INTRAEXAMINER RELIABILITY OF IDENTIFYING A

DYSFUNCTIONAL SEGMENT IN THE THORACIC

AND LUMBAR SPINE

Louise Potter, BSc,a Christopher McCarthy, PhD,b and Jacqueline Oldham, PhDc

ABSTRACT

a Centre for ReCentral Manchestetal’s NHS Trust, M

b Centre for ReCentral Manchestetal’s NHS Trust, M

c Centre for ReCentral Manchestetal’s NHS Trust, M

Sources of supresearch.

Submit requestsRehabilitation Scichester and MancTrust, Oxford Roa(e-mail: louise.j.poPaper submitted0161-4754/$32.Copyright D 20doi:10.1016/j.jm

Objective: To examine the intrarater reliability of identifying a manipulable lesion in the lumbar and thoracic spine.

Methods: An experienced osteopath used dynamic and static examination to assess 12 asymptomatic subjects for signs

of joint dysfunction in the thoracic and lumbar spine. The selected segment was marked with an UV invisible mark. A

second examiner visualized these marks with an UV lamp and recorded them on acetates for analysis; this process was

then repeated an hour later. The distance from the marks to a fixed point was measured and within-day intrarater reliability

was calculated using intraclass correlation coefficients (ICCs).

Results: The ICC(1,1) for the thoracic spine was 0.70 (95% confidence interval [CI], 0.27-0.90). In the lumbar spine the

ICC(1,1) was 0.96 (95% CI, 0.87-0.99).

Conclusion: This study shows that the lumbar spine joint perceived to be the joint most likely to benefit from a high-

velocity low-amplitude thrust can be identified with good within-day reliability in an asymptomatic sample using a defined

examination protocol. However, the reliability in identifying a joint exhibiting signs of segmental dysfunction in the

thoracic spine was poor. (J Manipulative Physiol Ther 2006;29:203-207)

Key Indexing Terms: Reproducibility of Results; Low Back Pain; Palpation; Segmental Dysfunction

Back problems cause considerable pain and disability.1

One commonly used treatment for the management

of back pain is spinal manipulation, which is used

by physiotherapists, osteopaths, chiropractors, and by some

medical practitioners.

The diagnosis of a biomechanical joint dysfunction is

fundamental to classification of musculoskeletal disease2

and a reliable biomechanical diagnosis is necessary to

habilitation Science, University of Manchester,r and Manchester Children’s University Hospi-anchester, UK.habilitation Science, University of Manchester,r and Manchester Children’s University Hospi-anchester, UK.habilitation Science, University of Manchester,r and Manchester Children’s University Hospi-anchester, UK.

port: No external funds were provided for this

for reprints to: Louise Potter, BSc, Centre forence, University of Manchester, Central Man-hester Children’s University Hospital’s NHSd, M13 9WL Manchester, UKtter@student.manchester.ac.uk).April 19, 2005; in revised form July 20, 2005.0006 by National University of Health Sciences.pt.2006.01.005

justify the use of spinal manipulation to correct it. Palpation

is the most commonly used method to assess joint

dysfunction3,4 and is considered to be one of the most

informative aspects of physical examination for musculos-

keletal pain.5 Different professions have numerous texts

describing methods for the evaluation of spinal mobility

using motion palpation.6,7 Although diagnosis of a joint

dysfunction is considered an important prerequisite to spinal

manipulation, little assessment has been made of the

clinician’s ability to reliably identify a joint that is exhibiting

signs of biomechanical dysfunction; there are very few

reliability studies reported in the literature.

Christensen et al8 examined the intra- and interobserver

reliability of prone motion palpation for segmental dysfunc-

tion in the upper thoracic spine (n = 107). Two chiropractors

examined the subjects and the results showed good hour-

to-hour (j = 0.68) and good day-to-day (j = 0.4)

intraobserver reliability. Interobserver reliability was poor

(j = 0.24). Hawk et al2 conducted a reliability study for the

identification of a manipulable lesion in the lumbar spine

(n = 18), comparing the agreement between 4 chiropractors

who had various levels of clinical experience. j Analysis

showed intraobserver reliability was better than interob-

server reliability and that the more experienced clinician had

greater intraobserver agreement, suggesting that increased

clinical experience led to greater reliability. These findings

were echoed by French et al9 who reported that the average

203

Fig 1. Inclusion and exclusion criteria used in the study.

Fig 2. Subject lying prone after visualization and recording of thesecond UV mark in the thoracic spine. The acetate shows therecording of the first and second UV mark and marking of molesand the corners of the acetate to help with consistent placing of theacetate sheet after the second assessment.

204 Journal of Manipulative and Physiological TherapeuticsPotter et al

March/April 2006Identifying a Dysfunctional Spinal Joint

intraexaminer reliability was moderate (j = .47) and

interexaminer reliability was poor (j = .27).

These previous studies are confounded by the clinician

having first to correctly identify and then label the

appropriate spinal segment, introducing a margin for error.

There is some debate about the ability of manual therapists to

correctly identify and label spinal segmental levels. Studies

by Burton et al10 and Simmonds and Kumar11 argued that

there was a measure of inaccuracy in identifying a nominated

spinal segment when using UV marker pens (with spread up

to 35mm). However, Downey et al12 and Billis et al13

showed good intratester but poor intertester reproducibility

in a similar study. Their study highlights the fact there is no

literature available on the size of the spinous process, which

can be palpated through the skin, and the authors used an

arbitrary measurement point. Absolute location of a spinal

segmental level could only be determined by the use of

radiographs, which would be ethically unacceptable.

Previous studies have shown differing degrees of

reliability in identifying a joint by naming its segmental

level, and there is some debate about the ability of manual

therapists to reliably identify and place a mark over a named

vertebral segment. Therefore, the purpose of this study was

to examine the intraobserver reliability of identifying a

manipulable lesion in the lumbar and thoracic spine, by

assessing agreement of marking the joint rather than by

having to name the appropriate level.

METHODS

SubjectsA convenience sample of 12 asymptomatic (5 male, 7

female) subjects was recruited from the staff and students at

a local university. It is often accepted that asymptomatic

subjects have nonpainful dysfunctional spinal joints and,

thus, subjects were selected from an asymptomatic popula-

tion as they would have less baseline variability. The subjects

all completed a pretest screening questionnaire to ensure they

were pain-free and to determine if they had any preexisting

medical history that wouldmake themunsuitable for amanual

examination. Inclusion and exclusion criteria are presented

in Fig 1. Height and weight measurements were recorded to

calculate the body mass index (weight in kilograms divided

by the square of height in meters). The experimental

procedure was explained to the subjects and informed

consent was then obtained. The study protocol was approved

by the Manchester Local Research Ethics Committee.

CliniciansThe examining researcher was an osteopath with 9 years

of clinical experience and the second researcher, recording

the findings on acetates, was a physiotherapist with 14 years

of clinical experience.

ProcedureThe first part of the examination was a visual postural

analysis, performed with the subject standing. The

researcher observed the subject’s spine from the side and

back to visualize any postural asymmetries that might be

due to the presence of dysfunction in the subject’s spine.14

The subject was then asked to perform a series of spinal

movements: flexion, extension, and lateral flexion to the

left and right to determine if there was an asymmetry in

the spinal movements or reduction in range of movement.

Passive motion in the lumbar spine was then assessed

with the subject lying on his or her left side on an adjustable

height plinth. The clinician moved the subject’s spine into

flexion and back to extension using the bent legs as a lever.

During this flexion/extension movement, the clinician

palpated each of the lumbar spine interspinous spaces to

allow the examiner to identify areas of hypo/hypermobility

in the lumbar spine and to detect the presence of a

dysfunctional vertebral segment.15 In this position, static

palpation was also used to detect areas of muscle hyper-

tonicity or other palpatory clues (eg, pain, redness, or heat)

to the presence of joint dysfunction.14

Fig 3. Bland and Altman 95% limit of agreement plot showing theintrarater agreement of the UV marks in the thoracic spinebetween assessment 1 and 2.

Fig 4. Bland and Altman 95% limit of agreement plot showing theintrarater agreement of the UV marks in the lumbar spine betweenassessment 1 and 2.

Potter et alJournal of Manipulative and Physiological Therapeutics

Identifying a Dysfunctional Spinal JointVolume 29, Number 3205

Finally, the subject was placed prone on the table, and a

passive examination to assess areas of hypo/hypermobility

of the thoracic spine was performed by the clinician

applying an anterior cephalic force through the spinous

processes of the thoracic spine.15 Although this is a

commonly used assessment method in clinical practice, it

was not an examination technique that the researcher was

very experienced in. However, placing the pen mark on the

appropriate segment, it would not be possible to use the

habitual supine position.

The results of the examination findings allowed the

examiner to identify the segment that was perceived to be

the joint most clinically relevant to receive a spinal

manipulation in both the thoracic and lumbar spine. When

the joint had been identified, it was marked on the subject’s

skin by the first examiner using an UV marking pen at the

inferior edge of the superior spinous process.

The subject then went into a separate room and lay in a

prone position on a treatment table. The second examiner,

using a handheld, battery-operated UV lamp to detect the

previously described markings, recorded the mark on 2

acetates; one for the lumbar spine and one for the thoracic

spine. The corners of the acetates were marked on the

subject’s back with a colored pen and visual landmark

points, such as moles and scars, were also recorded on the

acetates to ensure the acetates were placed in an identical

position for the second recording (Fig 2).

The whole process was repeated, again with subjects in

a random order, immediately after the first examination.

This occurred within 1 hour of the first examination and,

by reexamining the subjects within this short period, did

not give them an opportunity to do any physical activities

that might have altered the mechanical function of their

spine. The main difficulty encountered in the study was

blinding the examiner to each subject; this was addressed

by bringing the subjects in to the room in groups of 4 and

then assessing each subject once and then again in a

random order. This reduced the chance of the examiner

remembering an individual participant, and the fact that the

examiner did not give the joint a label meant that they

could not recall a named segment.

Data AnalysisThe distance from each mark was measured to a fixed

point at the edge of the acetate and the measurement

recorded for both the thoracic and lumbar spine. A 1-way

analysis of variance16 was performed to test for the

variability in difference between the 2 measurements for

both spinal areas. The intraclass correlation coefficient

(ICC) was calculated as a test of intraobserver agreement,17

and Bland and Altman 95% limit of agreement plots were

drawn using SPSS version 10.1 for Windows (SPSS Inc,

Chicago, Ill).

RESULTS

The data appeared normally distributed, as confirmed by

testing the data using a Kolmogrov-Smirnov test (P N .05).

The age of the sample (n = 12, 5 male, 7 female) was 35.7F11.95 years. Body fat might be considered a hindrance to

palpation, but the mean of the sample body mass index

was 23.9 (range, 18-30), which indicates the sample ranged

from underweight to overweight, with none in the obese

category, thus limiting the problems associated with excess

adipose tissue.

206 Journal of Manipulative and Physiological TherapeuticsPotter et al

March/April 2006Identifying a Dysfunctional Spinal Joint

Reliability of identification of a dysfunctional joint in the

thoracic spine was moderate to poor represented by an

ICC(1,1) of 0.70 (95% CI, 0.27 to 0.90). In the lumbar spine,

reliability was excellent; the ICC(1,1) was 0.96 (95% CI,

0.87 to 0.99). Bland and Altman18 95% limits of agreement

plots revealed no systematic difference between the position

of the 2 marks (Figs 3 and 4).

DISCUSSION

These results show that in an asymptomatic sample using

a standardized assessment protocol the within-day intrarater

reliability for detecting the joint most suitable to receive a

spinal manipulation was excellent for clinical measurement

in the lumbar spine but moderate to poor in the thoracic

spine.19 The strength of this study is that it tests the

clinician’s ability to identify the most dysfunctional joint in

the area of the spine being examined without need for the

joint to be named or labeled. Using the UV marking pen

protocol removed the possible measurement error inherent

in needing to label the correct spinal segment.

No previous comparable study could be found in the

literature, although both French et al9 and Hawk et al2

showed that intrarater reliability was good when the

examiner was asked to identify whether each joint is in

need of a spinal manipulative thrust to correct joint

dysfunction. This study differs as it identifies the 1 joint

in 1 area of the spine that would be clinically relevant to

receive a manipulation. Because measurements were used

for agreement, quantitative statistical analysis was possible.

These findings add to the evidence that manual palpation

of the lumbar spine is a reliable method for establishing a

joint suitable for spinal manipulation. The results should be

interpreted with caution as the sample size is small and the

subjects were asymptomatic. It is often acknowledged that

segmental dysfunction can be detected in an asymptomatic

subject and it would be expected that in a symptomatic

population it would be easier to detect the joint with the

most dysfunctional biomechanics. Panzer20 reviewed

reports that soft tissue palpation reproducing pain is the

most reliable marker to detect the level of dysfunction. It

would be expected that in a symptomatic population,

palpation would reproduce pain and therefore be more

reliable, although this needs further investigation.

Using a combination of examination methods (observa-

tion, static palpation, dynamic assessment, passive joint

palpation) reproduces a clinical examination and may make

the results more generalizable to everyday practice. It was

possible that examination of the thoracic spine was less

reliable because of the particular assessment protocol

adopted, which was necessary to allow the joint to be

marked with the pen. This may support the findings of

Hawk et al2 that the more experienced a clinician is at a

particular examination, the more reliable the palpation

findings are likely to be. It is also possible that the thoracic

spine, because of its densely packed bony anatomy, is

intrinsically more difficult to accurately palpate and assess.

This is obviously an area requiring further examination and

having a practitioner experienced in prone palpation may

improve the reliability of the findings in the thoracic spine.

CONCLUSION

This study shows that a clinically experienced osteopath

can identify with good reliability the lumbar spine joint

perceived to be the joint most likely to benefit from a high-

velocity low-amplitude thrust, in an asymptomatic sample,

using a defined examination protocol. However, when using

a similarly defined protocol, the same practitioner, not

experienced at this particular examination method, had

moderate to poor reliability in identifying a joint exhibiting

signs of segmental dysfunction in the thoracic spine.

REFERENCES

1. Anderson GB, Lucente T, Davis AM. A comparison ofosteopathic spinal manipulation with standard care for patientswith low back pain. N Engl J Med 1999;341:1426-31.

2. Hawk C, Phongphua C, Bleecker J, Swank L, Lopez D, RubleyT. Preliminary study of the reliability of assessment proceduresfor indications for chiropractic adjustments of the lumbar spine.J Manipulative Physiol Ther 1999;22:382-9.

3. Lewit K, Liebenson C. Palpation problems and implications.J Manipulative Physiol Ther 1993;16:586-90.

4. Russell R. Diagnostic palpation of the spine. J ManipulativePhysiol Ther 1983;6:181-3.

5. Evans DH. The reliability of assessment parameters: accuracyand palpation technique. In: Grieve GP, editor. Modern manualtherapy. Edinburgh7 Churchill Livingstone; 1986. p. 498-502.

6. Maitland G. Vertebral manipulation. London7 Butterworth &Com; 1986. p. 3-4.

7. Schafer RC, Faye LJ. Motion palpation and chiropractictechnique. Huntington Beach (Calif)7 Motion Palpation Insti-tute; 1989. p. 143-92.

8. Christensen HW, Vach W, Vach K, Manniche C, Haghfelt T,Hartvigsen L, et al. Palpation of the upper thoracic spine: anobserver reliability study. J Manipulative Physiol Ther 2002;25:285-92.

9. French SD, Green S, Forbes A. Reliability of chiropracticmethods commonly used to detect manipulable lesions inpatients with chronic low-back pain. J Manipulative PhysiolTher 2000;23:231-8.

10. Burton AK, Edwards VA, Sykes DA. Invisible skin markingfor testing palpatory reliability. J Man Med 1990;5:27-9.

11. Simmonds MJ, Kumar S. Part ii: location of structures bypalpation: a reliability study. Int J Ind Ergon 1993;11:145-51.

12. Downey BJ, Taylor NF, Niere KR. Manipulative physiothera-pists can reliably palpate nominated lumbar spinal levels. ManTher 1999;4:151-6.

13. Billis EV, Foster NE, Wright CC. Reproducibility and repeat-ability: errors of three groups of physiotherapists in locatingspinal levels by palpation. Man Ther 2003;8:223-32.

14. Gibbons P, Tehan P. Manipulation of the spine, thorax andpelvis. London7 Harcourt Publishers Ltd; 2000. p. 5-7.

Potter et alJournal of Manipulative and Physiological Therapeutics

Identifying a Dysfunctional Spinal JointVolume 29, Number 3207

15. Maitland G, Hengeveld E, Banks K, English K. Maitland’svertebral manipulation. Oxford7 Butterworth-Heinemann;2002. p. 93-166.

16. Rankin G, Stokes M. Reliability of assessment tools inrehabilitation: an illustration of appropriate statistical analyses.Clin Rehabil 1998;12:187-99.

17. Bland M. An introduction to medical statistics. 3rd ed. Oxford7Oxford University Press; 2000. p. 204-5.

18. Bland CJ, Altman DG. Statistical methods for assessingagreement between two methods of clinical measurement.Lancet 1986;327:307-10.

19. Portney LG. Foundations of clinical research: applications topractice. Norwalk7 Appleton & Lange; 1993. p. 509-13.

20. Panzer D. The reliability of lumbar motion palpation. JManipulative Physiol Ther 1992;15:518-24.