The relationship of Lumbar Flexion ...

[ Pobierz całość w formacie PDF ]
v
The Relationship of Lumbar Flexion
to Disability in Patients With Low
Back Pain
Background and Purpose.
Physical therapists routinely assess spinal
active range of motion (AROM) in patients with low back pain (LBP).
The purpose of this study was to use 2 approaches to examine the
relationship between impairment of lumbar spine flexion AROM and
disability. One approach relied on the use of normative data to
determine when an impairment in flexion AROM was present. The
other approach required therapists to make judgments of whether the
flexion AROM impairment was relevant to the patient’s disability.
Subjects.
Fifteen physical therapists and 81 patients with LBP com-
pleted in the study.
Methods.
Patients completed the Roland-Morris
Back Pain Questionnaire (RMQ), and the therapists assessed lumbar
spine flexion AROM using a dual-inclinometer technique at the initial
visit and again at discharge.
Results.
Correlations between the lumbar
flexion AROM measure and disability were low and did not vary
appreciably for the 2 approaches tested.
Conclusion and Discussion.
Measures of lumbar flexion AROM should not be used as surrogate
measures of disability. Lumbar spine flexion AROM and disability are
weakly correlated, suggesting that flexion AROM measures should not
be used as treatment goals. [Sullivan MS, Shoaf LD, Riddle DL. The
relationship of lumbar flexion to disability in patients with low back
pain.
Phys Ther
. 2000;80:240 –250.]
Key Words:
Impairment, Low back pain, Range of motion, Roland-Morris Back Pain Questionnaire.
M Scott Sullivan
†
Lisa Donegan Shoaf
Daniel L Riddle
†
Dr Sullivan died December 14, 1998.
240
Physical Therapy . Volume 80 . Number 3 . March 2000
therapy research in the area of impairment and
disability relationships, Jette stated, “Physical ther-
apy clinical research needs to explicitly state and
then investigate the nature of the hypothesized relation-
ship between different impairments and specific disabil-
ities. Included in such research is an examination of the
impact of changes in impairments on change in disabil-
ity and the investigation of important covariates that
alter these relationships. There is a paucity of examples
of such research in all the health professions’ literature,
not only in physical therapy.”
1(p969)
part, to identify impairments of ROM that influence the
patient’s disability. Identification of impairments is also
an integral component of treatment planning in physical
therapy.
3
Jette et al
4
reported that increased spinal ROM
was a treatment goal in 57% of care episodes for LBP;
this goal was the second most frequently cited following
the goal of reducing pain.
The data of Battieetal
2
and Jette et al
4
suggest that
physical therapists believe that spinal ROM and disability
are closely linked. Research has indicated, however, that
the correlation between spinal ROM and disability is weak.
In perhaps the most extensive study of the impairment-
disability relationship in patients with LBP, Waddell and
colleagues
5
measured different types of impairments
(eg, abdominal muscle performance, spinal ROM) in
120 patients with chronic LBP. Patients also completed a
Roland-Morris Back Pain Questionnaire (RMQ). Among
the impairments studied were those affecting lumbar
flexion and trunk flexion active range of motion
Physical therapists routinely assess for impairments of
spinal range of motion (ROM) in people with low back
pain (LBP). Batti and colleagues,
2
for example, found
that, when a large group of physical therapists in the
state of Washington were surveyed, 81% to 93% stated
that they would assess spinal ROM, given 3 hypothetical
patient cases. Presumably, spinal ROM is examined, in
MS Sullivan, PT, PhD, was Associate Professor, Department of Physical Therapy, School of Allied Health Professions, Medical College of Virginia
Campus, Virginia Commonwealth University, when this study was conducted.
LD Shoaf, PT, MS, is Assistant Professor, Department of Physical Therapy, School of Allied Health Professions, Medical College of Virginia
Campus, Virginia Commonwealth University.
DL Riddle, PT, PhD, is Associate Professor, Department of Physical Therapy, School of Allied Health Professions, Medical College of Virginia
Campus, Virginia Commonwealth University, 1200 East Broad, Richmond, VA 23298-0024 (USA) (driddle@hsc.vcu.edu). Address all correspon-
dence to Dr Riddle.
All authors provided concept/research design, writing, and data analysis. Dr Sullivan and Ms Shoaf provided data collection and project
management, and Dr Sullivan provided fund procurement. Physical therapists at Rockingham Memorial Hospital (Harrisonburg, Va), Island
Sports Physiotherapy (Coram, NY), and Medical College of Virginia Hospitals (Richmond, Va) assisted with data collection and provision of
subjects. Jill Binkley, Janet Freburger, and Paul Stratford provided reviews of an earlier version of the manuscript.
The study was approved by the Committee on the Conduct of Human Research at Virginia Commonwealth University.
This work was supported by a grant from The AD Williams Trust Fund, Medical College of Virginia Campus, Virginia Commonwealth University.
One of the 5 inclinometers used in this study was donated by The Saunders Group, Chaska, Minn.
This article was submitted February 22, 1999, and was accepted October 12, 1999.
Physical Therapy . Volume 80 . Number 3 . March 2000
Sullivan et al . 241
I
n a recent article discussing the need for physical
Table 1.
Relationships Between Spine Range-of-Motion Impairment and Disability in Patients With Low Back Pain
Author
Sample Size and Type
Disability Measure Range-of-Motion Measure Correlation (r)
Waddell et al
5
120 patients with LBP
.
3 mo
Roland-Morris Scale
Single inclinometer total flexion
2
.47
Dual inclinometer lumbar
flexion
2
.44
Single inclinometer total
extension
2
.33
Single inclinometer average of
right and left lateral flexion
2
.35
Rainville et al
11
89 patients with LBP
.
3 mo
Million Visual Analog
Scale
Dual inclinometer lumbar
flexion
.37
Single inclinometer total flexion
.33
Deyo and Diehl
6
80 patients, majority with acute LBP Sickness Impact Profile Fingertip-to-floor
.48
a
Waddell and Main
8
160 patients with LBP
.
3 mo
Waddell and Main
Disability Index
Tape measure method of Moll
and Wright
9
.35
Gronblad et al
12
55 patients with LBP
.
3 mo
Oswestry Disability
Questionnaire
Dual inclinometer lumbar
flexion
.09
Dual inclinometer lumbar
extension
2
.30
Tape measure mean of right
and left truck side bending
2
.24
Single inclinometer mean of
right and left rotation
.34
Deyo
10
129 patients, majority with acute
LBP
Sickness Impact Profile Fingertip-to-floor
.30
a
Roland-Morris Scale
.42
a
a
Spearman rho (
r
).
(AROM). Lumbar flexion was measured by the use of an
inclinometer positioned on the skin overlying the S2 and
then the L1 spinous processes while the patient was
upright and again when the spine was fully flexed. A
measure of lumbar flexion was then derived by subtract-
ing the values obtained in the starting position from the
values obtained in the fully flexed position. The corre-
lation (Pearson
r
) between lumbar flexion AROM and
disability was .44. Total flexion, a measure obtained by
positioning an inclinometer on the skin overlying L1
immediately before and after the patient maximally
flexes the spine from a standing position, was also weakly
correlated to disability (
r
.35) with the Waddell and Main
Disability Index. Other studies
10 –12
examining the ROM
impairment-to-disability relationship for patients with
LBP are summarized in Table 1. All studies summarized
in Table 1 used linear models to describe the impairment-
disability relationship. No studies were found that used
nonlinear models. The data in Table 1 suggest that
impairment-disability relationships are generally weak
for patients with LBP and that impaired spinal flexion
tends to be the spinal impairment most strongly related
to disability. In the studies summarized in Table 1, the
researchers only reported point estimates for the corre-
lations. Confidence intervals (CIs) were not reported,
and it may be that, if interval estimates were reported,
they may actually overlap for many of the studies.
5
.47), as measured with the
RMQ. The remaining impairments that were assessed
(eg, those involving spinal extension, lumbar lordosis,
pelvic flexion, spinal lateral flexion) had weaker
impairment-disability relationships (Pearson
r
5
5
.03–.35)
compared with the flexion measures.
We found only one study in which the relationship
between impairment and disability change scores follow-
ing treatment was examined. Deyo and Centor
13
exam-
ined 114 patients with LBP, 80% of whom had symptoms
for less than 1 month. The patients’ trunk flexion was
assessed using the fingertip-to-floor method, and they
completed an RMQ. Scores were obtained at an initial
visit and a 3-week follow-up visit. A Pearson
r
of .29 was
Other authors tend to agree with the work of Waddell
et al.
5
Deyo and Diehl
6
found a Pearson
r
of .48 for
the correlation between spinal flexion AROM (using a
fingertip-to-floor method) and disability (as measured by
the Sickness Impact Profile).
7
In an earlier study, Wad-
dell and colleagues
8
found that lumbar flexion AROM
measurements— obtained using the tape measure
242 . Sullivan et al
Physical Therapy . Volume 80 . Number 3 . March 2000
method described by Moll and Wright
9
—were weakly
correlated (Pearson
r
found for the correlation between the change in spinal
flexion and the change in RMQ scores.
relevance of the impairments. For example, if a patient
reportedly had difficulty with activities that required
sitting and bending forward and the therapist found that
the patient’s lumbar flexion was limited and painful
during AROM testing, the therapist may conclude that
the limited lumbar flexion is strongly associated with the
patient’s disability. In this case, the lumbar flexion
AROM impairment might be viewed as a clinically rele-
vant impairment. However, if a patient was judged to
have limited lumbar flexion AROM, but the patient only
had difficulty with walking-related activities, the therapist
may conclude that the limited lumbar flexion was not
associated with the patient’s disability. In this case, the
lumbar flexion AROM impairment would not be consid-
ered clinically relevant. We suspected that the linear
relationship between lumbar flexion AROM impairment
and disability would be stronger for patients judged to
have a clinically relevant impairment of lumbar flexion
AROM compared with patients whose lumbar flexion
AROM measure was judged to be not relevant to their
disability.
Some of the variation in the estimates of the relationship
of impairment to disability summarized in Table 1 may
be due to the different methods used to measure AROM
impairment. There are a variety of methods used to
assess lumbar spine flexion AROM.
14
One instrument
used to measure lumbar spine flexion AROM is the
inclinometer. At least 2 methods of measurement of
spinal flexion AROM using an inclinometer have been
described.
5,15,16
One method, recommended in the
American Medical Association’s
Guides to the Evaluation of
Permanent Impairment
,
15
has been criticized because of
the lack of substantive normative data that may be used
to determine when an impairment of lumbar spine
flexion AROM is present.
17
An alternative method for
determining lumbar spine flexion AROM has been
proposed by Troup and colleagues
16,18
and was used in
our study. A reliability study conducted on a sample of
335 subjects, most of whom were asymptomatic, sug-
gested that measurements obtained with this procedure
are reliable (Pearson
r
.91).
18
One advantage of the
method proposed by Troup and colleagues is that a
normative database has been developed. The data have
been stratified by age and sex, and they can be used to
determine whether impairment in lumbar flexion
AROM is present in patients with LBP.
17
No evidence
was found that indicated the inclinometer method used
in our study was valid for inferring the actual amount of
flexion in the lumbar spine. Evidence does exist to
indicate a dual inclinometer method similar to that used
in our study is valid based on comparisons with radio-
graphic measurements. Saur et al
19
found that the
Pearson
r
correlation between a dual inclinometer tech-
nique and a radiographic measure of lumbar flexion was
.98 for 54 patients with LBP.
We also used a normative data approach to assess the
lumbar flexion AROM impairment-disability relation-
ship. Because the method used to collect data in this
study was identical to the method used by Troup et al
16
and Sullivan et al,
17
we could compare our data with
the normative data. Theoretically, patients with more
severe limitations in lumbar flexion AROM should dem-
onstrate a stronger impairment-disability relationship
than patients whose AROM is judged to be “normal”
based on the normative data. We suspected that patients
whose lumbar flexion AROM was greater than 1 stan-
dard deviation below that of an age- and sex-matched
normative sample would have a stronger impairment-
disability relationship than patients who were within 1
standard deviation of the mean for the normative data.
Although a weak linear relationship between lumbar
spine flexion AROM and disability has repeatedly been
found in heterogeneous groups of patients, physical
therapists may still hypothesize that a strong linear
relationship between impaired lumbar spine flexion
AROM and disability exists for a given patient. We found
no studies in the literature that attempted to identify
patient characteristics that may influence the impairment-
disability relationship.
The purpose of our study was to assess the relationship
between lumbar flexion AROM impairment and disabil-
ity from 3 perspectives. First, we determined the relation-
ship between lumbar flexion AROM impairment and
disability for the entire sample. Second, we compared
the impairment-disability relationship for patients
judged to have a clinically relevant impairment with that
of patients judged not to have a clinically relevant
impairment. Third, we compared the impairment-
disability relationship for patients judged to have limited
lumbar flexion AROM based on normative data with that
of patients judged not to have limited flexion AROM.
The relationships were examined for measurements of
lumbar flexion AROM and disability obtained when
patients were admitted to the study and for the change
scores derived from measurements obtained at admis-
sion and at discharge. We tested several hypotheses:
One approach to identifying subgroups of patients with
stronger impairment-disability relationships is to deter-
mine whether the therapist concludes that the impair-
ment is clinically relevant.
Clinical relevance
, in this con-
text, deals with whether the therapist believes the
impairment is associated with the disability. We believe
many therapists not only look for the presence of
impairments, they also make judgments of the clinical
Physical Therapy . Volume 80 . Number 3 . March 2000
Sullivan et al . 243
5
1. We hypothesized that patients who were judged by
the participating therapists to have a clinically rele-
vant loss of lumbar flexion AROM would have a
stronger lumbar flexion AROM impairment-disability
relationship than patients who were not judged to
have a clinically relevant impairment of lumbar flex-
ion AROM. The first hypothesis was tested using the
lumbar flexion AROM impairment and disability
scores obtained during the patients’ first visit for
physical therapy.
Sample
Subjects.
A sample of convenience was chosen by
recruiting consecutive patients who met the inclusion
criteria at 5 outpatient physical therapy offices (3 facili-
ties were located in Virginia and 2 facilities were located
in New York). Inclusion criteria were: patients must be
between the ages of 18 and 75 years, patients must be
able to read English, and patients must be referred to
one of the participating facilities for treatment of LBP
with or without sciatica.
Low back pain
was defined as any
pain posterior to the midaxillary line between T12 and
the gluteal folds.
Sciatica
was defined as any lower-
extremity pain that was believed to be associated with
LBP, as determined by either the referring physician or
the physical therapist. Patients with any of the following
conditions, as determined by the referring physician,
were excluded: spondylolysis, spondylolisthesis, infec-
tious arthritis, spinal tumor, ankylosing spondylitis, or
idiopathic scoliosis. Patients who had spinal surgery or
who had neurological findings were admitted to the
study.
2. We hypothesized that the changes in the impairment
and disability scores of patients judged to have clini-
cally relevant lumbar flexion AROM impairments
would be more strongly correlated than the change in
the scores of patients judged not to have a clinically
relevant lumbar flexion AROM impairment. We
tested this hypothesis by using the change scores
derived from admission and discharge measures.
3. We hypothesized that patients with limited lumbar
flexion AROM, based on a normative data compari-
son, would have a stronger impairment-disability rela-
tionship than patients who did not have limited
lumbar flexion AROM. The third hypothesis was
tested using the lumbar flexion AROM impairment
and disability scores obtained during the patients’
first visit for physical therapy.
Between June 1994 and March 1995, a total of 116
patients were admitted to this study (Tab. 2). Thirty-five
patients, at some point in their rehabilitation, did not
return for completion of physical therapy treatment and
for follow-up measures. Eighty-one patients were fol-
lowed from the day of their initial physical therapy
evaluation until they were discharged from physical
therapy.
4. We hypothesized that changes in the impairment and
disability scores of patients judged to have limited
lumbar flexion AROM at admission, based on norma-
tive data, would be more strongly correlated than the
change in the scores of patients judged not to have
limited lumbar flexion AROM impairment. First, we
used the flexion AROM measurements obtained at
admission to identify 2 groups of patients: those
whose AROM was limited and those who did not have
limited AROM based on a normative data compari-
son. Second, the hypothesis was tested by using the
change in the admission and disability scores for the
2 groups.
2–20
years) participated in this study. Three clinics each
employed 2 therapists, 4 therapists worked at 1 clinic,
and 5 therapists worked at the fifth clinic. One of the 15
therapists was an orthopedic certified specialist, and all
therapists routinely treated patients with orthopedic
problems. At the time of the study, all physical therapists
worked full time in the participating outpatient ortho-
pedic settings.
10.2 years of experience, SD
5
3 years, range
5
Method
Procedure
Physical therapists employed at the participating facili-
ties recruited patients who met the inclusion criteria.
After agreeing to participate, each patient signed an
informed consent form and completed 2 questionnaires:
a brief demographic questionnaire and the RMQ.
20,21
Instructions for completion of the RMQ were printed
according to the methods described by the question-
naire’s originators. Two 10-cm visual analog scales— one
for current LBP and the other for pain other than
LBP—were attached to the RMQ and used for descrip-
tive purposes only.
Design
This was a pretreatment-posttreatment observational
study. Data were collected at 2 points in time: on the day
of the initial evaluation and on the day the patient was
discharged from physical therapy. We used the discharge
data because we believed these data would maximize the
variance in change scores. Some patients were likely to
change slightly or not at all, whereas others were
expected to show large changes in AROM and disability.
The mean time between admission and discharge was 51
days (SD
5
41 days, range
5
2–210 days).
244 . Sullivan et al
Physical Therapy . Volume 80 . Number 3 . March 2000
Physical therapists.
A total of 15 physical therapists
(X
5
[ Pobierz całość w formacie PDF ]

zanotowane.pl

doc.pisz.pl

pdf.pisz.pl

shinnobi.opx.pl