Accepted Manuscript

The validity of a clinical test for the diagnosis of lumbar spinal stenosis

Robert Dobbs, Stephen May, Philip Hope

PII: S1356-689X(16)30357-5

DOI: 10.1016/j.math.2016.05.332

Reference: YMATH 1860

To appear in: Manual Therapy

Received Date: 11 January 2016

Revised Date: 23 May 2016

Accepted Date: 25 May 2016

Please cite this article as: Dobbs R, May S, Hope P, The validity of a clinical test for the diagnosis oflumbar spinal stenosis, Manual Therapy (2016), doi: 10.1016/j.math.2016.05.332.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

http://dx.doi.org/10.1016/j.math.2016.05.332

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

Title page

The validity of a clinical test for the diagnosis of lumbar spinal stenosis

Robert Dobbs1, Stephen May2*, Philip Hope1

1 Physiotherapy Department, Chesterfield Royal Hospital, Chesterfield, Derbyshire

2 38 Collegiate Crescent, Faculty of Health and Wellbeing, Sheffield Hallam University, Sheffield, S10 2BP

s.may@shu.ac.uk

0114 2252370

*Corresponding author

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPTThe validity of a clinical test for the diagnosis of lumbar spinal stenosis

Abstract

Background: The diagnosis and management of acquired lumbar spinal stenosis (ALSS) is

an area of growing interest with an increase in its prevalence and detection in the older

population.

Objectives: To investigate the diagnostic accuracy of a modified extension test (MExT) for

diagnosing ALSS in subjects aged fifty or over.

Methods: Symptomatic response of the bi-component MExT was evaluated and

compared against MRI findings in 30 subjects. Estimates of sensitivity, specificity, likelihood

ratios and post-test probabilities were all calculated, and the capability of the test to

discriminate between grade and location of stenosis was also appraised.

Results: MExT sensitivity was high at 92% (95% CI, 72-99%) leading to a significant

negative likelihood ratio at -LR 0.2 (95% CI, 0.03-1.36); conversely, specificity was low at

40% (95% CI, 7-82%) with only a small positive likelihood ratio of +LR 1.53 (95% CI, 0.74-

3.16). All correlations between the MExT and concurrent grade, or location of stenosis

appeared weak and insignificant.

Conclusions: The MExT was found to demonstrate acceptable criterion validity in relation to

ruling-out a diagnosis when a negative result was observed; however, further validation

appears warranted.

Words 184

Keywords: Lumbar, Spinal stenosis, Extension Test, Diagnosis, Validity

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

1

The validity of a clinical test for the diagnosis of lumbar spinal stenosis

1. INTRODUCTION

World epidemiological projections indicate that the number of people within the

population aged sixty-five or over will have increased from sixteen to twenty five

per cent by the year 2040 (Population Reference Bureau, 2010; Maloney-

Backstrom et al., 2011). In response to such demands, health research agendas

have increasingly shifted focus towards the management of degenerative

conditions proposing direct efficiency savings through enhanced examination and

treatment rigour (IAGG, 2007). One such condition gaining attention is acquired

lumbar spinal stenosis (ALSS). Point prevalence estimates suggest degenerative

lumbar conditions may afflict up to twenty per cent of the older-aged population

(Keller et al., 2003; Lyle et al., 2005). Of those individuals referred to an

orthopaedic spinal specialist, approximately fourteen per cent are found to exhibit

severe stenotic change requiring surgical decompression (Aalto et al., 2006; Slatis

et al., 2007; Watters et al., 2008). As such, ALSS has fast become the leading

cause of spinal surgery in this population (Mannion et al., 2010; Steurer 2010),

whilst its deleterious influence on locomotor capability and psychosocial aspects of

health renders it a precursor for falls, depression and cardiovascular disease

(Middleton and Fish, 2009; Kim et al., 2011).

The clinical syndrome of ALSS develops as a consequence of incremental

damage to spinal tissues. Narrowing of the intervertebral disc reduces the

distance between adjacent vertebrae, whilst an alteration to biomechanical force

initiates arthritic change and ligamentous laxity (Fredrickson et al., 2001; Vo et al.,

2005; Papadakis et al., 2011). Subsequent thickening of the ligamentum flavum,

pedicles or vertebral facets diminish the space available for neural and vascular

structures, with ischaemic change ensuing within the spinal canal (central

stenosis), nerve root canal, or intervertebral foramina (foraminal stenosis) (Vo et

al., 2005). Sufferers of ALSS present with symptoms of lower extremity pain or

paraesthesia, occurring with or without back pain especially with positions of

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

2

extension, which limit walking capacity secondary to neurogenic claudication (Bal

et al. 2006; Malmivaara et al., 2007; Ogikubo et al., 2007).

1.1 Diagnostic methods

However no widely accepted diagnostic criterion for ALSS currently exist (Goh et

al., 2004; Haig et al., 2006; Genevay et al., 2010; Genevay and Atlas, 2010).

Recent reviews have cited heterogeneity and unsatisfactory research standards,

which prevent conclusions on the performance of diagnostic tests being drawn

(De Graaf, 2006; Genevay and Atlas, 2010). Moreover, appraisal of these tests

employed found many to be grounded in anecdotal evidence rather than robust

measurement (Haig et al., 2006; Browne and Roberts, 2008; Deyo et al., 2009).

For this reason supplementary analysis of all diagnostic procedures is needed. A

recent review reported postural effects on symptoms to have the most useful

diagnostic criteria, whereas clinical tests were less consistently useful (de

Schepper et al. 2013).

With a distinctive clinical presentation initial diagnosis of ALSS is made in relation

to patient report alone; however, clinical prediction rules based solely on patient

report have only been done on small populations (Sugioka et al., 2008). Accurate

findings from physical testing would help to validate patient report, however

selection of these tests has proven problematic. Neurological examination is

commonly observed to be normal (Bassewitz and Herkowitz, 2001; Genevay and

Atlas, 2010), thus the most pertinent physical findings typically relate to

symptomatic change in accordance with lumbar movement. Although both

treadmill and bike tests have been postulated as useful when differentiating ALSS

from claudication arising from peripheral vascular disease (Fritz et al., 1997;

Tenhula et al., 2000; Deen et al., 2000; Yukawa et al., 2002), research findings

appear inconsistent (Dong and Porter, 1989; Moon et al., 2005).

Thus postural tests gauging the symptomatic response to lumbar extension

appears pertinent to the initial physical examination. Studies have exhibited the

extension movement to narrow spinal space by up to twenty per cent (Panjabi et

al., 1983; Inufusa et al., 1996; Fritz et al., 1998), with percentages substantially

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

3

increasing when coupled with degenerative changes (Schronstrom and Willen,

2001; Westergaard et al., 2009; Kishner et al., 2010). Regardless of a strong

theoretical foundation, only one study by Katz et al., (1995) is frequently cited by

other authors when appraising validity (Fritz et al., 1998; Lurie, 2005; Vo et al.,

2005; Westergaard et al., 2009). Evaluating a range of physical components this

study observed a small positive likelihood ratio of 1.6 with thirty-seconds of lumbar

extension and provocation of at least thigh pain. Positive findings on coupled

quadrant movements of extension and side-flexion is recognised as ‘Kemps sign’

(Jenis and Howard, 2000), and is stated to occur frequently in foraminal stenosis

(Watanabe et al., 2010; Eguchi et al., 2010). Thus further evaluation of 'Kemps

sign' or combined extension and side-flexion movement seems warranted, as it

would be a simple and straightforward test to employ in the examination of those

with suspected ALSS, because of age and findings from the history.

1.2 Aims

The aim of this study was to prospectively evaluate the validity of the ‘Modified

Extension Test’ (MExT) an extension-based test proposed by the authors. This

test gauges the symptomatic response to lumbar movement, and is postulated to

support the diagnosis of ALSS whether located centrally or foraminally. Moreover,

no consistent information currently exists regarding whether symptoms occur more

rapidly following the commencement of extension in those with severe stenotic

change. Such discernments would support the identification of those individuals

more likely to require surgical consideration (Atlas et al., 1996; Amundsen et al.,

2000; Benoist, 2002; Slatis, 2007; Watters et al., 2008), and hence, was selected

for sub-analysis despite the inconsistencies that there was such a link in previous

literature (Wang et al., 2008; Haig and Tomkins, 2010; Kishner et al., 2010).

2. METHODS

2.1 Ethical approval

This was a prospective blinded concurrent-criterion related validity design to

establish MExT sensitivity, specificity and predictive capability in diagnosing

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

4

ALSS. Ethical approval was granted by the Nottingham Two proportionate review

sub-committee, with site specific permission and insurance obtained in conjunction

with the local CRH NHS trust research and development team. Institutional

clearance was further established on behalf of Sheffield Hallam University Health

and Care Ethics Committee. All patients provided signed, informed consent prior

to participation.

2.2 Sample

Subjects aged fifty or over were recruited from Chesterfield Royal Hospital (CRH)

over a six-month period. Consecutive patients attending either the Extended

Scope Practitioner (ESP) or musculoskeletal (MSK) physiotherapy clinic were

screened, and where deemed eligible were invited to participate. Recruitment was

based on presenting symptoms, with inclusion criterion requiring subjective report

of unilateral or bilateral pain or paraesthesia radiating below the gluteal fold. In

accordance with guidelines for MRI scan referral, symptoms were required to have

been present for a minimum of six weeks (Bussieres et al 2008). Patients who

had a lack of mental capacity to provide informed consent, or were unable to

understand English, or there was a pre-existing diagnosis of peripheral vascular

disease, or of Ankylosing Spondylitis were excluded. A STARD flowchart is

provided in Figure 1.

In the absence of a defined ‘true’ diagnostic gold standard, Magnetic Resonance

Imaging (MRI) was selected as the reference test upon which to evaluate the

index test’s (MExT) validity. Whilst concern has been documented regarding the

relative inconsistency between clinical symptoms and degree of observed

radiographic change (Wang et al., 2008; Haig and Tomkins, 2010; Kishner et al.,

2010), clinical guidelines recommend its use as a primary tool in ALSS diagnosis

(Watters et al., 2008). MRI has demonstrated both significant positive likelihood

ratios, (between 8.1 and 16.2), and negative likelihood ratios (between 0.3-0.19)

establishing its usefulness for ruling in or out a diagnosis (Fritz et al., 1998).

Moreover, the reliability of image interpretation for ALSS has been established

with Lurie et al., (2008) observing Kappa scores of (0.82) and (0.83) for inter-rater

and intra-rater reliability respectively. The definitions used are in tables 1 and 2.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

5

Table 1: Definitions of central ALSS severity based upon cross-sectional area of

the canal and presence of cerebrospinal fluid.

Severity Definition

Severe ALSS >50% narrowing with absence of cerebrospinal fluid

Moderate ALSS >50% narrowing with more nerve root than cerebrospinal fluid

Mild ALSS

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

6

conversely, when a red ball was selected component two, (extension/side-flexion),

was the initial test component.

Each subject received the MExT (Figure 2), whilst an independent rater timed

each component up to the sixty second cut-off point. The result of the MExT was

documented against each component, and when positive, the associated time for

symptoms to occur was recorded. A positive test was deemed to be when pain

below the gluteal fold was exacerbated or produced with one or both components

of the test; exacerbation or production of back pain alone did not count. The

participant was subsequently monitored for ten minutes and advised on what

measures should be undertaken in the event of adverse effects. In order to limit

partial verification bias each participant was referred for an MRI scan irrespective

of MExT result (Pewsner et al., 2004; Whiting et al., 2004). The subsequent

lumbar MRI scan was undertaken on a 1.5 Tesla, Siemens, Avanto scanner within

a week of the index test, before being analysed by a consultant radiologist blinded

to clinical findings and index test result. The results from both the index and

reference tests were returned to the principal investigator and evaluated, whilst

each subject received an ESP follow-up appointment to discuss further

management options.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

7

Figure 2. Index Test ‘MExT’ Procedure

Initial set up

Standing position

S Component 1 Limb Placement

The participant adopts the neutral standing position, with the therapist situated to the side of the individual. The therapist assumes a comfortable step stance posture using the medial aspect of their knee (closest to the participant) to maintain the knee in extension and provide support to the participant (see limb placement image below). The therapist then assumes an appropriate hold upon which to facilitate the test. The participant is asked to place their arm closest to the therapist across their chest with their hand close to the contralateral shoulder. The therapist places one hand on the lower lumbar spine and the other is located across the participants arm. The adopted position allows appropriate comfort for participant and therapist. Prior to test commencement, the participant is advised that they should inform the therapist if any symptomatic change occurs.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

8

Component 1

Component 2

The therapist asks for the timer to commence on the command ‘Go’ as they gently manoeuvre the participant into their full available extension, remaining in that position for up to one minute. The procedure is terminated prior to one minute if either the leg symptoms have been reproduced (a positive test), or if the participant reports excessive lumbar spine discomfort to continue, (an inconclusive test). If the participant fails to report reproduction of their leg symptoms during this time period the test is deemed negative. The therapist terminates the test with the command ‘stop’ before the participant is returned to the neutral position and provided with a one minute rest to allow symptoms to return to normal prior to initiating stage three.

The participant is once again advised to inform the therapist should any symptomatic change occur. The therapist asks for the timer to commence as the participant is manoeuvred into full extension, and then side-flexion towards the symptomatic side. On the occasion that symptoms are bilateral the examiner selects only the most symptomatic side to test, with results only determined positive if symptoms are reproduced on this same side. If symptoms are reproduced on the contralateral side the test is deemed negative. The procedure is terminated in relation to the predefined criteria (as in component one) with all findings documented. The researcher asks the participant to rest for up to ten minutes following the procedure, taking the opportunity to answer any additional questions the participant may have and to ensure no immediate adverse effects have occurred. The physiotherapist only performs the test once to mirror clinical practice, whilst additionally reducing the likelihood of developing post-test soreness.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

9

2.4 Pilot study

Preceding commencement of the trial, a pilot study of five subjects was completed

with results not included in the final analysis. Study procedures were appraised

and adapted accordingly, with inter-rater reliability of both stopwatch recordings

and MRI interpretation found to be above the limit set for acceptable

reliability >0.75 (Streiner and Norman, 2003). The minimal detectable change was

at a 95% confidence level.

One person performed each component of the test while two other people

simultaneously did the stopwatch timing. The intra-class correlation coefficient

(ICC) was 0.99 whilst the standard error of measurement (SEM) was 1.72 and the

minimal detectable change was 4.76. Moreover, MRI interpretation evaluation

was based upon agreement of stenotic grade, with a kappa coefficient value of

0.83 (95% CI 0.48, 1.0).

2.5 Data Analysis

Data was analysed using SPSS (Statistical Package for the Social Sciences)

version 16.0. Construction of a two by two contingency table allowed for initial

identification of test validity, with point-estimates of sensitivity and specificity

ascertained using a calculator alongside 95% confidence intervals (Bland, 2000).

Sensitivity and specificity are measures of test performance, discriminating

between subjects with or without a specific disease. In conjunction with statistical

and clinical recommendation acceptable levels were duly fixed at > 80% (Streiner

and Norman, 2003; Riegelman, 2005).

Positive and negative likelihood ratios were further calculated alongside 95%

confidence intervals to provide primary analysis of test validity. Likelihood ratios

estimate the extent to which an individual with ALSS is more likely to test positive

or negative when compared to someone without the disease. Evaluation utilising

likelihood ratios is preferable to that of the more commonly used positive and

negative predictive values, as likelihood ratios are not dependent upon disease

prevalence (McGee, 2002). As such, a positive likelihood ratio of > 2.0 was

perceived to highlight an important increase in ALSS likelihood (McGee, 2002),

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

10

conversely, a negative likelihood ratio of < 0.5 was adjudged to propose a reduced

ALSS likelihood. Post-test probabilities of a subject having ALSS in relation to a

specific MExT result were established from the likelihood ratios, and confirmed in

association with Fagan’s nomogram (Akobeng, 2007; Campo et al., 2010).

Sub-analysis of the relationship between symptom onset following MExT

commencement and the grading of ALSS whether mild, moderate or severe was

undertaken using Spearman’s rank correlation coefficient. Similarly, the Phi

coefficient was utilised to establish whether an association existed between

positive findings on a specific MExT component and a diagnosis of central or

foraminal ALSS. Significant findings were deemed to be p

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

11

Myotome lossb (%)

Reflex lossc (%)

Positive SLRd (%)

Intermittent Claudication (%)

1 (3)

3 (10).

3 (10)

25 (82) a = areas of sensory abnormality; b = weakness in muscle compared to the other side; c = diminished or absent; d =

reduced range compared to the other side

In relation to index and reference tests, eighty-seven per cent of the sample tested

positive to at least one component of the MExT. In accordance with reference test

findings sample prevalence of ALSS was calculated as eighty-three per cent, with

thirty per cent found to have central stenosis, thirty-three per cent foraminal

stenosis, and seventeen per cent both central and foraminal forms. Three per cent

were positive for stenosis at the lateral recess with ten per cent additionally found

to have an associated degenerative spondylolithesis at either the L4/5 or L5/S1

level.

3.2 Sensitivity, specificity and likelihood ratios

Sensitivity, specificity, and positive and negative likelihood ratios are reported in

table 4. Pre and post-test probabilities were calculated from pre and post-test odds

and confirmed against Fagan’s nomogram. Whilst the pre-test probability of a

subject within the sample having ALSS was 83% the post-test probability

marginally increased to 88% when a positive MExT was observed. Likewise post-

test probabilities based on the negative likelihood ratio observed a decrease in the

likelihood of ALSS to 49% when a negative result was detected.

Sub-analysis of correlations between grade of stenosis and symptom onset

following MExT commencement (Table 5), failed to demonstrate a significant

correlation for either central or foraminal elements in conjunction with Spearman’s

correlation coefficient. In relation to central stenosis a negative correlation of R =

- 0.147 was observed with symptom onset more rapid in severe stenosis, but this

correlation was not statistically significant (p = 0.684). Similar outcomes were

observed in relation to foraminal stenosis with a negative correlation of R = -

0.211, which was not statistically significant (p = 0.418).

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

12

Consideration of MExT selectivity in relation to stenosis location failed to identify

notable correlations between the matched components (Table 6). Findings of

central stenosis and a positive finding on the on the extension component of the

test observed a statistically non-significant negligible association (r = 0 .11, p =

0.54). There was a statistically non-significant association identified between the

extension / lateral flexion component of the test and foraminal stenosis (r = 0.09,

p = 0.62). The extension component of the test and foraminal stenosis had an

association of (r = 0.24, p =0.19) whilst the extension / lateral flexion component of

the test and central stenosis had an association of (r = 0.15, p = 0.41), but neither

was significant. The lack of statistical significance could be due to small sample

size and lack of adequate statistical power.

Table 4: MRI compared to MExT (in participant numbers)

MRI MExT

+ve scan findings -ve scan findings +ve test 23 3

-ve test

2 2

Sensitivity (95% CI) 92% (72-99%) Specificity (95% CI) 40% (07-82% + LR (95% CI) 1.53 (0.74-3.16) - LR (95% CI) 0.2 (0.003-1.36) +ve = one or both components were provocative of symptoms; -ve = neither

component provoked symptoms; CI = confidence intervals; LR = likelihood ratios

Table 5: Analysis of time to onset of symptoms (in seconds), and correlation with

stenosis severity

Stenosis grade (N)*

Mean time to symptom onset when positive on component 1 for central stenosis

Mean time to symptom onset when positive on component 2 for central stenosis

Mean time to symptom onset when positive on component 1 for foraminal stenosis

Mean time to symptom onset when positive on component 2 for foraminal stenosis

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

13

Mild (1)

05.41 S.D. *N/A

16.69 S.D. *N/A

18.24 S.D. *N/A

28.03 S.D. *N/A

Moderate (11)

19.59 S.D. 15.29

28.25 S.D. 21.24

16.14 SD 18.72

15.89 SD 10.50

Severe (13)

05.07 SD 3.31

13.78 SD 13.32

14.63 SD 13.03

18.13 SD 25.51

* five patients had a mixed picture moderate in the spinal canal, but severe in the lateral canal or vice versa

Table 6: Relationship between site of stenosis and MExt test (in participant

numbers)

Central stenosis

Foraminal stenosis

Central and foraminal stenosis

No stenosis

MExT extension positive / lateral flexion negative

1 1

MExt extension negative / lateral flexion positive

1 1 1 1

MExT extension / lateral flexion positive

6 8 4 2

MExT extension / lateral flexion negative

1 1 2

4. DISCUSSION

To the authors' knowledge, the current investigation is the first to propose and

evaluate the validity of the MExT, a bi-component movement test, postulated to

support the diagnosis of ALSS through physical examination of individuals who

were 50 years or older, and had unilateral or bilateral symptoms in the lower

extremity radiating distal to the gluteal fold. The current evaluation observed the

MExT to demonstrate high sensitivity (92%) with a clinically significant negative

likelihood ratio (-LR 0.2). In accordance with negative post-test probabilities, the

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

14

likelihood of a subject within our cohort having ALSS reduced from 83 % to 49 %

when a negative result was observed. Such findings would support the

prospective validity of the MExT to assist in ruling-out a diagnosis when a negative

result is observed. In relation to the test’s proficiency to assist in ruling-in a

diagnosis, the converse appeared true. Specificity point estimates were low (40%),

and positive likelihood ratios were only small, (+LR 1.53). Neither value achieved

the desired level for clinical acceptability, (>80% or > 2.0 respectively) with only a

five per cent increase in subject post-test probability when a positive result was

observed.

Analysis of whether a relationship existed between stenosis grade and symptom

onset following MExT commencement failed to demonstrate a significant

correlation for either central or foraminal elements. Correlations were statistically

non-significant irrespective of whether stenotic types were stratified by the

theoretical preferential component or calculated in relation to the component of

quickest onset of symptoms. Analysis of alternate MExT components in relation to

stenotic location additionally failed to identify a clear association. Whilst ‘Kemps

sign’ has been endorsed as an important clinical observation for foraminal

stenosis (Lyle et al., 2005; Watanabe et al., 2010), findings from this current study

would alternatively promote the evaluation of symptomatic response to both

extension alone and extension with side-flexion, irrespective of the stenotic

location suspected. As such, inclusion of both movement components would

allow for greater test sensitivity than any one movement alone.

As identified in an updated systematic review of the accuracy of diagnostic tests

(de Schepper et al. 2013) items from the history had the highest diagnostic value:

radiating leg pain worse on standing, absence of pain when sitting, easing of pain

when bending, and a wide-based gait. MRI was the most promising imaging test;

and about 50 clinical tests were also examined. These included abnormal gait,

weakness, reflexes, a combination of signs and symptoms, a questionnaire, pain

drawings, and walking on a treadmill. On the whole clinical tests had either good

levels of sensitivity or good levels of specificity, but not both; similar to the findings

in this study.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

15

Whilst the current cohort was consistent with other studies of this nature in regards

to age and gender (Katz et al., 1995; Lyle et al., 2005; Konno et al., 2007);

appraisal of post-test validity against previous extension tests varied. Katz et al.

(1995) reported a small positive likelihood ratio (+LR 1.6) on ‘thigh pain with thirty

seconds extension’; however, point estimates were notably different to the current

study with reduced sensitivity (51%), and greater specificity (69%). The increase in

test sensitivity may have related to longer test duration. Subjects in the MExT

maintained the extended posture for up to thirty seconds longer than previous

studies. Whilst test modifications may assist identification of ALSS with greater

frequency than traditional extensions tests, such alterations may additionally

account for the noted drop in specificity.

In relation to alternative clinical diagnostic tools the MExT observed a comparative

capability to rule out ALSS diagnosis. Sensitivity estimates and negative

likelihood ratios were similar to those of the prediction-rule grounded, ‘ALSS

diagnostic support tool’ (Konno et al., 2007). However, in relation to overall

properties the MExT appeared subordinate, failing to achieve the same level for

specificity (72 %) or positive likelihood ratios (+LR 3.31). Likewise greater

discriminative capability to rule-in a diagnosis was observed during a two-stage

treadmill test with all positive likelihood ratios exceeding those of the MExT

(+LR >2.5) (Fritz et al., 1997).

4.1 Limitations

The findings of this study should be considered in relation to a number of study

limitations. Through recruitment of only a small sample size (n = 30), the

evaluation had an increased likelihood of type II error. The small sample size and

lack of statistical power may have had the effect of producing non-significant

results. Confidence intervals around the primary data appeared wide, and hence,

estimations of test validity must be interpreted with caution. Prevalence of ALSS

within the cohort was high as a result of the inclusion criterion employed.

Specificity estimates may have suffered due to inadequate evaluation against

subjects with an alternative lumbar, neurogenic or vascular pathology.

Recruitment methods through secondary care may have inadvertently biased the

sample towards more advanced pathology. A large proportion of the cohort was

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

16

observed to exhibit moderate to severe ALSS. As such, it was not possible to fully

evaluate the MExT against more mild variants limiting its external validity to

primary care settings. Furthermore in such settings there would be a higher

degree of diagnostic uncertainty, but in such a context tests could have the

greatest ability to influence clinical decision making if they perform well. Whilst

MRI was selected as the reference standard due to its objectivity, recognised

discriminative capability and consistent clinical application, it must be

acknowledged that inconsistency between clinical symptoms and degree of

radiographic change exists. Another limitation was the lack of assessment of the

reliability between testers on decision making regarding whether there was a

positive or negative response to the tests.

4.2 Clinical and research implications

In relation to clinical practice it is not currently possible to support or refute the

MExT’s diagnostic validity due to the limitations discussed. Whilst preliminary

estimates suggest the MExT may demonstrate part criterion validity, further

evaluation is required. As such, the current evaluation should act as a feasibility

study upon which to base further investigation into extension test validity. Future

studies should aim to reproduce test procedures within larger cohorts across a

range of primary and secondary care settings. Replication of current findings

would support the value of the MExT in ruling out a diagnosis of ALSS, and hence,

would form a valuable part of the physical examination process.

Until such validation has been completed, clinician’s should continue to evaluate

all available subjective and objective findings in order to support or refute a

potential diagnosis. Moreover, in clinical practice clinician’s should continue to

use subjective report of radicular symptoms and the impact upon the individual’s

functionality as the pre-requisite for onward referral.

5. CONCLUSION

Preliminary examination of a modified extension test’s validity to diagnose lumbar

spinal stenosis demonstrated conflicting results. Whilst the test demonstrated a

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

17

clinically significant capability to rule-out diagnosis, concurrent capability to rule-in

diagnosis was unacceptable. All aspects of the test were not capable of

determining the severity of stenosis or of differentiating between central and

foraminal types. Supplementary research on larger cohorts is required to fully

evaluate test validity.

Word count 4,231 including figures

References

Aalto T, Malmivaara, A, Kovacs F, Herno A, Alen M, Salmi L, et al. Preoperative predictors for postoperative clinical outcome in lumbar spinal stenosis: systematic review. Spine 2006;31(18):E648-E663. Akobeng A. Understanding diagnostic tests1: sensitivity, specificity and predictive values. Acta Paediatrica 2007;96(3):338-341. Amundsen, T., Weber, H., Nordal, H., Magnaes, B., Abdelnoor, M., Lilleas, F. Lumbar spinal stenosis: conservative or surgical management: A prospective 10-year study. Spine 2000;25 (11):1424-1436. Bal S, Celiker R, Palaoglu S, Cila A. F wave studies of neurogenic intermittent claudication in lumbar spinal stenosis. Am J Phys Med Rehab 2006;85 (2):35–40. Bassewitz H, Herkowitz H. Lumbar stenosis with spondylolithesis: current concepts of surgical treatment. Clin Orth Rel Res 2001;384:54-60. Bland M. An Introduction to Medical Statistics (3rd edition). Oxford: Oxford University Press, 2000. Browne J, Roberts L. Extended scope practitioners management of lumbar spinal stenosis. J Bone Joint Surg 90(B);Supp III:488. Bussieres A, Taylor J, Peterson C. Diagnostic imaging practice guidelines for musculoskeletal complaints in adults-an evidence-based approach-part 3: spinal disorders. J Manip Physio Thera 2008;31(1):33-38. Campo M, Shiyko M, Lichtman S. Sensitivity and specificity: A review of related statistics and controversies in the context of physical therapist education. J Phys Thera Educ 2008;24(3):69-89. Deen H, Zimmerman R, Lyons M, McPhee M, Verheijde J, Lemens S. Test-retest reproducibility of the exercise treadmill examination in lumbar spinal stenosis. 2000;75 (10):1002-1007. De Graaf, Prak I, Bierma-Zeinstra S, Thomas S, Peul B, Koes B. Diagnosis of lumbar spinal stenosis: A systematic review of the accuracy of diagnostic tests. 2006;31(10):1168-1176.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

18

De Schepper EIT, Overdevest GM, Suri P, Peul WC, Oei EHG, Koes BW, Bierma-Zeinstra SMA, Luijsterburg PAJ. Diagnosis of lumbar spinal stenosis. An updated systematic review of the accuracy of diagnostic tests. Spine 2013;38:E469-E481. Deyo R, Mirza S, Turner S. Overtreating chronic back pain: time to back off?’ J Board Fam Med 2009;22:62-68. Dong G, Porter RW. Walking and cycling tests in neurogenic and intermittent claudication. Spine 1989;14(9):965–969. Eguchi Y, Ohtori S, Yamashita M, Yamauchi K, Susuki M, Orita S, et al. Diagnosis of lumbar foraminal stenosis. Chiba Med J 2010;86 (2):43. Fredrickson M, Lee S, Welsh J. Changes in posterior disc bulging and intervertebral foraminal size associated with flexion-extension movement: a comparison between L4-5 and L5-S1 levels. Spine 2001;26(1):10-17. Fritz J, Erhard R, Delitto A, Welch W, Nowakowski PO. Preliminary results of the use of a two-stage treadmill test as a clinical diagnostic tool in the differential diagnosis of lumbar spinal stenosis. J Spinal Dis 1997;10:410-416. Fritz J, Delitto A, Welch W, Erhard E. Lumbar spinal stenosis: A review current concepts in evaluation, management, and outcome measurements. Arch Phys Med Rehab 1998;79:700-708. Ganz J. Lumbar spinal stenosis: postoperative results in terms of preoperative posture-related pain. J Neurosurg 1990;72 (1):71-74. Genevay, S., Atlas, S. Lumbar spinal stenosis. Best Pract Res Clin Rheum 2010;24:253-265. Genevay S, Atlas S, Katz J. Variation in eligibility criteria from studies of radiculopathy due to a herniated disc and of neurogenic claudication due to lumbar spinal stenosis. Spine 2010;35 (7):803-811. Goh K, Khalifa W, Anslow P, Cadoux T, Donaghy M. The clinical syndrome associated with lumbar spinal stenosis. Euro Neurol 2004;52(4):242–249. Haig A, Tong H, Yamakawa K, Parres C, Quint D, Chiodo A et al. Predictors of pain and function in persons with spinal stenosis, low back pain, and no back pain. Spine 2006;31(25):2950–2957. Haig A, Tomkins C. Diagnosis and management of lumbar spinal stenosis. J Am Med Assoc 2010;303(1):71-72. IAGG. International Association of Gerontology and Geriatrics. Research agenda on ageing for the 21st century. 2007. Available at http://www.iagg.com.br/webforms/iaggNewsletter.aspx Inufusa A, An H, Lim T, Hasegawa T, Haughton V, Nowicki B. Anatomic changes of the spinal canal and intervertebral foramen associated with flexion-extension movement. Spine 1996;21(21):2412–2420. Jenis L, Howard S. Spine update: Lumbar foraminal stenosis. Spine 2000;25(3):389-394. Johnson O. Information theory and central limit theorem. Imperial College Press, 2004. Katz J, Dalgas M, Stucki G. Degenerative lumbar spinal stenosis. Diagnostic value of the history and physical examination. Arth Rheum 1995;38:1236-1241.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

19

Keller T, Szpalski M, Gunzburg R, Spratt K. Assessment of trunk function in single and multi-level spinal stenosis: a prospective clinical trial. Clin Biomech 2003;18:173-180. Kim H, Chun H, Han C, Moon S, Kang K, Kim H, et al. The risk assessment of a fall in patients with lumbar spinal stenosis. Spine 2011;36(9):E588-E592. Kinnear P, Gray C. SPSS 15 Made Simple. New York Psychology Press, 2008. Kishner S, Gunduz O, Munshi S, Gupta S, Goyeneche N. Electrodiagnosis in lumbar spinal stenosis: A review. Turkish J Phys Med Rehab 2010;56:75-80. Kranzler J. Statistics for the Terrified (4th edition). Pearson Education, London, 1997. Konno S, Hayashino Y, Fukuhara S, Kikuchi S, Kaneda K, Chiba K, et al. Development of a clinical diagnosis support tool to identify patients with lumbar spinal stenosis. Eur Spine J 1997;16(11):1951-1957. Lurie D. What diagnostic tests are useful for low back pain? Best Pract Res Clin Rheum 2005;19(4):557-575. Lurie J, Tosteson A, Tosteson T, Carragee E, Carrino J, Kaiser J,et al. Reliability of readings of magnetic resonance imaging features of lumbar spinal stenosis. Spine 2008;33(14):1605-1610. Lyle M, Manes S, Mcguiness M, Ziaei S, Iversen M. Relationship of physical exam findings and self-reported symptom severity & physical function in patients with degenerative lumbar conditions. Phys Ther 2005;85(2):120-133. Malmivaara A, Slatis P, Heliavaara M, Sainio P, Hirvonen N, Kankare J. Surgical or non-operative treatment for lumbar spinal stenosis? A randomized controlled trial. Spine 2007;32(1):1-8. Maloney-Backstrom K, Whitman J, Flynn T. Lumbar spinal stenosis – diagnosis and management of the aging spine. Man Ther 2011;16:308-317. Mannion A, Denzler R, Dvorak J, Grob D. Five-year outcome of surgical decompression of the lumbar spine without fusion. Eur Spine J 2010;19(11):1883-1891. McGee S. Simplifying likelihood ratios. J Gen Int Med 2010;17(8):647-650. Middleton K, Fish D. Lumbar spondylosis: clinical presentation and treatment approaches. Curr Rev Musc Med 2009;2(2):94-104. Moon E, Kim H, Park J, Shin D, Ha J, Shim D, et al. Comparisons of the predictive value of myelography, Computer tomography and MRI on the treadmill test in lumbar spinal stenosis. Yonsei Med J 2005;46:806-811. Ogikubo O, Forsberg L, Hansson T. The relationship between the cross-sectional area of the cauda equine and the preoperative symptoms in central lumbar spinal stenosis. Spine 2007;32(13):1423-1428. Papadakis M, Sapkas G, Papadopoulos E, Katonis P. Pathophysiology and biomechanics of the aging spine. Open Ortho J 2011;5:335-342. Panjabi M, Takata K, Goel V. Kinematics of the lumbar intervertebral foramen. Spine 1983;8:348-357. Petty NJ. Neuromusculoskeletal Examination and Assessment. A Handbook for Therapists (3rd Edition). Elsevier, Churchill Livingstone, Edinburgh, 2006.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

20

Pewsner D, Battaglia M, Minder C, Marx A, Bucher H, Egger M. Ruling a diagnosis in or out with SpPin and SnNout a note of caution. Br Med J 2004;329:209-213. Population Reference Bureau http://www.pssrb.org/publications/graphicsbank/PopulationTrends.aspx;2010. Porter R. Spinal stenosis and neurogenic claudication. Spine 1996;21(17):2046–2052. Riegelman R. Studying a study and testing a test: how to read the medical evidence. Lippincot-Williams and Wilkins, 2005 (5th Edition). Rigby A, Summerton N. Statistical methods in epidemiology. VIII. On the use of likelihood ratios for diagnostic testing with an application to general practice. Dis Rehab 2005;27(9):475-480. Schronstrom N, Willen J. Imaging lumbar spinal stenosis. Radiolog Clin Nth Am 2011;39:31-53. Slatis P, Malmivaara A, Heliovaara M, Saino P, Seitsalo S, Hurri H. Surgical or nonoperative treatment for lumbar spinal stenosis? A six-year follow-up of a randomized controlled trial. J Bone Joint Surg 2007;91B: Supp II, 290-281. Streiner D, Norman G. Health Measurement Scales (3rd edition). Oxford University Press, Oxford, 2003. Steurer J, Nydegger A, Held U, Brunner F, Hodler J, Porchet F et al. Lumbar spinal stenosis outcome study.’ BMC Musculoskeletal disorders. 2010;11:254. Sugioka T, Hayashino Y, Konno S, Kikuchi S, Fukuhara S. Predictive value of self-reported patient information for the identification of lumbar spinal stenosis. Family Pract 2008;25(4):237-244. Tenhula J, Lenke L, Bridwell K, Gupta P, Riew, D. Prospective functional evaluation of the surgical treatment of neurogenic claudication in patients with lumbar spinal stenosis. J Spine Dis 2000;13(4):276-282. Vo A, Kamen L, Shih V, Bitar A, Stitik T, Kaplan R. Rehabilitation of orthopaedic and rheumatologic disorders, 5. Lumbar spinal stenosis. Arch Phys Med Rehab 2005;86(3): Supp 1:69-76. Wang Y, Jeng C, Wu C, Chang H, Huang Y, Wang Y, et al. Dynamic effects of axial loading on the lumbar spine during MRI in patients with suspected spinal stenosis. J Formosan Med Assoc 2008;107(4):334-339. Watanabe K, Yamazaki A, Morita O, Sano A, Katsumi K, Ohashi M. Clinical outcomes of posterior lumbar interbody fusion for lumbar foraminal stenosis: preoperative diagnosis and surgical strategy. Indian J Orth 2014;48:107-10. Watters W, Baisden J, Gilbert T, Kreiner S, Resnik D, Bono C, et al. Degenerative lumbar spinal stenosis: an evidence-based clinical guideline for the diagnosis and treatment of degenerative lumbar spinal stenosis. Spine J 2008;8:305-310. Westergaard L, Hauerberg J, Springborg J. Outcome after surgical treatment for lumbar spinal stenosis: the lumbar extension test is not a predictive factor. Spine 2009;34(25):E930-E935.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

21

Whiting P, Rutjes A, Dinnes J, Reitsma J, Bossuyt P, Kleijnen, J. Development and validation of methods for assessing the quality of diagnostic accuracy studies. Health Technology Assessment 2004;8:25. Yukawa Y, Lenke L, Tenhula J, Bridwell K, Riew D, Blanke K. A comprehensive study of patients with surgically treated lumbar spinal stenosis with neurogenic claudication. J Bone Joint Surg 2002;84(A):1954-1959.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

Acknowledgements

The authors would like to thank Pieter Meiring and other staff at Chesterfield Royal

Hospital for facilitating the study, and the first author’s family for on-going

encouragement throughout the process.

Conflict of interest

The authors declare no conflict of interest.

Funding

No funding was received for this study.

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

Figure 1. STARD Flow Diagram

Excluded patients Reasons n= 1 - declined MRI

MeXT test n= 30

Positive MeXT n= 26

Negative MeXT n= 4

Inconclusive result n= 0

No MRI n= 0

No MRI n= 0

No MRI n= 0

MRI n= 4

MRI n= 26

MRI n= 0

Inconclusive n= 0

Inconclusive n= 0

Inconclusive n= 0

ALSS Present n= 23

ALSS absent n= 3

ALSS Present

n= 2

ALSS absent n= 2

ALSS Present

n= 0

ALSS absent n= 0

Eligible patients n= 31

Potentially eligible patients n= 276

Exclusion reason (N = 245):

Upper quadrant symptoms (109) ≤ 50 years (60)

No referred symptoms (76) No informed consent (0)

Peripheral vascular disease (0) Ankylosing Spondylitis (0)

MAN

USCR

IPT

ACCE

PTED

ACCEPTED MANUSCRIPT

Highlights

• Spinal stenosis is likely to be more commonly encountered in clinical practice. • The validity of a simple clinical test was compared to diagnosis by MRI. • The test had a reasonable sensitivity of 92%, but a poor specificity of 40%. • The test may be useful to rule out a diagnosis with a negative test result.