorporation of a basic
ROM exercise was consistent with other
studies that developed18 and validated9 a
clinical prediction rule for patients with
low back pain, and developed12 but failed
to validate13 a clinical prediction rule for
patients with neck pain who would im-
prove with thoracic manipulation.
Finally, as with other clinical predic-
tion rule derivation studies, this was a
single-arm design, and without a con-
trol group it is not possible to determine
whether the clinical prediction rule iden-
tifies prognosis (regardless of treatment)
or response to the specific treatment.2,25,49
Without a control group, we cannot know
whether improvement after the clinical
prediction rule treatment was actually
due to the treatment. This is considered a
significant limitation of the single-group
design, in that an initially hypothesized
clinical prediction rule does not have pre-
scriptive validity.49 Patients in our single-
group study who improved might have
had a more favorable natural history, and,
consequently, patients who were positive
on our initially hypothesized clinical pre-
diction rule might have done just as well
with no cervical TJM intervention.25 This
23"$3"$4"&&&5617645&'8
 has led to calls for dual-arm randomized
controlled trial designs in the develop-
ment of clinical prediction rules25,38,39;
however, treatment-based clinical predic-
tion rules have often been developed us-
ing a single-arm design, as it allows for a
large number of variables to be studied.42
It is worth noting that all patients in
this study had the benefit of a detailed
history and physical examination prior
to receiving TJM to the cervical spine
and ROM exercises. This study, there-
fore, developed a clinical prediction rule
that identified individuals with neck
pain who would respond favorably to the
combination of a thorough examination
(history and physical), cervical TJM, and
ROM exercises. We cannot determine
the contribution of the examination to
the outcome, and it is possible that the
examination might have provided some
therapeutic benefit to the patient.
Although an initially hypothesized
clinical prediction rule may not have
prescriptive validity, it is seen as help-
ful in some situations.49 The patients in
our single-group study who were posi-
tive on the clinical prediction rule com-
prised a prospective case series in which
improvement after the clinical predic-
tion rule treatment (TJM) was highly
likely.2 According to Straus et al,62 this
provides low-level evidence that pa-
tients in the clinic who are positive on
the clinical prediction rule are also likely
to improve after cervical TJM. Finally,
this single-group-design study to derive
a clinical prediction rule should be seen
as the first in a 3-step process to develop
a treatment-related clinical prediction
rule for patients with neck pain who will
respond rapidly and favorably to cervical
TJM.
CONCLUSION
W
e have developed a clinical
prediction rule that identifies
patients with neck pain who are
likely to rapidly and dramatically respond
to cervical spine TJM and ROM exercises.
Such a clinical prediction rule will be ad-
journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 |
!"#$%&'()*+),(-./0*,,&&&179
vantageous in determining which individ-
ual patients are likely to respond (and not
respond) to this treatment strategy. Future
studies are necessary to validate the clini-
cal prediction rule and determine if the use
of the clinical prediction rule results in im-
proved outcome and decreased costs in the
treatment of individuals with neck pain. !
KEY POINTS
FINDINGS: A preliminary clinical predic-
tion rule with 4 attributes to identify
individuals with mechanical neck pain
who respond to cervical TJM was iden-
tified: symptom duration of less than
38 days, positive expectation that ma-
nipulation will help, side-to-side differ-
ence in cervical rotation ROM of 10° or
greater, and pain with posteroanterior
spring testing of the middle cervical
spine. If at least 3 attributes were pres-
ent (+LR, 13.5), the probability of expe-
riencing a successful outcome increased
from 39% to 90%.
IMPLICATIONS: The findings provide pre-
liminary evidence to support the ability
to identify patients with neck pain who
will be more likely to benefit from cervi-
cal TJM by using select attributes from
the clinical examination.
CAUTION: Future studies are necessary to
validate these results and should also
include long-term follow-up and a com-
parison group to further examine the
predictive value of the attributes identi-
fied in the clinical prediction rule.
REFERENCES
1. Adams G, Sim J. A survey of UK manual thera-
pists’ practice of and attitudes towards manipu-
lation and its complications. Physiother Res Int.
1998;3:206-227.
2. Beattie P, Nelson R. Clinical prediction rules:
what are they and what do they tell us? Aust J
Physiother. 2006;52:157-163.
3. Bogduk N, Mercer S. Biomechanics of the cervi-
cal spine. I: normal kinematics. Clin Biomech
(Bristol, Avon). 2000;15:633-648.
4. Bolton A, Moran RW, Standen C. An investiga-
tion into the side of joint cavitation associ-
ated with cervical spine manipulation. Int J
Osteopath Med. 2007;10:88-96. http://dx.doi.
org/10.1016/j.ijosm.2007.10.004
5. Bolton JE. Sensitivity and specificity of outcome
measures in patients with neck pain: detecting
clinically significant improvement. Spine (Phila
Pa 1976). 2004;29:2410-2417; discussion 2418.
6. 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;40:455-463. http://
dx.doi.org/10.2519/jospt.2010.3229
7. Childs JD, Cleland JA, Elliott JM, et al. Neck
pain: clinical practice guidelines linked to the
International Classification of Functioning, Dis-
ability, and Health from the Orthopaedic Section
of the American Physical Therapy Association.
J Orthop Sports Phys Ther. 2008;38:A1-A34.
http://dx.doi.org/10.2519/jospt.2008.0303
8. Childs JD, Flynn TW, Fritz JM, et al. Screening
for vertebrobasilar insufficiency in patients with
neck pain: manual therapy decision-making in
the presence of uncertainty. J Orthop Sports
Phys Ther. 2005;35:300-306. http://dx.doi.
org/10.2519/jospt.2005.1312
9. 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 ma-
nipulation: a validation study. Ann Intern Med.
2004;141:920-928.
10. Childs JD, Fritz JM, Piva SR, Whitman JM.
Proposal of a classification system for patients
with neck pain. J Orthop Sports Phys Ther.
2004;34:686-696; discussion 697-700. http://
dx.doi.org/10.2519/jospt.2004.1451
11. Cleland JA, Childs JD, Fritz JM, Whitman
JM. Interrater reliability of the history and
physical examination in patients with me-
chanical neck pain. Arch Phys Med Rehabil.
2006;87:1388-1395. http://dx.doi.org/10.1016/j.
apmr.2006.06.011
12. Cleland JA, Childs JD, Fritz JM, Whitman JM,
Eberhart SL. Development of a clinical predic-
tion rule for guiding treatment of a subgroup of
patients with neck pain: use of thoracic spine
manipulation, exercise, and patient educa-
tion. Phys Ther. 2007;87:9-23. http://dx.doi.
org/10.2522/ptj.20060155
13. Cleland JA, Mintken PE, Carpenter K, et al.
Examination of a clinical prediction rule to
identify patients with neck pain likely to benefit
from thoracic spine thrust manipulation and a
general cervical range of motion exercise: multi-
center randomized clinical trial. Phys Ther.
2010;90:1239-1250. http://dx.doi.org/10.2522/
ptj.20100123
14. Cook C, Hegedus E, Showalter C, Sizer PS, Jr.
Coupling behavior of the cervical spine: a sys-
tematic review of the literature. J Manipulative
Physiol Ther. 2006;29:570-575. http://dx.doi.
org/10.1016/j.jmpt.2006.06.020
15. Cote P, van der Velde G, Cassidy JD, et al.
The burden and determinants of neck pain in
workers: results of the Bone and Joint Decade
2000-2010 Task Force on Neck Pain and Its
Associated Disorders. Spine (Phila Pa 1976).
589
23"$3"$4"&&&561764!&'8
 [ RESEARCH REPORT
2008;33:S60-74. http://dx.doi.org/10.1097/
BRS.0b013e3181643ee4
16. Di Fabio RP. Manipulation of the cervical spine:
risks and benefits. Phys Ther. 1999;79:50-65.
17. Field A. Discovering Statistics Using SPSS. 3rd
ed. London, UK: SAGE Publications; 2009.
18. Flynn T, Fritz J, Whitman J, et al. A clinical
prediction rule for classifying patients with low
back pain who demonstrate short-term im-
provement with spinal manipulation. Spine (Ph-
ila Pa 1976). 2002;27:2835-2843. http://dx.doi.
org/10.1097/01.BRS.0000035681.33747.8D
19. Friedrich KM, Trattnig S, Millington SA, Fried-
rich M, Groschmidt K, Pretterklieber ML.
High-field magnetic resonance imaging of
meniscoids in the zygapophyseal joints of the
human cervical spine. Spine (Phila Pa 1976).
2007;32:244-248. http://dx.doi.org/10.1097/01.
brs.0000251952.36735.fb
20. Fritz JM, Brennan GP. Preliminary examination
of a proposed treatment-based classifica-
tion system for patients receiving physical
therapy interventions for neck pain. Phys Ther.
2007;87:513-524. http://dx.doi.org/10.2522/
ptj.20060192
21. Fryer G, Adams JH. Magnetic resonance imag-
ing of subjects with acute unilateral neck pain
and restricted motion: a prospective case
series. Spine J. 2011;11:171-176. http://dx.doi.
org/10.1016/j.spinee.2010.12.002
22. Gibbons P, Tehan P. Manipulation of the Spine,
Thorax and Pelvis: An Osteopathic Perspective.
2nd ed. London, UK: Elsevier/Churchill Living-
stone; 2005.
23. Haldeman S, Kohlbeck FJ, McGregor M.
Stroke, cerebral artery dissection, and cervi-
cal spine manipulation therapy. J Neurol.
2002;249:1098-1104. http://dx.doi.org/10.1007/
s00415-002-0783-4
24. Haldeman S, Kohlbeck FJ, McGregor M. Un-
predictability of cerebrovascular ischemia
associated with cervical spine manipulation
therapy: a review of sixty-four cases after cervi-
cal spine manipulation. Spine (Phila Pa 1976).
2002;27:49-55.
25. Hancock M, Herbert RD, Maher CG. A guide
to interpretation of studies investigating sub-
groups of responders to physical therapy inter-
ventions. Phys Ther. 2009;89:698-704. http://
dx.doi.org/10.2522/ptj.20080351
26. Hicks GE, Fritz JM, Delitto A, McGill SM.
Preliminary development of a clinical predic-
tion rule for determining which patients with
low back pain will respond to a stabilization
exercise program. Arch Phys Med Rehabil.
2005;86:1753-1762. http://dx.doi.org/10.1016/j.
apmr.2005.03.033
27. Hogg-Johnson S, van der Velde G, Carroll LJ, et
al. The burden and determinants of neck pain
in the general population: results of the Bone
and Joint Decade 2000-2010 Task Force on
Neck Pain and Its Associated Disorders. Spine
(Phila Pa 1976). 2008;33:S39-51. http://dx.doi.
org/10.1097/BRS.0b013e31816454c8
590 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy
!"#$%&'()*+),(-./0*,,&&&19$
]
28. Hoving JL, Koes BW, de Vet HC, et al. Manual
therapy, physical therapy, or continued care by
a general practitioner for patients with neck
pain. A randomized, controlled trial. Ann Intern
Med. 2002;136:713-722.
29. Hubka MJ, Phelan SP. Interexaminer reliability
of palpation for cervical spine tenderness. J
Manipulative Physiol Ther. 1994;17:591-595.
30. Hurst H, Bolton J. Assessing the clinical
significance of change scores recorded on
subjective outcome measures. J Manipulative
Physiol Ther. 2004;27:26-35. http://dx.doi.
org/10.1016/j.jmpt.2003.11.003
31. Hurwitz EL, Carragee EJ, van der Velde G, et al.
Treatment of neck pain: noninvasive interven-
tions: results of the Bone and Joint Decade
2000-2010 Task Force on Neck Pain and Its
Associated Disorders. Spine (Phila Pa 1976).
2008;33:S123-152. http://dx.doi.org/10.1097/
BRS.0b013e3181644b1d
32. Inami S, Kaneoka K, Hayashi K, Ochiai N. Types
of synovial fold in the cervical facet joint. J
Orthop Sci. 2000;5:475-480. http://dx.doi.
org/10.1007/s007760070026
33. Ishii T, Mukai Y, Hosono N, et al. Kinematics
of the cervical spine in lateral bending: in vivo
three-dimensional analysis. Spine (Phila Pa
1976). 2006;31:155-160.
34. Ishii T, Mukai Y, Hosono N, et al. Kinematics of
the subaxial cervical spine in rotation in vivo
three-dimensional analysis. Spine (Phila Pa
1976). 2004;29:2826-2831.
35. Jaeschke R, Guyatt GH, Sackett DL. Users’
guides to the medical literature. III. How to use
an article about a diagnostic test. B. What are
the results and will they help me in caring for
my patients? The Evidence-Based Medicine
Working Group. JAMA. 1994;271:703-707.
36. Jaeschke R, Singer J, Guyatt GH. Measurement
of health status. Ascertaining the minimal clini-
cally important difference. Control Clin Trials.
1989;10:407-415.
37. Jaumard NV, Welch WC, Winkelstein BA. Spinal
facet joint biomechanics and mechanotrans-
duction in normal, injury and degenerative con-
ditions. J Biomech Eng. 2011;133:071010. http://
dx.doi.org/10.1115/1.4004493
38. Kamper SJ, Hancock MJ, Maher CG. Optimal
designs for prediction studies of whiplash.
Spine (Phila Pa 1976). 2011;36:S268-274. http://
dx.doi.org/10.1097/BRS.0b013e3182388202
39. Kent P, Hancock M, Petersen DH, Mjosund
HL. Clinimetrics corner: choosing appropriate
study designs for particular questions about
treatment subgroups. J Man Manip Ther.
2010;18:147-152. http://dx.doi.org/10.1179/1066
98110X12640740712419
40. Kerry R, Taylor AJ. Cervical arterial dysfunc-
tion: knowledge and reasoning for manual
physical therapists. J Orthop Sports Phys Ther.
2009;39:378-387. http://dx.doi.org/10.2519/
jospt.2009.2926
41. Korthals-de Bos IB, Hoving JL, van Tulder MW,
et al. Cost effectiveness of physiotherapy,
manual therapy, and general practitioner care
for neck pain: economic evaluation alongside a
randomised controlled trial. BMJ. 2003;326:911.
http://dx.doi.org/10.1136/bmj.326.7395.911
42. Laupacis A, Sekar N, Stiell IG. Clinical predic-
tion rules. A review and suggested modifica-
tions of methodological standards. JAMA.
1997;277:488-494.
43. MacDermid JC, Walton DM, Avery S, et al. Mea-
surement properties of the neck disability index:
a systematic review. J Orthop Sports Phys Ther.
2009;39:400-417. http://dx.doi.org/10.2519/
jospt.2009.2930
44. Malmstrom EM, Karlberg M, Fransson PA, Me-
lander A, Magnusson M. Primary and coupled
cervical movements: the effect of age, gender,
and body mass index. A 3-dimensional move-
ment analysis of a population without symp-
toms of neck disorders. Spine (Phila Pa 1976).
2006;31:E44-50.
45. McGinn TG, Guyatt GH, Wyer PC, Naylor CD,
Stiell IG, Richardson WS. Users’ guides to the
medical literature: XXII: how to use articles
about clinical decision rules. Evidence-
Based Medicine Working Group. JAMA.
2000;284:79-84.
46. Mercer SR, Bogduk N. Joints of the cervical
vertebral column. J Orthop Sports Phys Ther.
2001;31:174-182; discussion 183.
47. Miller RP, Kori SH, Todd DD. The Tampa Scale:
a measure of kinesiophobia. Clin J Pain.
1991;7:51-52.
48. Mintken PE, Cleland JA, Carpenter KJ, Bieniek
ML, Keirns M, Whitman JM. Some factors
predict successful short-term outcomes in
individuals with shoulder pain receiving cervico-
thoracic manipulation: a single-arm trial. Phys
Ther. 2010;90:26-42. http://dx.doi.org/10.2522/
ptj.20090095
49. Nee RJ, Coppieters MW. Interpreting research
on clinical prediction rules for physiotherapy
treatments. Man Ther. 2011;16:105-108. http://
dx.doi.org/10.1016/j.math.2010.10.005
50. Nicholas MK, Linton SJ, Watson PJ, Main CJ.
Early identification and management of psycho-
logical risk factors (“yellow flags”) in patients
with low back pain: a reappraisal. Phys Ther.
2011;91:737-753. http://dx.doi.org/10.2522/
ptj.20100224
51. Norušis MJ. PASW Statistics 18 Statistical Pro-
cedures Companion. Upper Saddle River, NJ:
Prentice Hall; 2010.
52. Pallant J. SPSS Survival Manual: A Step by Step
Guide to Data Analysis Using SPSS for Win-
dows. 3rd ed. New York, NY: McGraw-Hill/Open
University Press; 2007.
53. Panjabi MM, Crisco JJ, Vasavada A, et al.
Mechanical properties of the human cervical
spine as shown by three-dimensional load-
displacement curves. Spine (Phila Pa 1976).
2001;26:2692-2700.
54. Puentedura EJ, Landers MR, Cleland JA, Mint-
ken PE, Huijbregts P, Fernández-de-las-Peñas C.
Thoracic spine thrust manipulation versus cer-
23"$3"$4"&&&5617641&'8
 vical spine thrust manipulation in patients with
acute neck pain: a randomized clinical trial.
J Orthop Sports Phys Ther. 2011;41:208-220.
http://dx.doi.org/10.2519/jospt.2011.3640
55. Puentedura EJ, Louw A, Mintken PE. Neuro-
dynamic interventions for the management
of patients with headache. In: Fernández-de-
las-Peñas C, Arendt-Nielsen L, Gerwin RD,
eds. Tension-Type and Cervicogenic Headache:
Pathophysiology, Diagnosis, and Management.
Sudbury, MA: Jones and Bartlett Publishers;
2010:393-419.
56. Puentedura EJ, March J, Anders J, et al. Safety
of cervical spine manipulation: are adverse
events preventable and are manipulations be-
ing performed appropriately? A review of 134
case reports. J Man Manip Ther. 2012;20:66-
74. http://dx.doi.org/10.1179/204261861
1Y.0000000022
57. Rivett DA. Adverse events and the vertebral
artery: can they be averted? Man Ther.
2006;11:241-242. http://dx.doi.org/10.1016/j.
math.2006.08.001
58. Ross JK, Bereznick DE, McGill SM. Determining
cavitation location during lumbar and thoracic
spinal manipulation: is spinal manipulation
accurate and specific? Spine (Phila Pa 1976).
2004;29:1452-1457.
59. Scott D, Jull G, Sterling M. Widespread sensory
hypersensitivity is a feature of chronic whiplash-
associated disorder but not chronic idiopathic
neck pain. Clin J Pain. 2005;21:175-181.
60. Seffinger MA, Najm WI, Mishra SI, et al. Reliabil-
ity of spinal palpation for diagnosis of back and
neck pain: a systematic review of the literature.
Spine (Phila Pa 1976). 2004;29:E413-425.
61. Sterling M, Treleaven J, Edwards S, Jull G.
Pressure pain thresholds in chronic whiplash
associated disorder: further evidence of altered
central pain processing. J Musculoskelet Pain.
2002;10:69-81.
62. Straus SE, Richardson WS, Glasziou P, Haynes
RB. Evidence-Based Medicine: How to Practice
journal of orthopaedic & sports physical therapy | volume 42 | number 7 | july 2012 |
!"#$%&'()*+),(-./0*,,&&&194
and Teach EBM. 3rd ed. Edinburgh, UK: Elsevier/
Churchill Livingstone; 2005.
63. Swinkels-Meewisse EJ, Swinkels RA, Verbeek
AL, Vlaeyen JW, Oostendorp RA. Psychometric
properties of the Tampa Scale for Kinesiophobia
and the Fear-Avoidance Beliefs Questionnaire in
acute low back pain. Man Ther. 2003;8:29-36.
http://dx.doi.org/10.1054/math.2002.0484
64. Thiel H, Rix G. Is it time to stop functional
pre-manipulation testing of the cervical spine?
Man Ther. 2005;10:154-158. http://dx.doi.
org/10.1016/j.math.2004.06.004
65. Tonetti J, Peoc’h M, Merloz P, Pasquier B, Chiro-
ssel JP. Elastic reinforcement and thickness of
the joint capsules of the lower cervical spine.
Surg Radiol Anat. 1999;21:35-39.
66. Tong HC, Haig AJ, Yamakawa K. The Spurling
test and cervical radiculopathy. Spine (Phila Pa
1976). 2002;27:156-159.
67. Tseng YL, Wang WT, Chen WY, Hou TJ, Chen TC,
Lieu FK. Predictors for the immediate respond-
ers to cervical manipulation in patients with
neck pain. Man Ther. 2006;11:306-315. http://
dx.doi.org/10.1016/j.math.2005.08.009
68. Uhrenholt L, Hauge E, Charles AV, Gregersen
M. Degenerative and traumatic changes in
the lower cervical spine facet joints. Scand J
Rheumatol. 2008;37:375-384. http://dx.doi.
org/10.1080/03009740801998770
69. van Eerd M, Patijn J, Lataster A, et al. 5. Cervical
facet pain. Pain Pract. 2010;10:113-123. http://
dx.doi.org/10.1111/j.1533-2500.2009.00346.x
70. van Trijffel E, Anderegg Q, Bossuyt PM, Lucas C.
Inter-examiner reliability of passive assessment
of intervertebral motion in the cervical and
lumbar spine: a systematic review. Man Ther.
2005;10:256-269. http://dx.doi.org/10.1016/j.
math.2005.04.008
71. Vase L, Petersen GL, Riley JL, 3rd, Price DD.
Factors contributing to large analgesic effects in
placebo mechanism studies conducted between
2002 and 2007. Pain. 2009;145:36-44. http://
dx.doi.org/10.1016/j.pain.2009.04.008
72. Vase L, Riley JL, 3rd, Price DD. A comparison
of placebo effects in clinical analgesic trials
versus studies of placebo analgesia. Pain.
2002;99:443-452.
73. Waddell G, Newton M, Henderson I, Somerville
D, Main CJ. A Fear-Avoidance Beliefs Question-
naire (FABQ) and the role of fear-avoidance
beliefs in chronic low back pain and disability.
Pain. 1993;52:157-168.
74. Wainner RS, Fritz JM, Irrgang JJ, Boninger
ML, Delitto A, Allison S. Reliability and
diagnostic accuracy of the clinical examina-
tion and patient self-report measures for
cervical radiculopathy. Spine (Phila Pa 1976).
2003;28:52-62. http://dx.doi.org/10.1097/01.
BRS.0000038873.01855.50
75. Wand BM, Bird C, McAuley JH, Dore CJ,
MacDowell M, De Souza LH. Early interven-
tion for the management of acute low back
pain: a single-blind randomized controlled
trial of biopsychosocial education, manual
therapy, and exercise. Spine (Phila Pa 1976).
2004;29:2350-2356.
76. Webb AL, Collins P, Rassoulian H, Mitchell BS.
Synovial folds – a pain in the neck? Man Ther.
2011;16:118-124. http://dx.doi.org/10.1016/j.
math.2010.11.004
77. Williamson E. Fear Avoidance Beliefs Question-
naire (FABQ). Aust J Physiother. 2006;52:149.
78. Woolf CJ. Central sensitization: uncovering
the relation between pain and plasticity. Anes-
thesiology. 2007;106:864-867. http://dx.doi.
org/10.1097/01.anes.0000264769.87038.55
79. Zigenfus GC, Yin J, Giang GM, Fogarty WT.
Effectiveness of early physical therapy in
the treatment of acute low back musculo-
skeletal disorders. J Occup Environ Med.
2000;42:35-39.
@ MORE INFORMATION
WWW.JOSPT.ORG
591
23"$3"$4"&&&5617642&'8
 [ RESEARCH REPORT ]
APPENDIX

TREATMENT EXPECTATIONS
Indicate by circling the comment next to the treatment that corresponds to your amount of agreement with the following statement. Substitute each
treatment into the blank as you consider your response.

I believe ________________ will significantly help to improve this episode of my neck pain.

Note: If you have never received a particular treatment, base your answer on how much you think it would help if you were to receive this treatment.
Ask your physical therapist about any treatment that is not familiar to you.

Medication Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
Rest Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
Surgery Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
Modalities (ie, heat packs, ultrasound, TENS, etc) Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
Massage Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
Manipulation (ie, having your neck or back Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
“cracked” or “popped”)
Traction (lying on your back or stomach with Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
straps with a harness strapped on that
stretches out your neck or back)
Aerobic exercise (ie, walking, stationary cycling, Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
StairMaster, etc)
Range-of-motion exercises (ie, stretching) Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree
Strengthening exercises Completely disagree Somewhat disagree Neutral Somewhat agree Completely agree

FOR THE CLINICIAN

The above form was used to collect the data on expectation for this study. A similar form should be used to assess patient expectation in the clinic and
determine the presence of this attribute.

You can basically disregard all answers except the manipulation question.

If the patient circles “completely disagree,” “somewhat disagree,” or “neutral,” then there is NOT a positive expectation that manipulation will help their
neck pain, that is, negative for that predictor in the newly developed clinical prediction rule.

If the patient circles “somewhat agree” or “completely agree,” then there IS a positive expectation that manipulation will help their neck pain, that is,
positive for that predictor in the newly developed clinical prediction rule.

VIEW Videos on JOSPT’s Website

Videos posted with select articles on the Journal’s website (www.jospt.org)
show how conditions are diagnosed and interventions performed. For a
list of available videos, click on “COLLECTIONS” in the navigation bar in the
left-hand column of the home page, select “Media”, check “Video”, and
click “Browse”. A list of articles with videos will be displayed.

592 | july 2012 | volume 42 | number 7 | journal of orthopaedic & sports physical therapy

!"#$%&'()*+),(-./0*,,&&&19" 23"$3"$4"&&&5617642&'8