57
Interrater Reliability of Judgments of the Centralization
Phenomenon and Status Change During Movement Testing
in Patients With Low Back Pain
Julie M. Fritz, PhD, PT, ATC, Anthony Del&o, PhD, PT, Michelle Vignovic, MS, PT, Robert G. Busse, MS, PT
ABSTRACT. Fritz JM, Delitto A, Vignovic M, Busse RG.
Interrater reliability of judgments of the centralization phenom-
enon and status change during movement testing in patients
with low back pain. Arch Phys Med Rehabi12000;8 1:57-6 1.
Objective: To determine the interrater reliability of judg-
ments of status change, including the centralization phenom-
enon during examination of the lumbar spine, and to determine
the effects of training and experience on reliability.
Design: A videotape study of judgments by physical thera-
pists and physical therapy students of the results of movement
testing during the examination of patients with low back pain.
Setting: Outpatient physical therapy clinic.
Patients: Patients receiving physical therapy treatment for
low back pain.
Intervention: Patients with low back pain were videotaped
while performing a variety of movement tests including single,
repeated, and sustained movements. Forty licensed physical
therapists and 40 physical therapy students were provided with
operational definitions of the three potential judgments of status
change with movement testing; centralization, peripheraliza-
tion, status quo. All therapists and students viewed the video-
tape and made a judgment regarding the patient’s status change
in response to the test.
Main Outcome Measure: Percentage agreement and kappa
coefficient values for agreement.
Results: Inter-rater reliability was excellent for the total
sample of examiners (kappa = .793), for the licensed physical
therapists (kappa = .823), and for the students (kappa = .763).
Conclusions: Judgments of status change are reliable when
operational definitions are provided. Clinical experience does
not appear to substantially improve reliability.
Key Words: Low back pain; Reliability and validity; Move-
ment; Rehabilitation.
0 2000 by the American Congress of Rehabilitation Medi-
cine and the American Academy of Physical Medicine and
Rehabilitation
T HE MCKENZIE’ EXAMINATION and treatment system
is a widely used approach to the nonsurgical management
of patients with low back pain2 The McKenzie examination
utilizes patient history, a postural assessment, and active and
passive lumbar movement testing. Lumbar movement testing
may be performed in different positions (standing, sitting, prone,
From the Department of Physical Therapy, School of Health and Rehabilitation
Sciences, University of Pittsburgh, Pittsburgh, PA. Mr. Busse is presently associated
with the Spine Therapy Center, Oshkosh. WI.
Submitted for publication January 15. 1999. Accepted in revised form March 24,
1999.
No commercial party having a direct financial interest in the results of the research
supporting this article has or will confer a benefit upon the authors or upon any
organization with which the authors are associated.
Reprint requests to Julie Fritz, PhD, PT. 6035 Forbes Tower, Pittsburgh. PA 15260.
0 2000 by the American Congress of Rehabilitation Medicine and the American
Academy of Physical Medicine and Rehabilitation
0003-9993/00/8101-5410$3.00/0
supine) and may include repeated or sustained movements. Exami-
nation results are used to place patients into one of three syndromes
(postural, dysfunction, or derangement),r with treatment based
on the syndrome to which the patient is assigned.
An important component of the McKenzie examination
system is judging the effects of lumbar movements on the
patient’s status. Each test movement can be judged as having
one of three possible effects; centralization, peripheralization,
or no change (ie, status quo). Centralization was originally
described by McKenzie as a phenomenon occurring during
lumbar movement testing when the patient reports that the pain
moves from an area more distal or lateral to a location more
central or near midline position in the lumbar spine. Peripheral-
ization occurs when the patient reports the movement of pain
from an area more proximal in the lumbar spine to an area more
distal or lateral. Movements that do not produce centralization
or peripheralization are judged to be status quo.‘v3
The centralization phenomenon has been proposed to have
prognostic value in patients with low back pain. Donelson and
associatesi evaluated 87 patients with low back pain and
radiation into the buttock, thigh, or calf, and found that patients
in whom centralization occurred during lumbar movement
testing were more likely to achieve good or excellent results
with treatment. Long5 studied 243 patients with chronic low
back pain who were entering a work hardening program and
found that centralization during the initial evaluation was
associated with greater reductions in pain and higher percent-
ages of return to work after completion of the program. Karas
and colleagues6 examined 126 patients and found that the
inability to centralize symptoms during the initial evaluation
decreased the likelihood of return to work within 6 months. The
centralization phenomenon appears to be of prognostic value,
yet the reliability of its determination has not been studied.
Riddle and Rothstein evaluated the reliability of the Mc-
Kenzie examination system and found unacceptably poor
inter-rater reliability for the placement of patients into one of the
three syndromes (kappa value = .26). They suggested that a poten-
tial source of unreliabiity was the determination of centralization or
peripheral&ion with movements, although the reliability of this
judgment was not reported individually. The purpose of this study
was to determine the interrater reliability of judgments regard-
ing the effect of lumbar movement testing on pain in patients
with low back pain. A second purpose was to examine the effect
of clinical experience on reliability.
METHODS
Subjects
In this study, licensed physical therapists and physical
therapy students interpreted movement tests via videotaped
portions of clinical examination procedures performed on 12
patients receiving physical therapy treatment for low back pain.
This study was approved by the Institutional Review Board of
the University of Pittsburgh Medical Center, and all patients
Arch Phys Med Rehebil Vol 81, January 2000
58 RELIABILITY OF THE CENTRALlZATlON PHENOMENON, Fritz

gave their informed consent. Patients were excluded if they Table 1: Test Movements Used in the Final Videotape
demonstrated frank neurologic signs including loss of bowel/ Movement Definition
bladder control or saddle region paresthesia, ongoing preg-
nancy, or three or more signs of abnormal illness behavior as Side-bending in standing Patient is standing; the examiner asks
defined by Waddell and coworkers.8 Seven men and five (1) the patient to bend in the frontal
women participated in the study. Eight patients had symptoms plane to the right or left as far as pos-
extending into the buttock and/or lower extremity; four had sible, then return to the starting posi-
symptoms only in the lumbar spine. tion.
Flexion in standing (4) Patient is standing, the examiner asks
Examiners the patient to bend forward as far as
Participating physical therapists were recruited from local possible without flexing the knees,
outpatient clinics treating a mostly orthopedic caseload. Physi- then return to the starting position.
cal therapy students were recruited from two local entry-level Repeated flexion in The flexion in standing movement is
programs; all participating students had already completed standing (2) repeated 10 times.
formal instruction in evaluation and treatment of the spine, but Extension in standing (3) Patient is standing; the examiner asks
had not received specific instruction in the McKenzie approach the patient to bend backwards as far
to evaluation and treatment. Forty practicing therapists and 40 as possible without flexing the knees,
students gave informed consent to participate in this study. The then return to the starting position.
average age of the therapists was 32.2 years (standard deviation Repeated extension in The extension in standing movement is
[SD] = 8.0). with an average of 7.5 years of experience. The standing (3) repeated 10 times.
average age of the students was 26.8 years (SD = 5.8). Sustained extension in The extension in standing movement is
standing (2) maintained for 30 seconds before
Procedure returning to the starting position.
The 12 patients were videotaped while being examined by Pelvic translocation in Patient is standing; the examiner pas-
one therapist. Appropriate test movements to be used in the standing (2) sively shifts the patient’s pelvis in the
examination of each patient were determined by using the frontal plane while stabilizing the
decision-making algorithms proposed by Delitto and associ- shoulders, then returns the patient to
ates.3 In this system, single movements in standing including the starting position.
flexion, extension, and lateral bending are performed, followed Extension in prone (1) Patient is prone; the examiner asks the
by the performance of test movements in alternative positions patient to press-up by placing his or
(eg, quadruped, prone, etc), or repeated or sustained move- her hands on the examining surface
ments. The purpose is to identify a movement causing central- and extending the elbows while
ization or peripheralization; any movement causing peripheral- keeping the pelvis flat on the surface,
ization is not tested further.3 A standardized testing procedure then return to the starting position.
was used. For each test movement, the examiner first asked the Sustained extension in The extension in prone movement is
patient to describe the nature and location of his or her prone (1) maintained for 30 seconds before
symptoms and to rate the intensity of the symptoms before returning to the starting position.
movement. After the test movement, the examiner again asked Sustained extension with Patient is prone; the examiner passively
the patient to describe the location, nature and intensity of the pelvic translocation in shifts the patient’s pelvis in the
symptoms. A rest of 10 to 15 seconds was provided and the prone (1) frontal plane. The patient is asked to
patient was asked to further describe any changes in symptoms prop up on his or her elbows with the
that occurred during this rest period. pelvis flat on the examining surface.
The 12 videotapes from the individual test movements were This position is maintained for 30 sec-
combined into a single video that included at least one example onds before returning to the starting
of each of the 13 different movement tests that were performed. position.
The single videotape consisted of 24 individual movement Repeated flexion in sitting Patient is sitting; the examiner asks the
tests. Some test movements were included more than once (1) patient to bend forward as far as pos-
(table 1). sible, then return to the starting posi-
Therapists and students who participated in the study were tion. This movement is repeated 10
given instruction booklets that provided operational definitions times.
of the terminology used in the study (table 2) and explained the Flexion in quadruped (2) Patient is in a quadruped position; the
videotape procedure. The most important consideration in examiner asks the patient to rock
determining whether centralization or peripheralization has backwards approximating the heels
occurred is the location of symptoms, and not their intensity.5.‘g to the buttocks, then return to the
For example, a patient who experiences movement of paresthe- starting position.
sia from the calf into the buttock and a concomitant increase in Repeated flexion in quad- The flexion in quadruped movement is
central low back pain with a lumbar movement would be ruped (1) repeated 10 times.
judged to have centralized despite the increased pain intensity. Numbers in parentheses indicate the number of occurrences of each
Students and therapists took a written examination to test movement in the test video.
their understanding of the concepts presented in the booklet,
and the test videotape was shown immediately after the written
examination was completed. All of the students and 21 thera- The videotape presented 26 individual movement test proce-
pists completed the written examination and viewed the video- dures, including the verbal exchange between the examiner and
tape under the supervision of one of the researchers. The other patient in which the patient explained the location and intensity
19 therapists were mailed the testing materials and viewed the of his or her symptoms before and after the performance of the
videotape without the direct supervision of a researcher. test movement. After each individual test movement, the

Arch Phys Med Rehabil Vol81, January 2000

 RELIABILITY OF THE CENTBALIZATION
Table 2: Operational Definitions for Judgements of Status Change
With Movement Testing
Possible judgments of a patient’s status change with move-
ment testing:
Peripheralizes-A neurological sign or paresthesia is pro-
duced or the patient’s paresthesia or pain
move distally from the lumbar spine.
Centralizes-A neurological sign is improved, or pares-
thesia or pain is abolished or move from the
periphery toward the lumbar spine.
Status Quo-Symptoms may increase or decrease in inten-
sity, but do not centralize or peripheralize.
therapist or student was asked to make a judgment regarding the
status change of the patient due to the movement test. Possible
responses were: (1) centralized, (2) peripheralized, or (3) is
status quo. The therapists and students were given as much time
as they needed to arrive at a judgment, but could not review the
movement test from the videotape.
Data Analysis
Categorical data were obtained for each student and therapist
for each of the 26 test movements. The percentage agreement
and kappa coefficients9 and 95% confidence intervals were
calculated for total agreement among all students and therapists
(n = 80), students only (n = 40), and therapists only (n = 40).
The kappa coefficient is the preferred statistic for reporting the
reliability of nominal or ordinal scale data because it represents
the agreement among raters beyond that expected by chance. In
addition, agreement among therapists with less than 6 years of
clinical experience (n = 20) and 6 years or more of experience
(11 = 20) were calculated using a kappa coefficient.
RESULTS
Table 3 summarizes the percentage agreement and kappa
coefficients with 95% confidence intervals for the groups of
therapists and students. The percentage agreement for the total
sample (n = 80) was 87.8%; it was 89.7% for licensed physical
therapists and 85.9% for physical therapy students. The percent-
age agreement for individual movements tests ranged from
46.3% to lOO%, with a median percentage agreement of 97.5%
for the total sample (n = 80). The kappa value for the total
sample (n = 80) was .793, indicating substantial clinical agree-
ment.‘O For the licensed physical therapists (n = 40) the Kappa
value was .823, and for the physical therapy students (n = 40) it
was ,763. Physical therapists with less than 6 years of clinical
Table 3: Kappa Coefficients and Percentage Agreement for the
Total Sample and Specified Subgroups of Examiners
95% Confidence
Percentage Kappa Interval for
Agreement Coefficient Kappa Coefficient
Total sample (I-I = 80) 87.8% ,793 .777-.809
Licensed physical thera-
pists (n = 40) 89.7% .823 .814-,841
Physical therapy students
(n = 4 0 ) 85.9% ,763 .756-.770
Therapists with 6 or more
years of experience
(n = 2 0 ) 90.2% ,873 .861-,895
Therapists with less than 6
years of experience
(n = 2 0 ) 88.8% ,817 .806-,828
PHENOMENON, Fritz 59
experience had a kappa value of .817, whereas therapists with 6
or more years of experience had a kappa value of .873.
DISCUSSION
The purpose of this study was to assess the interrater
reliability of clinical judgments made regarding a patient’s
response to specific movement tests using videotaped examina-
tion segments. We chose to use videotaping to allow for a
substantial number of evaluators to make judgments without
placing undue stress on the patient, potentially altering his or
her status and confounding the reliability. The use of videotap-
ing has been recommended as an alternative to a test-retest
design for assessing interrater reliability of patients with low
back pain,” but its use has not previously been reported. Both
physical therapy students and licensed physical therapists with
varying amounts of clinical experience demonstrated substan-
tial reliability, with all kappa values being greater than .76.
A classification scheme, such as the one proposed by
McKenzie, must be reliable if it is to be of use clinically in
determining appropriate treatment procedures.‘* For a classifi-
cation scheme to be reliable, it is necessary that the key
examination procedures within the scheme have documented
clinical reliability. The reliability of movement testing has been
investigated to a degree by other authors. McCombe and
colleagues’3 and Strender and associates’4 evaluated the inter-
rater reliability of the judgment of pain production during single
test movements of flexion, extension, and lateral bending of the
lumbar spine with kappa values ranging from .3 1 to .76. Spratt
and coworkers15 conducted a similar study, but included single
and repeated test movements in standing and prone, and
reported adjusted percent agreement values between .46 and
.59. Each of these studies assessed judgments regarding the
production or aggravation of pain with movement testing. The
interrater reliability of judgements of centralization or peripher-
alization with movement testing has not been previously
examined. It is centralization and peripheralization, however,
and not pain production that have been proposed to have
prognostic value and importance in clinical decision-mak-
ing. I .3.4.6. I6
The classification of a patient using any system is essentially
a two-step process. First, the examiner must interact with the
patient and perform the individual tests and measurements
called for by the system. Second, the examiner must combine
the information gathered using his or her cognitive processes
and the decision rules specified by the system and make a
judgment regarding the classification of the patient. Insufficient
reliability could occur at either step. Riddle and Rothstein’
examined the interrater reliability of the McKenzie examination
scheme in the assessment of patients with low back pain and
found poor reliability. Because these authors did not report
individual reliability coefficients for the components of the
McKenzie examination system, it was not possible from their
work to determine which step was responsible for the lack of
reliability. The judgment of centralization or peripheralization
has been recommended as an important diagnostic finding by
several authors other than McKenzie; therefore, we considered
it important to establish the reliability of this judgment distinct
from the overall McKenzie system.
Several methods of improving interrater reliability of clinical
examination procedures and classification schemes have been
proposed. The amount of clinical experience and formal
instruction (ie, continuing education) in examination proce-
dures and classification rules has been suggested as a necessary
prerequisite to improving reliability.‘4v’7 Another recommenda-
tion has been to improve the operational definitions of the
examination procedures used and the decision rules for judging
Arch Phys Med Rehabil Vol 81, January 2000
60 RELIABILITY OF THE CENTRALIZATION
the test results.7*18 The results of our study support the latter
method of improving reliability. In this study, examiners were
provided with operational definitions and decisiop rules for
judging the movement testing. Furthermore, the examination
procedure was standardized and consistently performed across
patients.
We believe the operational definitions and the degree of
standardization used in the present study account for the high
reliability coefficients obtained. We also found little difference
in the reliability between students and practicing therapists, or
between therapists with more or less clinical experience. All
participating examiners, regardless of the level of clinical
experience and formal education, were able to demonstrate
excellent interrater reliability when provided with definitions
and decision rules, further evidence for the role of precise
operational definitions and standardized examination tech-
niques, as opposed to additional formal education and training,
in the improvement of clinical reliability. These results are
consistent with those of other authors who have not found a
profound difference between experienced and novice examiners
when performing orthopedic evaluation procedures.‘9.20
The need for clarification of the operational definitions of
judgments of status change with movement testing is evident
from a review of the use of the concept of centralization.
McKenzie defined the centralization phenomenon as “the
situation in which pain arising from the spine and felt laterally
from the midline or distally, is reduced and transferred to a more
central position when certain movements are performed.“’ This
definition has been expanded by some to include decreased pain
with movement testing.6 We believe that the original intent of
the centralization phenomenon as proposed by McKenzie
should be recaptured, namely that symptom location, not
aggravation, is the key judgment. We therefore propose the
operational definitions of status change listed in table 2, which
demonstrated high levels of reliability in this study.
The use of specific exercise regimens, such as those proposed
by McKenzie, has not been substantiated by randomized
controlled trials. General reconditioning exercises and low-
stress aerobic activities have been recommended for patients
with a c u t e low back pain based on systematic reviews of the
available literature.2’-‘3 Exercise therapy, including stretching
and back muscle strengthening exercises, has been found to be
effective in patients with chronic low back pain.22q24-27 There is
no evidence however, that specific exercise routines are more
effective than general approaches or that it is possible to
identify patients likely to respond to specific exercise tech-
~qUes~22.23,28
One possible explanation for the inability to develop a
method to identify specific patient characteristics that make an
individual more likely to respond to a specific exercise ap-
proach is the lack of reliable and valid examination and
treatment decision-making systems. McKenzie’s system for
placing patients into syndromes for the purpose of specifying
exercise strategies has been shown to lack overall reliability.’ A
recent randomized trial also failed to show superiority for the
McKenzie system over manipulation in the treatment of pa-
tients with low back pain.27 Individual concepts within the
McKenzie system, such as the centralization phenomenon,
however, may prove to be reliable and useful examination
procedures.
Several potential shortcomings of our study need to be
acknowledged. A potential for selection bias exists in the
choices of patients to be videotaped and the individual test
movements to be used in the final videotape. Patients were
selected to be representative of different classifications of
Arch Phys Mad Rehebil Vol 81, January 2000
PHENOMENON, Fritz
patients based on the system of Delitto and colleaguesi in an
attempt to ensure a variety of patient presentations. The
individual movement tests selected for the final videotape were
chosen by the researchers to include as many different combina-
tions of test movements and postures, and also to include
roughly equal numbers of patients thought by the examiners to
have peripheralized, centralized, and remained status quo. If the
prevalence of one of the three judgments had been especially
low, the Kappa values may have been artificially deflated.29
Another potential limitation is the lack of supervision of a data
collector during the viewing of the videotape by 19 of the 40
participating physical therapists.
We also recognize the inherent limitation of the use of
videotaping. The judgment of a clinician watching a movement
test on a videotape is not an exact reflection of the judgment
made in a clinic during a patient evaluation. We believe,
however, that the advantages of videotaping to ensure stability
of the patient across multiple judgments outweighs the disadvan-
tages. We also believe that any improvement in reliability based
on the use of videotapes rather than actual clinical examination
would not be substantial enough to overturn the results of this
study.
CONCLUSION
The determination of a patient’s response to movement
testing has been proposed as a key finding in the clinical
examination and classification of patient’s with low back pain.
The reliability of such judgments had not previously been
assessed. The present s t u d y found substantial agreement be-
tween examiners provided with clear definitions and decision
rules regarding the judgments of status change, regardless of the
level of training or clinical experience of the examiner.
References
1. McKenzie RA. The lumbar spine: mechanical diagnosis and
therapy. Waikanae, New Zealand: Spinal Publications Limited;
1989.
2. Battie MC, Cherkin DC, Dunn R, Ciol MA, Wheeler KJ.
Managing low back pain: attitudes and treatment preferences of
physical therapists. Phys Ther 1994;74:2 19-26.
3. Delitto A, Erhard RE, Bowling RW. A treatment-based classifica-
tion approach to low back syndrome: identifying and staging
patients for conservative management. Phys Ther 1995;75:470-89.
4. Donelson R. Silva G, Murphy K. Centralization phenomenon: Its
usefulness in evaluating and treating referred pain. Spine 1990;15:
211-3.
5. Long AL. The centralization phenomenon: its usefulness as a
predictor of outcome in conservative treatment of low back pain (a
pilot study). Spine 1995;20:25 13-21.
6. Karas R, McIntosh G, Hall H, Wilson L, Melles T. The relationship
between nonorganic signs and centralization of symptoms in the
prediction of return to work for patients with low back pain. Phys
Ther 1997;77:354-60.
7. Riddle DL, Rothstein JM. Intertester reliability of McKenzie’s
classifications of the syndrome types present in patients with low
back pain. Spine 1993;18:1333-44. . .
8. Waddell G. McCulloch JA. Kummel E. Venner RM. Nonoreanic
signs in low-back pain. Spine 1980;5:117-25.
9. Cohen J. A coefficient of agreement for nominal scales. Educ
Psycho1 Meas 1960;20:37-46.
IO. Landis RJ, Koch GG. The measurement of observer agreement for
categorical data. Biometrics 1977;33: 159-74.
Il. Delitto A, Shulman AD, Rose SJ, Strube MJ. Erhard RE, Bowling
RW, et al. Reliability of a clinical examination to classify patients
with low back syndrome. Phys Ther Pratt 1992; I : l-9.
12. Buchbinder R, Goel V, Bombardier C, Hogg-Johnson S. Classifica-
tion systems of soft tissue disorders of the neck and upper limb: Do
they satisfy methodological guidelines? J Clin Epidemiol 1996;49:
141-9.
 RELIABILITY OF THE CENTRALIZATION
13. McCombe PF, Fairbank JC, Cockersole BC, Pynsent PB. 1989
Volvo Award in clinical sciences. Reproducibility of physical signs
in low-back pain. Spine 1989;14:908-18.
14. &render LE, Sjoblom A, Sundell K, Ludwig R, Taube A.
Interexaminer reliability in physical examination of patients with
low back pain. Spine 1997;22:814-20.
15. Spratt KF, Lehmann TR. Weinstein JN, Sayre HA. A new approach
to the low-back physical examination. Behavioral assessment of
mechanical signs. Spine 1990;15:96-102.
16. Donelson R, Grant W, Kamp C. Medcalf R. Pain response to
sagittal end-range spinal motion. Spine 1991;16 Suppl:206-12.
17. Donelson R. Training program to learn McKenzie’s work [letter].
Spine 1994;19:1414-5.
18. Spratt KF, Weinstein JN, Lehmann TR, Woody J, Sayre H. Efficacy
of flexion and extension treatments incorporating braces for the
low back pain patients with retrodisplacement, spondylolisthesis
or normal sagittal translation. Spine 1993; 18: 1839-49.
19. Somers DL, Hanson JA, Kedzierski CM, Nestor KL, Quinlivan
KY. The influence of experience on the reliability of goniometric
and visual measurement of forefoot position. J Orthop Sports Phys
Ther 1997;25: 192-202.
20. Huber FE, lrrgang JJ, Hamer C, Lephart S. Intratester and
intertester reliability of the KT- 1000 arthrometer in the assessment
of posterior laxity of the knee. Am J Sports Med 1997;25:479-85.
21. Bigos S, Bowyer 0, Braen G. Brown K, Deyo R, Haldeman S, et
al. Acute low back problems in adults. AHCPR Publication
95-0642. Rockville (MD): Agency for Health Care Policy and
PHENOMENON, Fritz 61
Research, Public Health Service, Department of Health and
Human Services; 1994.
22. van Tulder MW, Koes BW, Bouter LM. Conservative treatment of
acute and chronic nonspecific low back pain. A systematic review
of randomized controlled trials of the most common interventions.
Spine 1997;22:2128-56.
23. Waddell G. Feder G, McIntosh M, Hutchinson A. Low back pain
evidence review. London: Royal College of General Practitioners;
1996.
24. Manniche C, Lundberg E, Christensen I, Bentzen L, Hesselsoe G.
Intensive dynamic back exercises for chronic low back pain: a
clinical triai. Pain 1993;41:53-63.
25. Lidstrom 1. Ohlund C. Eek C. Wallin L. Peterson LE. Nachemson
A. Mobility, strength, and fitness after a graded activity program
for patients with subacute low back pain. Spine 1992;17:6;11-32.
26. Devo RA. Walsh NE. Martin DC. Schoenfeld LS. Ramamurthv S.
A controlled trial of transcutaneous electrical stimulation (TENS)
and exercise for chronic low back pain. N Engl J Med 1990;322:
1627-34.
27. Cherkin DC, Deyo RA, Battie M, Street J, Barlow W. A compari-
son of physical therapy, chiropractic manipulation, and provision
of an educational booklet for the treatment of patients with low
back pain. N Engl J Med 1998;339:1021-99.
28. Faas A. Exercises: which ones are worth trying, for which patients,
and when? Spine 1996;21:2874-8.
29. Spitznagel EL, Helzer JE. A proposed solution to the base rate
problem in the kappa statistic. Arch Gen Psychiatry 1985;42:
725-8.
Arch Phys Med Rehabil Vol 81, January 2000