The Effectiveness Of Walking Stick Use For Neurogenic.pdf

15

ORIGINAL ARTICLE

The Effectiveness of Walking Stick Use for Neurogenic Claudication: Results From a Randomized Trial and the Effects on Walking Tolerance and Posture Christine M. Comer, MSc, Mark I. Johnson, PhD, BSc, PGCHE, Paul R. Marchant, PhD, CStat, BSc, MSc, FRAS, Anthony C. Redmond, PhD, MSc, Howard A. Bird, MD, FRCP, Philip G. Conaghan, MBBS, PhD, FRACP, FRCP ABSTRACT. Comer CM, Johnson MI, Marchant PR, Redmond AC, Bird HA, Conaghan PG. The effectiveness of walking stick use for neurogenic claudication: results from a randomized trial and the effects on walking tolerance and posture. Arch Phys Med Rehabil 2010;91:15-9. Objectives: To determine the immediate effects of using a stick on walking tolerance and on the potential explanatory variable of posture, and to provide a preliminary evaluation of the effects of daily walking stick use on symptoms and function for people with neurogenic claudication. Design: A 2-phase study of neurogenic claudication patients comprising a randomized trial of 2 weeks of home use of a walking stick and a crossover study comparing walking tolerance and posture with and without a walking stick. Setting: A primary care– based musculoskeletal service. Participants: Patients aged 50 years or older with neurogenic claudication symptoms (N⫽46; 24 women, 22 men, mean age⫽71.26y) were recruited. Intervention: Walking stick. Main Outcome Measures: Phase 1 of the trial used the Zurich Claudication Questionnaire symptom severity and physical function scores to measure outcome. The total walking distance during a shuttle walking test and the mean lumbar spinal posture (measured by using electronic goniometry) were used as the primary outcome measurements in the second phase. Results: Forty of the participants completed phase 1 of the trial, and 40 completed phase 2. No significant differences in symptom severity or physical function were shown in score improvements for walking stick users (stick user scores ⫺ control scores) in the 2-week trial (95% confidence interval [CI], ⫺.24 to .28 and ⫺.10 to .26, respectively). In the second phase of the trial, the ratio of the shuttle walking distance with a stick to without a stick showed no significance (95% CI,

From the Academic Unit of Musculoskeletal Disease, University of Leeds (Comer, Redmond, Bird, Conaghan); Faculty of Health, Leeds Metropolitan University (Johnson), and Faculty of Information and Technology (Marchant), Leeds Metropolitan University, NIHR Leeds Musculoskeletal Biomedical Research Unit and Section of Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds (Conaghan), Leeds, UK. Supported by a project grant from Leeds Primary Care Trusts Research Consortium. Comer is funded in part by the Arthritis Research Campaign UK (grant no. 18183). Trial registration is ISRCTN registered (ISRCTN35836727) and NRR registered (N0436193958). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Correspondence to Philip G. Conaghan, MBBS, PhD, FRACP, FRCP, Section of Musculoskeletal Disease, Chapel Allerton Hospital, Chapeltown Road, Leeds LS7 4SA, United Kingdom, e-mail: [email protected]. Reprints are not available from the author. 0003-9993/10/9101-00452$36.00/0 doi:10.1016/j.apmr.2009.08.149

.959 –1.096) between the groups. Furthermore, the use of a walking stick did not systematically promote spinal flexion; no significant difference was shown for mean lumbar spinal flexion for stick use versus no stick (95% CI, .351°–.836°). Conclusions: The prescription of a walking stick does not improve walking tolerance or systematically alter the postural mechanisms associated with symptoms in neurogenic claudication. Key Words: Rehabilitation; Spinal stenosis; Walking stick. © 2010 by the American Congress of Rehabilitation Medicine EUROGENIC CLAUDICATION is described as the clasN sic clinical syndrome associated with lumbar spinal stenosis and is characterized by symptoms of pain, numbness, weakness, or paresthesia in the legs exacerbated by walking and relieved by stooping forward or sitting.1,2 Surgical procedures for NC are complex and costly and may be associated with limited improvement in symptoms or function.3,4 Consequently, conservative (nonsurgical) treatment is normally advocated as first-line management for patients with NC. To date, there has been little high-quality research investigating the effectiveness of conservative treatments. Many conservative treatment programs attempt to capitalize on theoretic benefits of postural modification. The “shopping cart sign” is commonly reported in the literature, referring to the phenomenon that patients with NC obtain symptomatic relief from walking with a shopping trolley for support,5,6 which allows them to lean forward slightly. Because of this phenomenon, a recommendation to use a walking stick is sometimes incorporated into the conservative management of these patients. This study had 2 aims: (1) to determine the immediate effects of using a walking stick on walking tolerance and posture in order to understand potential mechanisms of action of the intervention and (2) to provide a preliminary evaluation of the effectiveness of the daily use of a walking stick in improving symptoms and function. METHODS To address the 2 aims, the study comprised 2 discrete phases: phase 1 consisted of a pragmatic randomized trial of 2 weeks of home use of a walking stick in which 1 group was prescribed a walking stick and the control group was not, and phase 2 consisted of a cross-sectional study comparing walking tolerance and posture with and without a walking stick during a shuttle walk test using a crossover design (fig 1). The study

List of Abbreviations CI NC

confidence interval neurogenic claudication

Arch Phys Med Rehabil Vol 91, January 2010

16

WALKING STICKS FOR NEUROGENIC CLAUDICATION, Comer

Fig 1. Trial design and timetable.

was approved by Harrogate Local Research Ethics Committee and was given Research Governance approval by The Leeds Teaching Hospitals National Health Service Trust and West Yorkshire Primary Care Research and Development Unit. The sample size for the study was powered for the primary outcome, which was the change in shuttle walking test distance with a walking stick in phase 2 of the study. It was designed to detect what was considered to be a clinically meaningful improvement of 20% in population mean walking distance when using a stick based on a mean walking distance of 150m in similar patients from a previous study.7 A standard deviation of 38m from the previous study was assumed,7 and a significance level of 5% and power of 95% were used. Calculation (after Kirkwood and Sterne8 for a single mean) gave a sample size of 21 participants for each of the 2 subgroups arising from the order of stick use in the shuttle walking tests. To account for attrition, recruitment was set at 24 participants per group. Potential study participants with clinical symptoms of NC were identified within the Leeds primary care– based Musculoskeletal Service. All patients gave informed consent before entering the trial. Inclusion for both phases of the study was based on a clinician diagnosis of NC combined with a limitation in walking tolerance because of NC (table 1). NC symptoms were defined as lower-limb pain, cramping, heaviness, or paresthesia brought on by walking and relieved by sitting or stooping forward. Radiologic confirmation of lumbar spinal stenosis was not an inclusion criterion because NC represents

a clinical syndrome recognized and treated by physiotherapists in the primary care setting. Full inclusion criteria for phases 1 and 2 of our study are given in table 1. Phase 1 Clinical Protocol: 2-Week Home Trial of Stick Use Participants who were not using a walking stick routinely were randomized into 1 of 2 groups for a 2-week trial: a group of participants who were given a walking stick for daily use or a control group of participants who were not given a walking stick. Randomization was performed by using a computergenerated randomization scheme prepared in 5 blocks of 8 participants. A sealed envelope system was used for blinding the trial investigator to treatment allocation until the point of treatment delivery. Prescribing a walking stick involved issuing a standard National Health Service wooden walking stick with a curved handle adjusted to the height of the participant’s wrist joint with the arm held comfortably at the side. This protocol adhered to the normal recommended prescription of walking sticks in standard orthopedic and physiotherapy practice.9 Participants with predominantly unilateral symptoms were advised to hold the stick in the hand opposite to the symptomatic side, whereas those with bilateral symptoms were advised to use the stick in whichever hand felt most comfortable. Each participant was required to show safe use of the walking stick while in the clinic before being instructed to use the stick when walking during all day-to-day activities. The

Table 1: Inclusion and Exclusion Criteria for Trial Phases 1 and 2 Inclusion Criteria

Exclusion Criteria

Age 55 years or greater Unilateral or bilateral neurogenic claudication symptoms (ie, exercise induced leg pain on walking, relieved in sitting or flexion) Patient-reported limitation in walking tolerance because of NC symptoms

Cognitive impairment preventing full understanding or participation in study Medical conditions limiting walking (lower-limb joint pathology, neurologic, cardiovascular, or respiratory conditions)

*Phase 1 only.

Arch Phys Med Rehabil Vol 91, January 2010

Signs or symptoms of acute cauda equina syndrome or severe or worsening neurologic status requiring medical or surgical assessment. (This includes significant or worsening nerve root or cauda equina function, significant or sinister weight loss, pyrexia, unremitting pain, and significant inflammatory joint disease.) Current use of a walking aid*

17

WALKING STICKS FOR NEUROGENIC CLAUDICATION, Comer

following data were collected at both baseline and at 2 weeks: (1) Zurich Claudication Questionnaire (primary outcome), (2) Oswestry Disability Index, (3) pain history including visual analog scales for leg pain and for back pain, and (4) Hospital Anxiety and Depression Scale. The primary hypothesis was that changes in symptom severity scores of the Zurich Claudication Questionnaire scale were equal in both groups. A descriptive analysis of data was performed initially to explore changes over the 2-week intervention period for both groups. The hypothesis was tested by comparison of the mean change by using an independent samples t test. Phase 2 Clinical Protocol: Walking Tolerance Test On completion of the 2-week home trial period, all participants were invited to participate in phase 2 of the trial. Subjects who were unwilling to participate in phase 1 were given the option of being recruited directly to phase 2. Phase 2 involved undertaking 2 shuttle walking tolerance tests; 1 test was performed by using a walking stick, and the other test was performed without a stick. The order of the tests was randomized, and allocation was concealed by using presealed envelopes as in phase 1. The shuttle walking test is a standardized test10 that involved the participant walking repeated laps of a 10-m distance on a flat surface. The start of each new lap of the shuttle walk test was indicated by prerecorded beeps, with the time between beeps becoming incrementally shorter during the test. The test ended when the participant could no longer complete a shuttle before the next beep or when the participant needed to stop because of pain or other symptoms. During the shuttle walking test, the sagittal plane posture of the lumbar spine (flexion/extension movements) was recorded by using an electronic strain gauge goniometer.a The electrogoniometer consists of 2 end blocks (distal block 50mm length, proximal block 120mm length) connected by a biaxial strain gauge that provides flexion data in 2 planes. In line with the manufacturer’s operating instructions, the distal end block was fixed with tape over the level of S1, with the proximal end block fixed at the level of T12 to L1 so that the strain gauge connecting the 2 probes is at near minimal length. The dataacquisition unit was calibrated to 0 with participants in a fully upright stance, with the heels, sacrum, and shoulders flat against a vertical wall. Spinal posture was recorded at 5Hz and logged in real time on the data-acquisition unit before being downloaded and analyzed by using Biometrics DataLOG soft-

ware version 3.00.a This provided an individual temporal mean measurement of lumbar spinal flexion (in degrees) from an upright stance during each shuttle walking test with and without the use of a stick. The outcome for phase 2 was the difference in distance in meters walked with or without a walking stick. Evaluation of the shuttle walking test data involved initial descriptive and graphic analysis to explore differences in the mean walking distance and in posture between the 2 test conditions. A general linear model respecting the crossover design was used to test the hypothesis that there was no difference in the distance walked and no difference in spinal posture with or without a walking stick. RESULTS Forty-six patients (24 women, 22 men; mean age, 71.26y [range, 58 – 89y]) entered the study over a 15-month period. Of the 46 participants overall, 34 completed both phases of the study. Six participants undertook phase 1 of the study but were not able or prepared to take part in phase 2 of the study because of transport and time constraints (n⫽5) or illness (n⫽1). Six participants who did not undertake phase 1 (current stick users) were recruited directly to phase 2 of the trial. Therefore, a total of 40 participants were recruited in each phase of the study. There was no loss to follow-up in either phase once participants had been recruited (fig 2). Phase 1 Results: 2-Week Home Trial of Stick Use Baseline characteristics for the 40 participants in the 2-week home trial are presented in table 2. There were no dropouts or missing data, and all participants stayed within the allocated treatment arm. Therefore, analysis was performed on an intention-to-treat basis. Both groups showed small improvements in the symptom severity score of the Zurich Claudication Questionnaire. The mean difference in improvement between the 2 groups (stick group ⫺ no-stick group scores) was 0.02 points (95% CI, ⫺0.24 to 0.28), which was neither statistically nor clinically significant. The mean difference (stick group ⫺ no-stick group) in improvement in the Zurich Claudication Questionnaire physical function scale scores of 0.08 points (95% CI, ⫺0.10 to 0.26) was neither clinically meaningful nor statistically significant. Secondary outcome measures for pain severity, function, and psychologic factors also showed changes in scores that did not reach statistical significance (table 3).

Patients Recruited to Trial (n=46)

Walking stick users (n=6)

Non - walking stick users (n=40) Randomized

Two weeks continue stick use (n=6)

Randomized

Shuttle Walk Test: Stick then no stick (n=3)

Shuttle Walk Test: No stick then stick (n=3)

Two weeks stick use (n=20)

Two weeks stick use (n=20)

Randomized

Shuttle Walk Test: Shuttle Walk Stick then no then stick Test: Stick (n=7) no stick

Unwilling / unable to undertake Shuttle Walk Test (n=6)

Shuttle Walk Test: No stick then stick (n=10)

Randomized

Shuttle Walk Test: Stick then no stick (n=11)

Shuttle Walk Test: No stick then stick (n=6)

Fig 2. Flow of participants to trial.

Arch Phys Med Rehabil Vol 91, January 2010

18

WALKING STICKS FOR NEUROGENIC CLAUDICATION, Comer Table 2: Baseline Characteristics for Phase 1 Participants Group

Group Prescribed Walking Stick

Control Group

Age (y) VAS leg pain VAS low back pain ZCQ symptom severity ZCQ physical function Oswestry Disability Index HAD depression HAD anxiety

71.0⫾6.7 6.90⫾1.95 5.49⫾2.93 3.08⫾0.61 2.48⫾0.61 39.20⫾11.70 6.90⫾3.49 6.45⫾3.68

70.8⫾8.3 6.72⫾2.57 6.91⫾2.55 3.25⫾0.76 2.44⫾0.45 44.55⫾16.36 7.84⫾3.30 7.89⫾4.58

NOTE. Values are mean ⫾ SD. Abbreviations: HAD, Hospital Anxiety and Depression scale; VAS, visual analog scale; ZCQ, Zurich Claudication Questionnaire.

Phase 2 Results: Walking Tolerance Test Baseline characteristics of the 40 participants who completed the shuttle walking crossover trial are presented in table 4. Inferential analysis was undertaken by using the standard statistical model.11 Data were analyzed by using a mixed model with a random effect for the participants as well as the fixed effects for the 2 factors: (1) use of a walking stick and (2) the order of the tests. Effect of a walking stick on walking distance. All 40 participants completed 1 shuttle walking test with a walking stick and 1 without a stick. With randomization of order and subsequent model fitting, the order of the tests was found to exert no significant systematic effect on the walking distance. The mean distance completed without a stick was 217.5⫾118.7m, and with a stick it was 215.5⫾108.1m. Before inferential analysis, walking distance values were log transformed to ensure that the residuals from the fitted model (eij) were normally distributed. The 95% CI for the increase in mean log distance with a stick was found to be ⫺0.044 to 0.094, which after back transformation via the exponential function yielded a 95% CI for the median ratio of distance walked with and without a walking stick (0.959 –1.096). Therefore, the ratio was seen to be about 1, and the upper confidence limit suggests at most a 10% increase of distance achieved when using a walking stick. These results indicate that there was no significant effect on the walking distance, either statistically or clinically meaningful, when using a walking stick. Effect of walking stick on lumbar spinal posture. Twentyeight of the participants undertaking the shuttle walking test had detailed electrogoniometric data available. Data for the remainder of the participants were either incomplete or missing because of recording faults. The mean lumbar flexion during the walking test without a walking stick was ⫺1.60°⫾8.63°, and with a stick it was ⫺1.28°⫾8.78°. The 95% CI for the mean increase in spinal flexion when using a stick was ⫺0.351 to 0.836, indicating no statistically significant difference. In

Table 4: Baseline Characteristics of Participants in Phase 2 Group

Median (5th and 95th Percentiles)

Age (y) Low back pain duration (mo) Leg symptom duration (mo) VAS leg pain VAS low back pain ZCQ symptom severity ZCQ physical function

70 (59 and 85) 42 (0 and 482) 24 (4 and 179) 6.7 (2.1 and 9.7) 6.3 (0 and 9.6) 3.07 (2 and 4.69) 2.50 (1.6 and 3.2)

Abbreviations: VAS, visual analog scale; ZCQ, Zurich Claudication Questionnaire.

this case, there was an effect of period (order of test) such that the 95% CI for the change in flexion in “period 2” versus “period 1” (⫺1.501 to ⫺0.314) showed a statistically significant decrease. Although the overall mean flexion was 0.32°⫾1.73° greater when using a stick compared with not using a stick, this difference was mostly in those using a stick in the first test in which there was an increase in flexion of 1.12°⫾1.26°. DISCUSSION There are few trials investigating the effectiveness of nonsurgical treatments for NC, and this is the first randomized trial of walking aid use in people with NC. The results of the first phase of this study showed that using a walking stick had no discernable effect on either symptoms or function over a 2-week period in this patient group. Equally, no discernable difference in walking tolerance was shown with or without a walking stick in the second phase of the trial. This contrasts with findings in a previous study12 evaluating the effects of a walking aid; in Goldman et al’s12 case series study of patients using a wheeled walker, good or excellent improvement (more than a 250% increase) in selfreported walking distance was observed in 71% (37/52) of patients with NC. However, the direct comparability with our data is limited by the use of a different type of walking aid and the use of self-reported walking distance as an outcome measure, which has been shown to have limitations compared with objective measures of walking tolerance.13 This second phase of the trial provides some insight into the possible reasons why using a walking stick did not improve symptoms or function. The mechanism expected to produce improvements was a hypothesized increase in spinal flexion when using a stick. This was not evident, however, and it may be that shorter walking sticks are required to produce measurable postural change. Goldman14 has noted previously that taller people with NC report good improvement in walking tolerance when leaning on a standard-height wheeled walker, but those of a shorter stature required a smaller wheeled walker to benefit. Furthermore, Goldman15 noted that some patients

Table 3: Pain, Dysfunction, and Psychologic Outcomes in Phase 1 Outcome Measure

Difference in Score Changes Between Groups Mean (95% CI)

Mean Change in Walking Stick Group Mean (95% CI)

Mean Change in Control Group Mean (95% CI)

VAS leg pain VAS low back pain ODI HAD depression HAD anxiety

0.52 (⫺0.52 to 1.56) 0.81 (⫺0.66 to 2.29) ⫺2.22 (⫺8.35 to 3.91 0.39 (⫺1.25 to 1.32) ⫺1.13 (⫺2.79 to 0.52)

⫺0.38 (⫺1.11 to 0.35) ⫺1.13 (⫺2.47 to 0.20) ⫺3.22 (⫺8.76 to 2.33) ⫺0.25 (⫺1.18 to 0.68) 0.45 (⫺0.89 to 1.79)

0.14 (⫺0.64 to 0.92) ⫺0.32 (⫺1.14 to 0.50) ⫺1.00 (⫺4.07 to 2.07) ⫺0.21 (⫺1.16 to 0.73) 0.68 (⫺1.74 to 0.37)

Abbreviations: HAD, Hospital Anxiety and Depression scale; ODI, Oswestry Disability Index; VAS, visual analog scale.

Arch Phys Med Rehabil Vol 91, January 2010

WALKING STICKS FOR NEUROGENIC CLAUDICATION, Comer

maximize spinal flexion postures by resting their forearms on the transverse handlebar when walking with a shopping trolley for support, but this is not possible with a wheeled walker with 2 separate handles or with a walking stick. This raises a further possibility that unloading of the spine may also play a role in the mechanism of symptom relief along with postural modification. A previous study16 showed that walking tolerance in people with NC was influenced more by loading or unloading of the spine with weights or harnesses than by adopting different spinal postures. It is possible that using a walking stick does not offer sufficient unloading to improve symptoms and function. Study Limitations The 2-phase design introduced the potential for the shuttle walking tolerance tests to be influenced by previous use or nonuse of a walking stick during the 2-week home trial phase. However, this potential confounder was explored explicitly in the analysis, and no systematic effects were identified. In addition, the crossover design of phase 2 carries an inherent risk that the response in the second shuttle walk test might be influenced by carryover effects from the first shuttle walk test. There was no systematic effect on walking distance from the order of the tests in this trial, but test order did affect changes in spinal posture with and without a walking stick. Therefore, a washout period of greater than the 30 minutes allowed in this trial would be advocated for future studies incorporating postural analysis during walking. CONCLUSIONS The trial results indicate that using a walking stick did not improve posture, symptoms, or function in our participants with NC. Further work to evaluate the effectiveness of a shorter than normal walking stick in promoting a flexed lumbar spinal posture and improving symptoms and function might be warranted. On the basis of our trial, the prescription of a walking stick cannot be routinely recommended as an intervention for all NC patients, although the benefits of a more substantial walking aid warrant further investigation. Acknowledgment: We thank Heikki Vanharanta, MD, DMSc, for his contribution to the study concept and design during the development stage of the trial. References 1. Porter RW. Spinal stenosis and neurogenic claudication. Spine 1996;21:2046-52. 2. Willen J, Danielson B, Gaulitz A, Niklason T, Schonstrom N, Hansson T. Dynamic effects on the lumbar spinal canal: axially

3. 4.

5.

6.

7.

8. 9. 10.

11. 12.

13.

14.

15.

16.

19

loaded CT-myelography and MRI in patients with sciatica and/or neurogenic claudication. Spine 1997;22:2968-76. Atlas SJ, Delitto A. Spinal stenosis: surgical versus nonsurgical treatment. Clin Orthop Relat Res 2006;Feb(443):198-207. Gibson JN, Waddell G. Surgery for degenerative lumbar spondylosis: updated Cochrane Review. Spine (Phila Pa 1976) 2005;30: 2312-20. Bulstrode C, Buckwalter J, Carr A, et al, editors. Oxford textbook of orthopaedics & trauma. 1st ed. Oxford: Oxford University Press; 2002. Evans JG, Williams TF, Beattie BL, Michel JP, Wilcock GK, editors. The Oxford textbook of geriatric medicine. 2nd ed. Oxford: Oxford University Pr; 2003. Pratt RK, Fairbank JC, Virr A. The reliability of the Shuttle Walking Test, the Swiss Spinal Stenosis Questionnaire, the Oxford Spinal Stenosis Score, and the Oswestry Disability Index in the assessment of patients with lumbar spinal stenosis. Spine (Phila Pa 1976) 2002;27:84-91. Kirkwood B, Sterne J, editors. Essential medical statistics. 2nd ed. Oxford: Blackwell Science; 2003. Watson MS, editor. Oxford handbook of palliative care. 1st ed. Oxford: Oxford University Press; 2005. Singh SJ, Morgan MD, Scott S, Walters D, Hardman AE. Development of a shuttle walking test of disability in patients with chronic airways obstruction. Thorax 1992;47:1019-24. Matthews JN, editor. An introduction to randomised controlled clinical trials. London: Arnold Publishers; 2000. Arnold; 2000. Goldman SM, Barice EJ, Schneider WR, Hennekens CH. Lumbar spinal stenosis: can positional therapy alleviate pain? J Fam Pract 2008;57:257-60. Zeifang F, Schiltenwolf M, Abel R, Moradi B. Gait analysis does not correlate with clinical and MR imaging parameters in patients with symptomatic lumbar spinal stenosis. BMC Musculoskelet Disord 2008;9:89. Goldman SM. Neurogenic positional pedal neuritis. Common pedal manifestations of spinal stenosis. J Am Podiatr Med Assoc 2003;93:174-84. Goldman S. Value of a grocery cart and walker in identification and management of symptomatic spinal stenosis in diabetic patients presenting with peripheral neuropathy or claudication [Letter]. Diabetes Care 2003;26: p 1943. Og˘uz H, Levendog˘lu F, Og˘ün TC, Tantu A. Loading is more effective than posture in lumbar spinal stenosis: a study with a treadmill equipment. Eur Spine J 2007;16:913-8.

Supplier a. Goniometer with DataLOG data acquisition unit P3X8, and DataLOG software ver 3.00; Biometrics Ltd, Unit 25, Nine Mile Point Industrial Estate, Cwmfelinfach, Gwent NP11 7HZ, UK.

Arch Phys Med Rehabil Vol 91, January 2010