HONK KONG PHYSIOTHERAPY JOURNAL

Research Paper Hong Kong Physiotherapy Journal Vol. 38, No. 2 (2018) 125–131 DOI: 10.1142/S1013702518500105 Hong Kong Physiother. J. 2018.38:125-131. Downloaded from www.worldscientific.com Hong Kong Physiotherapy Journal by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. http://www.worldscientific.com/worldscinet/hkpj Agreement of clinical examination for low back pain with facet joint origin Mantana Vongsirinavarat*, Wahyuddin Wahyuddin and Ratchaneewan Adisaiphaopan Faculty of Physical Therapy, Mahidol University, Nakornpathom 73170, Thailand *[email protected] Received 31 January 2017; Accepted 28 July 2017; Published 14 August 2018 Background: Low back pain (LBP) with facet joint origin is a common diagnosis of patients referred to physical therapy clinic. An expert consensus of diagnostic criteria has been proposed. However, the reliability of the assessment has not been proved. Objective: To test the degrees of agreement between two physical therapists for nine physical examination items and the diagnosis of facet joint origin. Methods: The examination according to diagnostic criteria was performed independently by two physical therapists in 45 patients with chronic LBP. The percent agreements and Kappa coe±cients of each exami- nation item and diagnostic conclusion were calculated. Results: The percent agreements of nine examined items ranged from 73.3–91.1%. The Kappa coe±cients, widely ranged from 0.250–0.690 (p ¼ 0:48 to < 0:001), showed statistically signi¯cant agreements for all examination items. The low level of agreements was partly due to improper distributions of test results. The agreement of conclusion was 86.7% and Kappa coe±cient was 0.492 (p ¼ 0:001) which re°ected good agreement of facet diagnosis. Conclusion: There were adequate agreements for clinical examination of LBP with facet joint origin. The low level of agreement suggested the clinicians to have operational de¯nition and rigorous training sessions although the examinations seemed to be routinely performed. Keywords: Low back pain; facet joint; assessment; reliability; agreement. Introduction physician and physical therapy visits.1 Several anatomical sites in low back were considered the Among musculoskeletal complaints, low back pain source of pain including facet joint, intervertebral (LBP) is the symptom most frequently leading to *Corresponding author. Copyright@2018, Hong Kong Physiotherapy Association. Published by World Scienti¯c Publishing Co Pte Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 125

126 M. Vongsirinavarat, W. Wahyuddin & R. Adisaiphaopan Hong Kong Physiother. J. 2018.38:125-131. Downloaded from www.worldscientific.com disc, ligaments, nerve root and dura, muscles, origin showed controversial results.18,19 However, by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. and fascia.2 A number of factors such as neuro- these criteria are valuable in clinical reasoning and logical, mental stress, and social status are also support the biomechanic evidences of pathology of reportedly contributing to LBP.3,4 the facet joints.17 Use of these criteria for evalu- ating the patients would result in more homoge- Facet or zygapophyseal joint was indicated to nous of subjects and the manual therapy be the cause in 42% of persons with LBP using speci¯cally on facet joints would be more e®ective. controlled comparative local anesthetic diagnostic However, the reliability of using these criteria in blocks for diagnostic con¯rmation,5 especially in clinic has not been reported. Therefore, this study the chronic cases.6 A community-based survey aimed to test the agreement of physical therapists in older adults showed evident facet degeneration in using the criteria to examine and diagnose on CT imaging associated with LBP.7 patients with facet joint problem causing LBP. A study reported that when physical therapists Methods use the combination of McKenzie lumbar spine assessment algorithms with a series of patho- This study was a single-group, repeated-measures anatomical diagnostic tests in patients with reliability study. The testing was conducted in a chronic LBP, the most frequent diagnosis (49%) university physical therapy center. The study was facet joint.8 To diagnose the facet joint pain in protocol was approved by the Ethic Committee of clinic, the literatures suggested the uses of symp- Mahidol University (MU-CIRB); protocol No. toms,9 symptom changes with movement and ac- MU-IRB, COA. No. 2014. 033.2103, and Protocol tivities,10 local symptoms without nerve root No. MU-IRB 2014/006.0901. The data collection pain,11,12 and unilateral pain without referred pain was undertaken from January to December 2015. lower than knee, and no symptom along derma- The examination according to the criteria reported tome or myotome.13,14 The movements activating in the study of Wilde17 included the interview of pain by increasing pressure on the joint and symptom behaviors (unilateral local pain, referred stretching the capsule are extension, rotation, and pain not beyond knee, and no nerve root pain); lateral °exion.10,11 In rotation and extension, pain response in movement tests (less pain in the pain mechanics is supposed as the inferior °exion, pain in extension, and pain in extension articular process slips on superior articular with lateral °exion and rotation to the same side process which could activate nociceptors in the of facet); and manual tests (activate pain with capsule.15,16 unilateral pressure on the joint or transverse pro- cess, pressure on the joint found decreased range or A study in 2007 presented the consensus of increased resistance on the painful side, and muscle experts about the clinical features of facet joint spasm same side to the facet). pain.17 The three round Delphi survey resulted in 12 indicators relevant to pathoanatomical mecha- The participants were consecutive patients aged nism of pain. The criteria included \\(1) unilateral 18–60 years old with LBP longer than three local pain, (2) activate pain with unilateral pres- months. On the day of examination, the patient sure on the joint or transverse process, (3) no nerve had pain measured by VAS 21–79 from 100 mm. root pain, (4) less pain in °exion, (5) referred pain Subjects were excluded if they presented with his- not beyond knee, (6) pressure on the joint found tory of suspected spinal fracture or severe trauma; decreased range or increased resistance on the cauda equine syndrome with sensory impairment, painful side, (7) muscle spasm same side to the leg weakness and incontinence; medical diagnosis facet, (8) pain in extension, (9) pain in extension of spondylolisthesis, foraminal or central stenosis, with lateral °exion and rotation to the same side of scoliosis or other spinal deformities; extended pe- facet, (10) the injection into the joint relief pain, riod of steroid use; taking pain medication within (11) pain improves with °uoroscopically guided 24 h; pregnancy or menstruation. double-anesthetic blocks into the medial branch of the dorsal ramus which innervated the joints, and The examiners were two physical therapists (12) could not diagnosed from X-ray\". with clinical experiences in the musculoskeletal area of 20 and 9 years. They reviewed and prac- Clinically, the physical therapists could evaluate ticed the testing procedure together before the criteria 1 to 9 to determine whether the pain is caused by facet joint. Some studies attempted to validate clinical features of pain from facet joint

Hong Kong Physiother. J. 2018.38:125-131. Downloaded from www.worldscientific.com beginning of subject recruitment. After history Clinical examination for LBP 127 by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. taking, the ¯rst examiner assessed the participants. After a brief rest, the second examiner, who did not ranged from 3 months to 10 years. The pain see the examination process or knew the results, intensity on the day of assessment was 4:47 Æ 1:57. performed the assessment again. Both examiners also gave their impressions that the LBP was from The results of examination are presented in facet joint origin or not. The subjects were asked to Table 1. The numbers of yes and no determined by conceal the information of the ¯rst testing session. each examiner were di®erent for all items. How- Each session took about 10 min. The examination ever, the number of cases concluded to have or not order was randomized. The results of examination to have facet lesion determined by two examiners were analyzed for the agreements of each criteria was the same. and conclusion of facet joint diagnosis. Table 2 shows the agreement levels of each The sample size of subjects was estimated to examination item and the conclusion. All items had minimize the standard error associated with the percent agreement greater than 70%. The highest percent agreement between two arbitrary raters.20 level of percent agreement was the referred pain Setting the error margin at 15%, at least 44 not beyond knee. The lowest percent agreements patients were needed in the study. The percentage were reporting of pain location as unilateral and of agreement and generalized Kappa statistics were the palpation of muscle spasm on the same side of used to determine the agreement between thera- facet. pists for each of the dichotomous scale items of the examination. The levels of Kappa statistics agree- The Kappa coe±cients showed statistically ment were as follows: 0:00 < K < 0:20 poor or signi¯cant agreements between examiners for all slight agreement; 0:21 < K < 0:40 fair; 0:41 < examination items. There were di®erent levels of K < 0:60 moderate; 0:61 < K < 0:80 substantial agreement. Three items had fair agreement level, or good; 0:81 < K < 1:00 very good or almost four had moderate agreement, and two had sub- perfect.21 stantial agreement. The agreement of diagnosis if the patients had facet joint lesion was fair. Results Discussion There were 45 patients with LBP, 32 females and 13 males, participating in this study. The average To date, there are no speci¯c clinical and radio- age of subjects was 33.64 Æ 10.12 years. All sub- graphic indications of pain originated from facet jects had chronic LBP with duration of symptom joint.22 A systematic review showed that the con- trolled anesthetic block of the facet or its nerve supply, the medial branch had good psychometric properties for diagnosis of LBP.22 However, the test is invasive and not speci¯c, therefore it is not Table 1. Results of the examination of each criterion. Examination items Examiner 1 Examiner 2 Yes (%) No (%) Yes (%) No (%) (1) Unilateral local pain 23 (51.1) 22 (48.9) 29 (64.4) 16 (35.6) (2) Referred pain not beyond knee 37 (82.2) 8 (17.81) 41 (91.1) 4 (8.9) (3) No nerve root pain 36 (80.0) 9 (20.0) 39 (86.7) 6 (13.3) (4) Less pain in °exion 20 (44.4) 25 (55.6) 23 (51.1) (5) Pain in extension 34 (75.6) 11 (24.4) 38 (84.4) 22 (48.9) (6) Pain in extension with lateral °exion and rotation 33 (73.3) 12 (26.7) 36 (86.7) 7 (15.6) 6 (13.3) to the same side of facet (7) Muscle spasm same side to the facet 30 (66.7) 15 (33.3) 42 (93.3) 3 (6.7) (8) Activated pain with unilateral pressure on the 37 (82.2) 8 (17.8) 41 (91.1) 4 (8.9) joint or transverse process 40 (88.9) 5 (11.1) 39 (86.7) 6 (13.3) (9) Pressure on the joint found decreased range or 38 (84.4) 7 (15.6) 38 (84.4) 7 (15.6) increased resistance on the painful side Conclusion of facet joint origin

128 M. Vongsirinavarat, W. Wahyuddin & R. Adisaiphaopan Table 2. Percentage of agreement and Kappa Coe±cient of criteria and diagnosis. Examination items % Agreement Kappa coe±cient p-valuea Hong Kong Physiother. J. 2018.38:125-131. Downloaded from www.worldscientific.com (1) Unilateral local pain 73.3 0.436 0.001** by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. (2) Referred pain not beyond knee 91.1 0.622 < 0.001** (3) No nerve root pain 80.0 0.286 (4) Less pain in °exion 84.4 0.690 0.048* (5) Pain in extension 86.7 0.588 < 0.001** (6) Pain in extension with lateral °exion and rotation to 77.8 0.324 < 0.001** the same side of facet 73.3 0.250 0.017* (7) Muscle spasm same side to the facet 86.7 0.433 (8) Activated pain with unilateral pressure on the joint 0.011* 88.9 0.483 0.002** or transverse process (9) Pressure on the joint found decreased range or 86.7 0.492 0.001** increased resistance on the painful side 0.001** Conclusion of facet joint origin ap-values for Kappa Statistics; *statistical signi¯cance at p < 0:05; **statistical signi¯cance at p< 0:01. commonly performed even in orthopedics clinic. result in di®erent interpretations by di®erent The criteria for diagnosing by symptom response examiners. The interview and movement tests and physical examination still needed for clinical might be lacking of de¯nition and training con- use had face validity from the expert consensus in a sideration, since they are usually supposed as basic Delphi study.17 clinical skill of therapists. For the clinical examination tested in this study, There were three examination items which three criteria of symptom behaviors, including needed tactile determination in this study includ- unilateral local pain, referred pain not beyond ing muscle spasm same side to the facet, activated knee, and no nerve root pain, had various levels of pain with unilateral pressure on the joint or agreements with Kappa coe±cients ranged from transverse process, and pressure on the joint found 0.286 to 0.622 (% agreement from 73.3–91.1). decreased range or increased resistance on the There were three tests determining pain responses painful side. These items had low to fair agree- to movements, including less pain in °exion, pain ments with Kappa coe±cients of 0.250 to 0.483 in extension, and pain in extension with lateral (% agreement from 73.3–88.9). Systematic reviews °exion and rotation to the same side of facet. This reported generally low reliability of palpation- examination domain also showed the same trend as based assessment.25–27 There were con°icting evi- the symptom behaviors with Kappa coe±cients dences about the reliability of evaluating muscle from 0.324 to 0.690 (% agreement from 77.8–86.7). tension or spasm as well as the intersegmental The evidences of reliability of pain response to sti®ness of vertebral disc.24,27–29 Strong evidence repeated movements were reportedly controversial indicating low reliability pain on palpation and in a systematic review.23 However, based on the trigger points was also reported. Consistent results movement impairment classi¯cation system ex- showed that the judgments based on visual and amined, Van Dillen et al. reported the very good tactile information were usually di±cult to perform level of agreements (K > 0:75) of physical exami- reliably.25–29 nation items of symptom behaviors in patients with LBP.24 In their study, the researchers estab- The agreement for the pressure on the joint lished and de¯ned the wording of questions and found decreased range or increased resistance on response choices before testing. However in our the painful side was moderate in this study. How- study, the interview and movement tests were ever, multiple reviews have suggested low agree- performed in the same manner as routinely done in ment of this manual assessment.27–29 The better the physical therapy clinic which did not control agreement shown in this study might be due to the way to ask or test patients. With this manner, both the subject and examiner characteristics. The the answers and responses from each patient might patients recruited in this study had chronic symptom duration longer than three months with

Clinical examination for LBP 129 Hong Kong Physiother. J. 2018.38:125-131. Downloaded from www.worldscientific.com moderate pain scale which might be related to the therefore these issues were not standardized which by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. changes of spinal resistance. The examiners might be the source of disagreement in patient might have similar manual technique and inter- responses. However, due to skewness of symptom pretation of pressure on joint since they graduated response, the Kappa coe±cients were only low to from the same physical therapy school, have been moderate in this study. The items of \\muscle working in the same clinical setting, and have spasm same side to the facet\", \\no nerve root several chances to discuss and practice in terms of pain\", and \\pain in extension with lateral °exion manual therapeutic procedure together. They also and rotation to the same side of facet\" which had reviewed and practiced the testing procedure to- Kappa coe±cients less than 0.40 (fair level of gether before beginning of subject recruitment in agreement) might need special cautions when per- this study. The three-point grading (hypomobility, formed and interpreted for facet diagnosis. Further normal, and hypermobility) which used in study with variety of sign and symptoms of LBP this study was also recommended in review by would be bene¯cial to con¯rm the reliability of Wong et al.27 physical examination of facet joint in physical therapy clinic. The validity study compared with The agreement level of conclusion was moderate standard tests, i.e., nerve block and intra-articular in this study. The review of agreement levels of injection would also verify these criteria of diag- di®erent diagnosis systems used in physical thera- nosis. The reliability study assessing therapists py clinics reported variability of agreements with di®erent clinical experiences would be valu- depended on the methodology and de¯nitions used able to prove clinical practicality of the testing in the study.24 Since the physical examination in protocol. Also, the study using reliable instru- physical therapy clinic depends largely on com- mental spinal sti®ness measurements27,31 might munication, manual skills and judgments of add clinical insight to this speci¯c lesion condition. examiners, the explicitly de¯ned techniques, oper- ational de¯nition, and consistent training would Acknowledgments be necessary for improving reliability.25 The authors would like to thank all participants The major limitation of result interpretation of for their cooperation. The physical therapy sta® of this study is due to the statistics used. The results the Physical Therapy Center also provided a great of low Kappa coe±cient values in this study were help in the subject recruitment process. partly due to the small number of some response category results from characteristics of the study Con°ict of Interest sample. This would result in skewed response dis- tribution and e®ect on Kappa statistics.30 More The authors have no con°icts of interest relevant studies which used greater variety of symptom and to this paper. examination responses would be needed to con¯rm the agreement of therapists. Funding/Support In addition, the study to test validity of No ¯nancial or material support of any kind was the criteria set is warrant. The construct validity received for the work described in this paper. examination using factor analysis would result in the known redundant items which guide to more concrete set of examination instrument. Conclusion Author Contributions There was adequate reliability between two Vongsirinavarat M took part in study conception examiners showed by the percent agreements and design, data collection and analysis, as well as greater than 70% for all items used for con¯rming writing and revision of the manuscript. Wahyuddin facet joint lesion. The test protocol of all assess- W was involved in study design and manuscript ment items was reviewed and practiced together by revision. Adisaiphaopun R performed subject screen- both examiners. However, the words of questions in ing and data collection. All authors have given ¯nal history taking part and the manual techniques approval of the version to be published. used were assumed to be routine practice in clinic,

130 M. Vongsirinavarat, W. Wahyuddin & R. Adisaiphaopan Hong Kong Physiother. J. 2018.38:125-131. Downloaded from www.worldscientific.com References 15. Starkey C, Brown SD, Ryan JL. Examination of by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. Orthopedic and Athletic Injuries. 3rd ed. Phila- 1. Mohseni MA, Stephenson BR, Richardson B. delphia, PA: F.A. Davis Company, 2010. Spinal manipulation in the treatment of low back pain: A review of the literature with particular 16. Hresko MT. Thoracic and lumbosacral spine. In: emphasis on randomised controlled clinical trials. Steinberg G, ed. Orthopaedics in Primary Care. Phys Ther Rev 1998;3:185–94. 2nd ed. Baltimore: Williams & Wilkins, 1992. 2. Kuslich SD, Ulstrom CL, Michael CJ. The tissue 17. Wilde VE, Ford JJ, McMeeken JM. Indicators of origin of low back pain and sciatica: A report of lumbar zygapophyseal joint pain: Survey of an ex- pain response to tissue stimulation during opera- pert panel with the Delphi technique. Phys Ther tion on the lumbar spine using local anesthesia. 2007;87(10):1348–61. Orthop Clin North Am 1991;22:181–7. 18. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine 3. Bogduk N. Psychology and low back pain. Int J G, Bogduk N. Clinical features of patients with Osteopath Med 2006;9:49–53. pain stemming from the lumbar zygapophysial joints. Is the lumbar facet syndrome a clinical en- 4. Waddell G. The Back Pain Revolution. Edinburgh: tity? Spine 1994;19:1132–7. Churchill Livingstone, 2004. 19. Pang WW, Mok MS, Lin ML, Chang DP, Hwang 5. Manchikanti L, Pampati V, Fellows B, Ghafoor A. MH. Application of spinal pain mapping in the The inability of the clinical picture to characterize diagnosis of low back pain — analysis of 104 cases. pain from facet joints. Pain Physician 2000; Acta Anaesthesiol Sin 1998;36:71–4. 3:158–66. 20. Gwet K. Handbook of Inter-Rater Reliability: 6. Manchikanti L, Kaye AD, Boswell MV, et al. The De¯nitive Guide to Measuring the Extent A systematic review and best evidence synthesis of of Agreement Among Multiple Raters, 3rd ed. e®ectiveness of therapeutic facet joint interventions Maryland, USA: Advanced Analytics, LLC, in managing chronic spinal pain. Pain Physician 2012. 2015;18:E535–82. 21. Landis JR and Koch GG. The measurement of 7. Suri P, Hunter DJ, Rainville J, Guermazi A, Katz observer agreement for categorical data. Biometrics JN. Presence and extent of severe facet joint 1977;33:159–74. osteoarthritis are associated with back pain in older adults. Osteoarthritis Cartilage 2013;21 22. Sehgal N, Dunbar EE, Shah RV, Colson J. (9):1199–206. Systematic review of diagnostic utility of facet (zygapophysial) joint injections in chronic spinal 8. Flavell CA, Gordon S, Marshman L. Classi¯ca- pain: An update. Pain Physician 2007;10(1): tion characteristics of a chronic low back pain 213–28. population using a combined McKenzie and patho-anatomical assessment. Man Ther 2016; 23. May S, Littlewood C, Bishop A. Reliability of 26:201–7. procedures used in the physical examination of non-speci¯c low back pain: A systematic review. 9. Manchikanti L, Hirsch JA, Falco FJE, Boswell Aust J Physiother 2006;52:91–102. MV. Management of lumbar zygapophysial (facet) joint pain. World J Orthop 2016;7(5): 24. Van Dillen LR, Sahrmann SA, Norton BJ, et al. 315–37. Reliability of physical examination items used for classi¯cation of patients with low back pain. Phys 10. Manchikanti L, Pampati V, Fellows B, et al. Ther 1998;78:979–88. In°uence of psychological factors on the ability to diagnose chronic low back pain of facet joint origin. 25. Maher CG, Adams R. Reliability of pain and Pain Physician 2001;4(4):349–57. sti®ness assessments in clinical manual lumbar spine examination. Phys Ther 1994;74(9):801–9. 11. Sahrmann S. Diagnosis and Treatment of Move- ment Impairment Syndromes. 1st ed. St. Louis, 26. Strender LE, Sjo€blom A, Sundell K, Ludwig R, MO: Mosby, 2001. Taube A. Interexaminer reliability in physical ex- amination of patients with low back pain. Spine 12. Eubanks JD, Lee MJ, Cassinelli E. Prevalence of 1997;22(7):814–20. lumbar facet arthrosis and its relationship to age, sex, and race: An anatomic study of cadaveric 27. Wong A, Kawchuk G. The clinical value of lumbar specimens. Spine 2007;32:2058–62. posterioanterior segmental sti®ness: A narrative review. 2017 PM&R 2017. 13. Kalichman L, Li L, Kim DH, et al. Facet joint osteoarthritis and low back pain in the community- 28. Se±nger MA, Najm WI, Mishra SI, et al. Reli- based population. Spine 2008;33:2560–5. ability of spinal palpation for diagnosis of back and neck pain: A systematic review of the literature. 14. Kirkadly-Willis WH. The relationship of Spine 2004;29:E413–25. structural pathology to nerve root. Spine 1984; 9:49–52. 29. Snodgrass SJ, Haskins R, Rivett DA. A structured review of spinal sti®ness as a kinesiological outcome

of manipulation: Its measurement and utility in Clinical examination for LBP 131 diagnosis, prognosis and treatment decision- making. J Electromyogr Kinesiol 2012;22:708–23. 31. Wong AY, Kawchuk G, Parent E, Prasad N. 30. Lantz CA. Application and evaluation of the kappa Within- and between-day reliability of spinal statistic in the design and interpretation of chiro- sti®ness measurements obtained using a computer- practic clinical research. J Manipulative Physiol controlled mechanical indenter in individuals with Ther 1977;20:521–8. and without low back pain. Man Ther 2013;18 (5):395–402. Hong Kong Physiother. J. 2018.38:125-131. Downloaded from www.worldscientific.com by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles.

Research Paper Hong Kong Physiotherapy Journal Vol. 38, No. 2 (2018) 133–139 DOI: 10.1142/S1013702518500117 Hong Kong Physiother. J. 2018.38:133-139. Downloaded from www.worldscientific.com Hong Kong Physiotherapy Journal by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. http://www.worldscientific.com/worldscinet/hkpj Side-to-side elbow range of movement variability in an ulnar neurodynamic test sequence variant in asymptomatic people Michelle Meng Yim Tong, Vincent Cheng-Hsin Liu and Toby Hall* School of Physiotherapy and Exercise Science, Curtin University of Technology, Perth, Western Australia 6102, Australia *[email protected] Received 25 March 2017; Accepted 28 July 2017; Published 14 August 2018 Background: Range of motion (ROM) asymmetry between sides is one indicator of a positive neurodynamic test, but this has been less well studied for the ulnar nerve. Objective: The purpose of this study was to investigate side-to-side variation in elbow ROM during an ulnar neurodynamic test sequence, including contralateral cervical side °exion, in 40 asymptomatic subjects. Methods: A traditional goniometer was used to measure elbow °exion ROM at two end points, onset of resistance (R1) and symptom onset (P 1). Two repeated measures of R1 and P 1 were taken on each side. Results: Reliability for R1 and P 1 was found to be good (ICC ! 0:83, SEM 5:37) with no signi¯cant di®erence in mean ROM between sides. A signi¯cant relationship between sides was seen (r values ! 0:48) and R2 values > 0:23; this indicates at least 23% of the variance observed in one limb was accounted for by range in the opposite limb. This relationship was slightly stronger for R1 than P 1. Lower bound scores indicate that intra-individual ROM di®erence > 23 for R1 and 22 for P 1 would exceed normal ROM asymmetry. Conclusion: These ¯ndings provide clinicians with background information of ROM asymmetry during the ulnar neurodynamic test. Keywords: Ulnar Neurodynamic test; upper limb; variability; reliability. *Corresponding author. Copyright@2018, Hong Kong Physiotherapy Association. Published by World Scienti¯c Publishing Co Pte Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 133

Hong Kong Physiother. J. 2018.38:133-139. Downloaded from www.worldscientific.com 134 M. M. Y. Tong, V. C.-H. Liu & T. Hall The purpose of this study was to investigate by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. side-to-side variation in elbow ROM during the Introduction ulnar neurodynamic test sequence, using the se- quence described by Hall and Elvey4 in asymp- Increased neural tissue mechanosensitivity evalu- tomatic people. Two end points were investigated: ated by neurodynamic tests1,2 is frequently repor- P 1, perceived by the subject at onset of the dis- ted during the examination of patients with comfort, and R1, determined by examiner, onset of musculoskeletal disorders.3,4 In particular, the resistance. The results of this study should provide ulnar nerve neurodynamic test is recommended in clinician with background information regarding the examination of cubital tunnel syndrome,5,6 elbow ROM variability for the ulnar neurodynamic thoracic outlet syndrome,7 and C8 nerve root test in normal subjects, which may enable the de- radiculopathy.8 termination of a positive test in symptomatic people. To de¯ne a positive neurodynamic test, the fol- lowing have been recommended. Firstly, the Methods patient's symptoms must be reproduced and sensi- tizing maneuvres must increase or decrease symp- Study design toms.2,3,8–10 Secondly, there should be a discrepancy in joint range between sides.2 Finally, increased re- A within-subject comparative measurement design sistance is perceived by the examiner on the side of was used to identify di®erences between sides symptoms.2,3,11 Side-to-side discrepancy in ROM during the ulnar neurodynamic test in asymptom- and reproduction of symptoms are considered the atic people. The objective of this study was to de- most essential criteria for interpretation of neuro- termine the minimum side-to-side elbow ROM dynamic tests,2,12 and also useful comparable signs asymmetry required to classify an abnormal re- to evaluate treatment.13 sponse to this speci¯c test. Elbow ROM is frequently used as an outcome Participants measure in studies investigating upper limb neu- rodynamic tests, due to the ease of side-to-side Forty asymptomatic subjects (19 females and 21 comparison.1,14,15 Various studies show that males, mean age 30.14 years) were included in this asymmetry in elbow ROM between sides is com- study. Participants were excluded if they had a mon during neurodynamic testing in asymptomatic current or previous history of trauma to the cer- people.16–18 Within-person side-to-side variability vical spine, thoracic spine, shoulder, elbow, wrist, for the median nerve was reported between 15.5 or hand. They were also excluded if they had any and 27.16,18 Within-person side-to-side variability limitation of ROM in the upper quadrant. All for the radial nerve was reported between 11.2 participants underwent pre-test screening to en- and 20.16,18 Similarly, between limb values of sure that they had pain-free and normal range of elbow ROM for the ulnar nerve was 21.16 The upper limb joint movement. This study received di®erence in values for elbow ROM reported by approval from Curtin University Human Research Covill and Petersen16 and Stalioraitis et al.18 may Ethics Committee. All participants were provided be due to di®erent testing sequences. Speci¯cally, with information and gave informed consent. Using the addition of contralateral cervical side °exion in the two-tailed paired t-test, with an alpha level the latter study reduced variability ROM between 0.05, power of 0.8, and a medium e®ect size of 0.5, sides, potentially increasing the probability that 34 subjects were calculated to be needed for this ROM can be used to determine a positive neuro- study. dynamic test.18 Structural di®erentiation to de- termine if symptom provocation is neurogenic in Equipment and measurements origin for the case of upper limb neurodynamic tests is determined by assessing the e®ect of adding The independent variable was side (left or right). contralateral cervical side °exion. Hence, it is im- The dependent variable was range of elbow °exion. portant to know side-to-side variation in ROM Extraneous variables include body mass index, age, during neurodynamic tests during di®erent neuro- gender, and hand-dominant side. A traditional dynamic test variants, including cervical side °ex- ion, as this can provide the clinician with an expectation of what could potentially be normal variance.

Hong Kong Physiother. J. 2018.38:133-139. Downloaded from www.worldscientific.com goniometer was used to record elbow ROM. Side-to-side elbow ROM variability 135 by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. Acceptable validity for measurement of elbow ROM using a traditional goniometer when com- Neurodynamic test sequence for the pared with radiograph measurement has been Ulnar nerve reported.19 In that study, the intra-class correla- tion coe±cient (ICC) ranged from 0.94 to 0.97 for The tested arm was positioned in 90 shoulder the goniometric measurements and from 0.98 to abduction, 90 shoulder external rotation with the 0.99 for the radiographic measurements.19 The two shoulder girdle maintained in neutral. The cervical methods correlated and the maximum error of the spine was placed in maximal lateral °exion to the goniometric measurement was 7.0 for °exion, 95% contralateral side.4 The elbow was fully extended, of the time.19 with the forearm in maximum pronation, and wrist/¯ngers maximally extended. Elbow °exion Procedure was initiated, and at P 1 and R1, elbow movement was paused while ROM recorded. The neurody- Participants were tested according to a standard namic test continued until both end points had clinical testing protocol, without ¯xation devices. been achieved. The subjects were given one famil- The untested limb was placed in a relaxed position iarization trial on each side followed by two trials with the hand resting on the abdomen. The cer- where measurements were recorded in between a vical spine was placed in contralateral side °exion 10-second rest interval. to the side tested, without rotation. The shoulder girdle on the tested side was held in neutral Data Analysis elevation and/depression position manually by the examining therapist to mimic the clinical Data analysis was carried out using SPSS v19. situation. (SPSS Inc., 444 N. Michigan Avenue, Chicago, Illinois, 60611). All data were normally distributed. The ulnar neurodynamic test sequence was Intra-tester reliability for repeated measures on tested on each side in random order. The parti- each arm was calculated using ICC (2,1), standard cipants underwent one familiarization trial. Two error of the measurement (SEM), and minimal de- measurements of range of elbow movement were tectable change (MDC). Mean elbow ROM and recorded after the familiarization trial using the standard deviation was determined for the Ulnar traditional goniometer by a separate independent neurodynamic test sequence for both arms. researcher while the main researcher maintained Dependent t-tests were used to compare within- the arm position during the measurement process. subject range of motion (ROM) between the right The goniometer was not visible to the main and left arms for each test. Relationship in ROM researcher to avoid bias. The goniometer axis between limbs was calculated using the Pearson was aligned with the medial epicondyle, with correlation coe±cient and coe±cient of determi- the proximal arm aligned with the midline of nation (r2). The mean absolute values (MAVs) humerus and the distal arm aligned with the line were calculated to determine di®erences between formed by the medial epicondyle and radial limbs while ensuring that all values remain positive. styloid process. A lower bound score was used to determine the cut-o® point at which the degree of di®erence Good intra-tester reliability of goniometric between limbs could be considered greater than measurement has been shown when the mean of that accounted for by measurement error and var- two or three measurements is taken,20–22 hence iability. This was carried out according to the only two measurements were taken for each end method reported in another study by multiplying point and each test. The end points were R1 and the standard deviation of the MAV by the P 1, as these have been shown to have excellent z-score (1.65) of a one-tailed t-test ( ¼ 0:05) inter- and intra-rater reliability.21,23,24 Partici- and adding the MAV (lower bound score ¼ (SD) pants were instructed to say \\now\" upon the onset (1.65) þ MAV).16 of any sensation change during the neurodynamic test and the movement paused for measurement Results purposes. The examiner said \\R\" when the onset of resistance was felt and again the movement was All data were checked and found to be normally paused for measurement. distributed. The results for intra-therapist reliability

Hong Kong Physiother. J. 2018.38:133-139. Downloaded from www.worldscientific.com 136 M. M. Y. Tong, V. C.-H. Liu & T. Hall more variability between limbs than P 1 during the by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. ulnar neurodynamic test. Table 1. Reliability statistics for elbow ROM for ulnar neurodynamic test (n ¼ 40). Discussion Measurement ICC [2,1] (95% CI) SEM MDC This study investigated side-to-side variation in elbow ROM at P 1 and R1 for the ulnar neurody- Right Ulnar R1 0.83 (0.67, 0.91) 5.4 14.9 namic test sequence as described by Hall and Right Ulnar P 1 0.84 (0.69, 0.91) 4.7 13.0 Elvey.4 Small mean di®erences were detected be- Left Ulnar R1 0.90 (0.81, 0.95) 3.7 10.1 tween sides for R1 (1.6) and P 1 (1.1). However, Left Ulnar P 1 0.90 (0.81, 0.95) 4.2 11.6 despite these small mean di®erences, this did not equate to a strong correlation between sides as are shown in Table 1. For both R1 and P 1, the seen in Table 2. This might be explained by the range recorded during Ulnar neurodynamic test, relatively large MAVs for ROM di®erences be- ICC (2,1) values was greater than 0.83 indicating tween limbs, which indicate large intra-individual good reliability.16 In addition, the SEM and MDC di®erences in ROM between limbs as shown in for each assessment point were also relatively Table 3. small. The MAVs for discrepancy between sides were Means and standard deviations for elbow ROM similar in order of magnitude for both R1 and P 1. during the ulnar neurodynamic test are presented The lower bound scores were calculated from the in Table 2. The mean di®erence between the left MAV, and indicate that elbow ROM di®erence and right sides, for both R1 and P 1, was very between limbs for the ulnar neurodynamic test small, with at most 1.5 between sides. At any must be greater than 23 for R1 and 22 for P 1 assessment point, there was no signi¯cant di®er- for the ROM ¯ndings to be considered relevant ence between the left and right sides as re°ected by beyond normal variation and measurement error. the 95% con¯dence intervals (Table 2). These ¯ndings were similar to those reported by Covill and Petersen,16 who had also investigated A Pearson correlation analysis revealed a sig- the Ulnar neurodynamic test. In that study, the ni¯cant relationship between the limbs, with r MAV was 6.1 and lower bound score was 20.9. values greater than 0.48. Furthermore, the R2 Small di®erences in MAVs for ROM between values were greater than 0.23, indicating that at Covill and Petersen16 and the current study were least 23% of the variance observed in one limb was likely to be attributed to di®erences in end-point accounted for by range in the opposite limb. This measurement, type of goniometer used, and vari- relationship was slightly stronger for R1 than P 1. ation in neurodynamic test sequence. Also, P 2 and These data point to a relationship for elbow ROM R2 were the end points measured by Covill and between limbs, indicating that elbow ROM of one Petersen,16 while in the present study, P 1 and R1 side can be used to some degree to predict elbow were recorded instead. To the best of our knowl- ROM of the opposite limb. edge, no other study has reported lower bound scores for the ulnar neurodynamic test. The MAV and lower bound scores shown in Table 3 revealed some degree of variability be- tween the right and left limbs for any assessment point. Elbow ranges recorded at R1 had slightly Table 2. Mean range, mean di®erences between left and right sides (SD) with 95% con¯dence interval (CI), Pearson correlation coe±cient (r), and coe±cient of determination (R2) (n ¼ 40). Measurement Mean range (SD) Mean di®erence r R2 Left Right scores (95% CI) Ulnar R1 110.8 (1.6) 112.4 (13.0) 1.6 (À2.5, 5.6) 0.53 0.28 Ulnar P 1 110.8 (13.2) 111.9 (11.8) 1.1 (À2.8, 5.0) p < 0:001 0.23 0.48 p < 0:001

Table 3. Mean absolute di®erences (MAV) in elbow Side-to-side elbow ROM variability 137 ROM between right and left sides together with lower bound scores for neurodynamic testing (n ¼ 40). for P 2. As such, the use of P2 as an end-point measure might explain the greater mean elbow Measurement MAV (SD) Lower bound scores ROM reported by Covill and Petersen.16 There are di®erences of opinion as to the use of end-point Ulnar R1 9.7 (8.1) 23.1 measure with P 1 being more clinically relevant as Ulnar P 1 9.0 (8.2) 22.5 it is included in the de¯nition of a positive neuro- dynamic test to reproduce the patient's symp- Hong Kong Physiother. J. 2018.38:133-139. Downloaded from www.worldscientific.com MAVs of 10.1 and 6.7 for the median and ra- toms3,8 without bringing on greater pain. by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. dial neurodynamic tests were also reported by The correlation (R2) in elbow ROM between Covill and Petersen,16 which are higher than that sides was higher being 0.23 for R1 and P 1 in this reported by Stalioraitis et al.18 where ROM values study compared to 0.13 reported by Covill and were 5.5 for R1 and 5.8 for P 1 for the median Petersen.16 Although values varied slightly, the nerve and 4.2 for R1 and 4.8 for P 1 for the radial clinical interpretation of these small R2 values was nerve. As such, it might be expected that the MAV that it may not be possible to use elbow ROM comparison between sides as one of the criteria to for the ulnar nerve reported by Covill and Peter- support a positive ulnar neurodynamic test. In sen16 would be greater than those in the current other words, the range in one limb accounts for only a small proportion of the predicted range in study, but the reverse was seen. One explanation for the opposite limb during this neurodynamic test. This would indicate that other criteria should be these di®erences between the three studies might be used to identify a positive neurodynamic test, with ROM di®erence between limbs being of minor in the type of measurement device which in the importance. current study was a traditional goniometer. In Intra-tester reliability was found to be good for the left and right sides of ulnar neurodynamic test contrast, an electrogoniometer has often been used (Table 1). These ¯ndings are consistent with other in the previous research.16,18 Additionally, the type studies reporting reliability of other neurodynamic tests when measuring elbow ROM with an elec- of neurodynamic sequence might also a®ect the trogoniometer.18,21,24 Neither the left nor right side showed any indication of substantially better intra- di®erence in MAV. A di®erent test sequence that tester reliability. Hence, a possible conclusion would be the degree of intra-tester reliability which did not include the contralateral cervical side °ex- is not related to the side measured and it would ion was used by Covill and Petersen.16 Cervical side suggest that future neurodynamic test studies should include intra-tester reliability within every °exion is an important component of neurodynamic study. testing used in structural di®erentiation which The use of traditional goniometer has been recommended to depict a realistic clinical situa- increases strain on the nervous system without dif- tion.16 Despite this, a traditional goniometer is clumsy to use, as the therapist must control many fering the mechanical load on the musculoskeletal components during the ulnar neurodynamic test. system.23 The e®ect of including contralateral cer- Trying to correctly place the goniometer during the neurodynamic testing process greatly adds to vical side °exion in the median neurodynamic test this complexity and makes it di±cult to accurately sequence was to reduce mean elbow ROM to 132.8 measure ROM. Future studies should look for simpler methods of measuring ROM that can be in asymptomatic subjects whereas the same se- easily applied and read by the clinician. quence without contralateral cervical side °exion Conclusion had achieved elbow ROM of 149.23 The results of this study provide clinicians with Mean ROM di®erence between sides for R1 and baseline knowledge of normal ROM variation P 1 was noted to be similar to Stalioraitis et al.18 with the exception of R1 for the median nerve which had a mean di®erence of only 0.9. This similarity may be attributed to the consistency in neurodynamic test sequence used in both studies. The mean elbow ROM values were smaller in this study than those reported by Covill and Petersen12 for the ulnar neurodynamic test. This could be explained by the various di®erences in methodology used by the two studies, most nota- bly, the di®erent end-point measures. As observed by Vanti et al.,21 for the median neurodynamic test, mean elbow ROM was 155 for P 1 and 164

Hong Kong Physiother. J. 2018.38:133-139. Downloaded from www.worldscientific.com 138 M. M. Y. Tong, V. C.-H. Liu & T. Hall Phys Ther 2008;38(9):A1–34. doi:10.2519/jospt. by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. 2008.0303. between sides during the ulnar neurodynamic test 2. Nee RJ, Jull GA, Vicenzino B, Coppieters M. The using a commonly used measurement device. The validity of upper-limb neurodynamic tests for lower bound score of 23 for measurement at R1 detecting peripheral neuropathic pain. J Orthop and 22 for P 1 would suggest that side-to-side Sports Phys Ther 2012;42(5):413–24. doi:10.2519/ variation of more than 23 would exceed normal jospt.2012.3988. variability and would likely not be due to mea- 3. Butler D. The Sensitive Nervous System. 1st ed. surement errors and is therefore clinically relevant. Adelaide: Noigroup Publications, 2000. Large variation in ROM between sides indicates 4. Hall T, Elvey RL. Neural tissue evaluation and that ROM is less helpful in determining a positive treatment. In: Donatelli RA, ed. Physical Therapy neurodynamic test than other test criteria. One of the Shoulder. 5the ed. Chap. 6. Saint Louis: explanation for large side-to-side variation in ROM Churchill Livingstone, 2012:131–45. is the cumbersome nature of using a traditional 5. Coppieters M, Bartholomeeusen K, Stappaerts K. goniometer during neurodynamic testing. Incorporating nerve-gliding techniques in the con- servative treatment of cubital tunnel syndrome. Con°ict of Interest J Manipulative Physiol Ther 2004;27(9):560–68. doi:10.1016/j.jmpt.2004.10.006. The authors declare that there is no con°ict of 6. Goyal M, Mehta S, Rana N, et al. Motor nerve interest relevant to the study. conduction velocity and function in carpal tunnel syndrome following neural mobilization: A ran- Funding/Support domized clinical trial. Int J Health Allied Sci 2016;5 (2):104–10. doi:10.4103/2278-344x.180434. This research did not receive any speci¯c grant 7. Sanders RJ, Hammond SL, Rao NM. Diagnosis of from funding agencies in the public, commercial, or thoracic outlet syndrome. J Vasc Surg 2007;46 not-for-pro¯t sectors. The authors of this study (3):601–4. doi:10.1016/j.jvs.2007.04.050. would like to acknowledge Karen Lam for her 8. Shacklock MO. Clinical Neurodynamics: A New contributions to this study. System of Musculoskeletal Treatment. Edinburgh: Elsevier Butterworth-Heinemann. Author Contributions 9. Manvell JJ, Manvell N, Snodgrass SJ, Reid SA. Improving the radial nerve neurodynamic test: An Study design was initiated by Dr. Toby Hall. Data observation of tension of the radial, median and Collection was collected by Ms. Michelle Tong ulnar nerves during upper limb positioning. Man with the assistance of Mr. Vincent Liu. Michelle Ther 2015;20(6):790–6. doi:10.1016/j.math.2015. and Vincent had written the majority of this paper 03.007. under the supervision of Dr. Toby Hall. Data 10. Mintken P, Puentedura E, Louw A. Neurodynamic Analysis was done by Dr. Toby Hall. Subsequent interventions and physiological e®ects: Clinical revisions of the drafts made for publication were neurodynamics in neck and upper extremity pain. done by Michelle under the supervision of In: de las Peñas CF, Cleland JA, Huijbregts PA, Dr. Toby Hall. Each author has reviewed and eds. Neck and Arm Pain Syndrome. Edinburgh: approved the submission of this updated draft of Churchill Livingstone, 2011:496–515. the manuscript and takes full responsibility for the 11. Hall T, Elvey RL, Davies N, Dutton L, Moog M. manuscript. E±cacy of manipulative physiotherapy for the treatment of cervicobrachial pain. In: Tenth Bien- References nial Conf MPAA. Melbourne: Manipulative Phy- siotherapists Association of Australia, 1997. 1. Childs JD, Cleland JA, Elliott JM, Teyhen DS, 12. Butler D, Gi®ord L. The concept of adverse me- Wainner RS, Whitman JM. Neck pain: Clinical chanical tension in the nervous system Part 1: practice guidelines linked to the International Clas- Testing for \\Dural tension\". Physiotherapy 1989; si¯cation of Functioning, Disability, and Health 75(11):622–9. doi:10.1016/s0031-9406(10)62374-7. from the Orthopedic Section of the American 13. Vicenzino B, Collins D, Wright A. The initial Physical Therapy Association. J Orthop Sports e®ects of a cervical spine manipulative physio- therapy treatment on the pain and dysfunction of lateral epicondylalgia. Pain 1996;68(1):69–74. 14. Nee RJ, Vicenzino B, Jull GA, Cleland JA, Cop- pieters MW. Neural tissue management provides

Hong Kong Physiother. J. 2018.38:133-139. Downloaded from www.worldscientific.com immediate clinically relevant bene¯ts without Side-to-side elbow ROM variability 139 by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. harmful e®ects for patients with nerve-related neck and arm pain: A randomised trial. J Physi- measurements: Comparative study with a radio- other 2012;58(1):23–31. doi:10.1016/s1836-9553(12) graphic method. Clin Orthop Relat Res 2011;469 70069-3. (11):3134–40. doi:10.1007/s11999-011-1986-8. 15. Wainner RS, Fritz JM, Irrgang JJ, Delitto A, 20. Lohkamp M, Small K. Normal response to upper Allison S, Boninger ML. Development of a clinical limb neurodynamic test 1 and 2A. Man Ther 2011; prediction rule for the diagnosis of carpal tunnel 16(2):125–30. doi:10.1016/j.math.2010.07.008. syndrome. Arch Phys Med Rehabil 2005;86(4):609– 21. Vanti C, Bon¯glioli R, Calabrese M, Marinelli F, 18. doi:10.1016/j.apmr.2004.11.008. Violante FS, Pillastrini P. Relationship between 16. Covill LG, Petersen SM. Upper extremity neurody- interpretation and accuracy of the upper limb namic tests: Range of motion asymmetry may not neurodynamic test 1 in carpal tunnel syndrome. J indicate impairment. Physiother Theory Pract 2012; Manipulative Physiol Ther 2012;35(1):54–63. 28(7):535–41. doi:10.3109/09593985.2011.641198. doi:10.1016/j.jmpt.2011.09.008. 17. Martinez MD, Cubas CL, Girbes EL. Ulnar nerve 22. Yaxley GA, Jull GA. A modi¯ed upper limb ten- neurodynamic test: Study of the normal sensory sion test: An investigation of responses in normal response in asymptomatic individuals. J Orthop subjects. Aust J Physiother 1991;37(3):143–52. Sports Phys Ther 2014;44(6):450–6. doi:10.2519/ doi:10.1016/S0004-9514(14)60536-5. jospt.2014.5207. 23. Coppieters M, Stappaerts K, Janssens K, Jull G. 18. Stalioraitis V, Robinson K, Hall T. Side-to-side Reliability of detecting `onset of pain' and `sub- range of movement variability in variants of the maximal pain' during neural provocation testing of median and radial neurodynamic test sequences the upper quadrant. Physiother Res Int 2002;7 in asymptomatic people. Man Ther 2014;19(4): 338– (3):146–56. 42. doi:http://dx.doi.org/10.1016/j.math. 2014.03.005. 24. Oliver GS, Rushton A. A study to explore the re- 19. Chapleau J, Canet F, Petit Y, La°amme GY, liability and precision of intra and inter-rater Rouleau DM. Validity of goniometric elbow measures of ULNT1 on an asymptomatic popula- tion. Man Ther 2011;16(2):203–6. doi:10.1016/j. math.2010.05.009.

Research Paper Hong Kong Physiotherapy Journal Vol. 38, No. 2 (2018) 141–147 DOI: 10.1142/S1013702518500129 Hong Kong Physiother. J. 2018.38:141-147. Downloaded from www.worldscientific.com Hong Kong Physiotherapy Journal by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. http://www.worldscientific.com/worldscinet/hkpj Strati¯cation of stroke rehabilitation: Five-year pro¯les of functional outcomes Bryan Ping Ho Chung Physiotherapy Department, Tai Po Hospital Tai Po, New Territories, Hong Kong [email protected] Received 6 June 2017; Accepted 13 September 2017; Published 14 August 2018 Background: Stroke rehabilitation in inpatient setting requires high intensity of manpower and resources. Early strati¯cation of patients with stroke could facilitate early discharge plan and reduce avoidable length of stay (LOS) in hospital. Strati¯cation of patients with stroke in clinical setting is usually based on functional scores which are quite time-consuming and require a special training to complete the full score. Objective: The objective of the study was to explore whether Modi¯ed Functional Ambulation Category (MFAC) can serve as a strati¯cation tool of patients with stroke in inpatient rehabilitation. Methods: This was a retrospective, descriptive study of the demographic, functional outcomes of patients with stroke in an inpatient rehabilitation center. A total of 2,722 patients completed a stroke rehabilitation program from 2011 to 2015 were recruited. The patients were divided into seven groups according to their admission MFAC. The between-group di®erence in LOS, functional outcomes at admission and discharge including Modi¯ed Rivermead Mobility Index (MRMI) and Modi¯ed Barthel Index (MBI) as well as MRMI gain, MRMI e±ciency, MBI gain, and MBI e±ciency were analyzed. Results: Subjects with admission categories of MFAC 2 and 3 had a highly signi¯cant (p < 0:001) MRMI gain (6.2 and 6.6, respectively) and subjects with admission categories of MFAC 3 to 5 had highly signi¯cant (P < 0:001) MRMI e±ciency (0.34, 0.40, and 0.39, respectively). The subjects with admission categories of MFAC 2 to 5 had a highly signi¯cant (p < 0:001) MBI gain (9.7, 10.2, 9.3, and 7.0, respectively) and the subjects with admission categories of MFAC 4 to 5 had a highly signi¯cant (p < 0:001) MBI e±ciency (0.70 and 0.72, respectively). The subjects with admission categories of MFAC 1 and 2 had a highly signi¯cant (p < 0:001) LOS (27.7 and 26.6, respectively). MFAC pro¯le was also established to represent the distri- bution of discharge MFAC of subjects according to their admission MFAC. The chance of subjects with admission categories of MFAC 1 and MFAC 2 progress to any kind of walker (MFAC > 2) is 12.7% and 58.2%, respectively. The chance of subjects with admission MFAC 3, MFAC 4 and MFAC 5 progress to independent walker (MFAC > 5) is 6.7%, 14.8%, and 50.3%, respectively. Both admission MFAC and Copyright@2018, Hong Kong Physiotherapy Association. Published by World Scienti¯c Publishing Co Pte Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 141

142 B. P. H. Chung admission MBI had strong correlations with discharge MFAC (r ¼ 0:84, P < 0:0001 and r ¼ 0:78, P < 0:0001, respectively), discharge MRMI (r ¼ 0:82, P < 0:0001 and r ¼ 0:78, P < 0:0001, respectively) and discharge MBI (r ¼ 0:78, P < 0:0001 and r ¼ 0:94, P < 0:0001, respectively). Conclusion: This study showed that patients on admission with moderate disability in term of MFAC had the greatest mobility gain and basic activities of daily living (ADL) gain from inpatient stroke rehabilitation. Admission MFAC could be a strati¯cation tool of patients with stroke in inpatient rehabilitation. Keywords: Stroke rehabilitation; physiotherapy; functional outcome; strati¯cation. Hong Kong Physiother. J. 2018.38:141-147. Downloaded from www.worldscientific.com Introduction Methods by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. Stroke, also known as cerebrovascular accident Demographic characteristics and (CVA), is an acute disturbance of focal or global average functional outcomes cerebral function with signs and symptoms lasting more than 24 h or leading to death, presumably of This was a retrospective descriptive study of vascular origin.1 Stroke was the fourth leading patients with principal diagnosis of CVA, stroke, cause of deaths in Hong Kong in 20142 and around or hemiplegia and had received stroke rehabilita- 25,000 of patients with stroke are admitted to tion program in a hospital in Hong Kong between public hospitals under the Hong Kong Hospital the periods of 1st January 2011 to 31st December Authority annually.2 The most widely recognized 2015. Patient's demographic and hospital infor- impairment causing stroke is motor impairment mation including age, gender, premorbid accom- which restricts functional mobility including modation, stroke type, days post stroke, discharge walking.3,4 Therefore, to improve functional mo- destination and LOS were retrieved from the da- bility of patients with stroke is one of the main tabase of physiotherapy department of the hospital goals of rehabilitation.4 However, stroke rehabili- and Clinical Management System of Hong Kong tation in inpatient setting requires high intensity of Hospital Authority. LOS is the total number of manpower and resources. Early strati¯cation of days spent in the rehabilitation program. The de- patients with stroke is useful for clinicians to rec- mographic characteristics and average functional ognize patients' possible functional outcomes, level outcomes of all subjects were demonstrated in of disability and requirement for social support so number and percentage or mean and standard as to facilitate early discharge plan and reduce deviation. avoidable length of stay (LOS) in hospital. Strati- ¯cation in clinical setting is usually based on ad- Clinical outcomes mission functional scores such as Functional Independence Measure (FIMTM),5 Barthel Index The subjects were divided into seven groups (BI),6 Modi¯ed BI (MBI),7 and classi¯cation sys- according to their admission MFAC. The between- tem such as Inpatient Rehabilitation Facility- group di®erences of clinical outcomes were ana- Case-Mix Group (IRF-CMG).8 However, these lyzed. Clinical outcomes including LOS, admission assessment tools are quite time-consuming and re- Modi¯ed Rivermead Mobility Index (MRMI), dis- quire a special training to complete the full score. charge MRMI, MRMI gain, MRMI e±ciency, Using functional scores in terms of simple mobility admission MBI, discharge MBI, MBI gain, and MBI categories to stratify patients with stroke and e±ciency. MRMI gain is the di®erence between predict functional scores of patients with stroke at discharge MRMI and admission MRMI. MRMI discharge is worth exploring. Modi¯ed Functional e±ciency is the average change in total MRMI Ambulation Category (MFAC) is a 7-point Likert ratings per day, which were calculated for each Scale9 which is easy-to-use, inexpensive and com- subject by subtracting admission MRMI from dis- monly used to classify walking capacity of patients charge MRMI ratings and then dividing by the with stroke in Hong Kong10,11 and Korea.12 The LOS measured in days. MBI gain and MBI e±- objective of the study was to explore whether ciency were measured by the same methods. The MFAC can serve as a strati¯cation tool of patients gain of a score indicates the total gain within the with stroke in inpatient rehabilitation. LOS. The e±ciency of a score indicates the daily

Hong Kong Physiother. J. 2018.38:141-147. Downloaded from www.worldscientific.com gain of the score. Both gain and e±ciency of scores Strati¯cation of stroke rehabilitation 143 by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. were included to facilitate comparisons to other studies. (Cronbach's alpha ¼ 0.93) to early stage patients with stroke. The MRMI consists of eight test items, MFAC pro¯le including turning over, changing from lying to sitting, maintaining sitting balance, going from MFAC pro¯le represented the distribution of pos- sitting to standing, standing, transferring, walking sibility of discharge MFAC by subjects' admission indoors, and climbing stairs. The score of MRMI category of MFAC. The results were expressed in a ranges from 0 to 40. One main characteristic of the matrix table and in a rank order for each group MRMI is that participants are scored by observa- of subjects. The possibilities were expressed in tion of their performance on the items directly.16 percentage (%). Modi¯ed Barthel Index Correlation MBI was used to assess subjects' basic activities of Since previous studies showed that functional daily living (ADL) in this study. MBI measures the outcome at admission is positively correlated with subject's performance on 10 functional items in- the functional outcome at discharge,13,14 the cor- cluding self-care, continence, and locomotion. The relation of admission MFAC and admission MBI to values assigned to each item are based on the discharge MFAC, discharge MRMI, and discharge amount of physical assistance required to perform MBI were measured. the task and added to give a total score ranging from 0 to 100 (0 ¼ fully dependent, 100 ¼ fully Modi¯ed Functional Ambulation independent) with higher scores indicating higher Category levels of physical function.11 There are no subtotal score because there is no subscale.11 The internal MFAC is a 7-point Likert Scale (1–7) that is used consistency reliability coe±cient for MBI is 0.90.17 to classify a subject's walking capacity. It was modi¯ed from Functional Ambulation Classi¯ca- Statistical Analysis tion (FAC). FAC includes six ordinal categories (0–5) of support needed for gait, but does not In order to stratify subjects with greatest functional evaluate whether or not an aid was used.15 MFAC gain from the stroke rehabilitation program, the divided gait into seven categories (Category I to subjects were divided into seven groups according to Categories VII), ranging from no ability to walk the admission categories of MFAC. The between- and requires manual assistance to sit or is unable to group di®erences of the characteristics including sit for 1 min without back or hand support (MFAC 1; LOS, admission MRMI, discharge MRMI, admis- Category: I, stage: Lyer) to the ability to walk sion MBI and discharge MBI, MRMI gain, MRMI independently on level and non-level surfaces, e±ciency, MBI gain, and MBI e±ciency were ana- stairs, and inclines (MFAC 7; Category: VII, stage: lyzed by one way analysis of variance (ANOVA), Outdoor walker).9 The inter-rater reliability of the post-hoc Bonferroni's test was administered to iden- MFAC (intraclass correlation coe±cient (ICC)) tify subsets of each group. For each characteristic, was 0.982 (0.971–0.989), with a kappa coe±cient of the groups with relative high score and without 0.923 and a consistency ratio of 94% for stroke within-group post-hoc di®erence were highlighted. patient12 and the ICC of the MFAC in patients with hip fractures is 0.96, with a construct validity In order to compare the correlation of admission of r ¼ 0:81 on the Elderly Mobility Scale (EMS).9 MFAC and admission MBI to functional outcomes of subjects, Spearman's rank correlation coe±cient Modi¯ed Rivermead Mobility Index ðrsÞ was used to test the relationship between admission MFAC and admission MBI to discharge MRMI was used to assess subjects' mobility in this MFAC, discharge MRMI and discharge MBI. Results study. The MRMI is highly reliable between raters were considered statistically signi¯cant when (ICC ¼ 0:98) and has high internal consistency p < 0:05. Data were analyzed with the use of the SPSS À V20 statistical package (SPSS Inc., Chicago, LL).

144 B. P. H. Chung Table 1. Demographic characteristic and average functional outcomes of patients (N ¼ 2; 722). Ethics Statement Demographic characteristic N (%) Mean (SD) Ethics approval was granted by the Joint Chinese Hong Kong Physiother. J. 2018.38:141-147. Downloaded from www.worldscientific.com University of Hong Kong — New Territories East Age — 74.8 (12.24) by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. Cluster Clinical Research Ethics Committee. Gender — 1,433 (52.7) — Results Male 1,289 (47.3) — Female Demographic characteristics and Premorbid MFAC 139 (5.1) — average functional outcomes Category 1 168 (6.2) — Category 2 97 (3.6) — Medical records of 2,722 of 3,085 subjects admitted Category 3 193 (7.0) — to a rehabilitation hospital were analyzed for this Category 4 70 (2.6) — study. A total of 363 subjects were excluded, in Category 5 390 (14.3) — which, 187 subjects were transferred back to acute Category 6 1,665 (61.2) — hospital due to unstable medical conditions, 18 Category 7 subjects were discharged against medical advice Premorbid accommodation 2,333 (85.7) — (DAMA) and 158 subjects died. Of the 2,722 Home 389 (14.3) — subjects, the mean age was 74.6 in which 1,433 Institution (52.7%) subjects were male and 1,289 (47.3%) Stroke type (%) 2,312 (84.9) — subjects were female. There were 2,333 (85.7%) Cerebral infraction 410 (15.1) — subjects lived at home whereas 389 (14.3%) sub- Cerebral hemorrhage — jects were institutionalized. There were 2,312 First stroke 1,955 (71.8) (84.9%) subjects su®ered from cerebral infarction Recurrent stroke 767 (28.2) 12.5 (7.65) and 410 (15.1%) subjects su®ered from cerebral Days post stroke (day) — 22.3 (18.24) hemorrhage. A total of 1,955 (71.8%) subjects were LOS in rehabilitation — ¯rst-time stroke patients and 767 (28.2%) subjects Discharge destination — — had recurrent stroke. The mean days post stroke from rehab — was 12.5 days and the mean LOS in the rehabili- Home 1,690 (62.1) 13.7 (11.07) tation program was 22.3 days (Table 1). Before Institution 1,032 (37.9) 18.1 (12.83) the admission, 139 (5.1%) subjects were lyers Admission MRMI 4.4 (6.14) (MFAC ¼ 1), 168 (6.2%) subjects were sitters Discharge MRMI — 0.26 (0.43) (MFAC ¼ 2), 97 (3.6%) subjects were dependent MRMI gain — 32.5 (29.19) walkers (MFAC ¼ 3), 193 (7.0%) subjects were MRMI e±ciency — 39.7 (32.33) assisted walkers (MFAC ¼ 4), 70 (2.6%) subjects Admission MBI — 7.2 (12.00) were supervised walkers (MFAC ¼ 5), 396 (14.3%) Discharge MBI — 0.40 (0.77) subjects were independent indoor walkers MBI gain — (MFAC ¼ 6) and 1,665 (61.2%) subjects were in- MBI e±ciency — dependent outdoor walkers (MFAC ¼ 7) — (Table 1). The average MRMI was increased from 13.7 at admission to 18.1 at discharge. The average sitters (MFAC ¼ 2), 439 (16.1%) subjects were MRMI gain was 4.4 and the average MRMI e±- dependent walkers (MFAC ¼ 3), 570 (20.9%) ciency was 0.26 (Table 1). The average MBI was subjects were assisted walkers (MFAC ¼ 4), 190 increased from 32.5 at admission and 39.7 at dis- (7.0%) subjects were supervised walkers charge. The average MBI gain was 7.2 and the (MFAC ¼ 5), 74 (2.7%) subjects were independent average MBI e±ciency was 0.4 (Table 1). indoor walkers (MFAC ¼ 6) and 14 (0.5%) sub- jects were independent outdoor walkers Clinical outcomes (MFAC ¼ 7) (Table 2). The LOS of subjects with admission categories of MFAC 1 and 2 (27.7 and The outcomes among the seven groups of subjects 26.6, respectively) had signi¯cant di®erences with di®erent admission MFAC were shown in (p < 0:05) when compared to subjects with ad- Table 2. On admission, 889 (32.7%) subjects were mission categories of MFAC 3 to 7 (21.8, 16.1, 10.9, lyers (MFAC ¼ 1), 546 (20.6%) subjects were 8.3, and 6.1, respectively). The MRMI gain of subjects with admission categories of MFAC 2 and 3 (6.2 and 6.6, respectively) had signi¯cant di®erences (p < 0:05) when compared to subjects with admission categories of MFAC 1 and 4 to 7

Strati¯cation of stroke rehabilitation 145 Table 2. Clinical outcomes of subjects. Outcome Admission MFAC P -value 1 (n ¼ 889) 2 (n ¼ 546) 3 (n ¼ 439) 4 (n ¼ 570) 5 (n ¼ 190) 6 (n ¼ 74) 7 (n ¼ 14) LOS 27.7 (22.74) 26.6 (17.52) 21.8 (14.36) 16.1 (11.47) 10.9 (7.08) 8.3 (7.54) 6.1 (4.94) P < 0:001* Admission MRMI 2.0 (3.12) 10.2 (4.26) 15.7 (4.28) 23.4 (5.55) 30.5 (4.54) 37.2 (2.59) 37.1 (8.74) P < 0:001* Discharge MRMI 4.7 (7.05) 16.4 (8.37) 22.3 (7.96) 28.1 (6.71) 33.9 (4.91) 38.2 (2.46) 37.5 (7.42) P < 0:001* Admission MBI 5.6 (10.98) 25.9 (18.91) 41.3 (20.14) 54.3 (22.03) 71.6 (18.78) 80.6 (21.36) 82.3 (23.51) P < 0:001* Discharge MBI 8.6 (15.60) 35.6 (23.80) 51.5 (22.50) 63.6 (22.56) 78.6 (18.07) 84.3 (20.67) 85.5 (24.21) P < 0:001* MRMI gain 2.7 (5.91) 6.2 (6.97) 6.6 (6.86) 4.8 (4.91) 3.4 (3.90) 1.0 (1.69) 0.4 (1.34) P < 0:001* MRMI e±ciency 0.11 (0.35) 0.26 (0.34) 0.34 (0.41) 0.40 (0.53) 0.39 (0.51) 0.15 (0.26) 0.05 (0.17) P < 0:001* MBI gain 3.0 (8.94) 9.7 (14.84) 10.2 (12.81) 9.3 (12.10) 7.0 (8.51) 3.7 (6.11) 3.2 (4.76) P < 0:001* MBI e±ciency 0.10 (0.35) 0.38 (0.70) 0.51 (0.74) 0.70 (1.04) 0.72 (0.96) 0.51 (0.88) 0.52 (0.65) P < 0:001* Hong Kong Physiother. J. 2018.38:141-147. Downloaded from www.worldscientific.com Notes: All scores are reported as a mean (SD). *For each characteristic, one-way ANOVA with admission MFAC as a ¯xed by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. factor, post-hoc di®erences exist between the shaded and non-shaded scores, but no post-hoc di®erence among the shaded scores. (2.7, 4.8, 3.4, 1.0, and 0.4, respectively). The categories of MFAC and the top row lists the dis- MRMI e±ciency of patients with admission cate- charge categories of MFAC. At the intersection of gories of MFAC 3 to 5 (0.34, 0.40, and 0.39) had the rows and columns, the possibility in percentage (%) signi¯cant di®erences (p < 0:05) when compared to of discharge categories of MFAC of subjects in a subjects with admission categories of MFAC 1, 2, rank order for each admission categories of MFAC 6, and 7 (0.11, 0.26, 0.15, and 0.05, respectively). of subjects could be identi¯ed. For example, the The MBI gain of subjects with admission categories change of subjects with admission category of of MFAC 2 to 5 (9.7, 10.2, 9.3, and 7.0) had sig- MFAC 2 (sitter) progress to dependent walkers, ni¯cant di®erences (p < 0:05) when compared to assisted walkers, supervised walkers, indoor walk- subjects with admission categories of MFAC 1, 6, ers and outdoor walkers upon discharge is 23.9%, and 7 (3.0, 3.7, and 3.2, respectively). The MBI e±- 22.2%, 10.1%, 2%, and 0%, respectively (Table 3). ciency of subjects with admission categories of The chance of subjects with admission categories of MFAC 4 and 5 (0.70 and 0.72) had signi¯cant dif- MFAC 1 and 2 progress to any kind of walker (MFAC ferences (p < 0:05) when compared to subjects with 3 to 7) is 12.7% and 58.2%, respectively. The chance of admission categories of MFAC 1 to 3 and 6 to 7 (0.10, subjects with admission categories of MFAC 3 to 5 0.38, 0.51, 0.51, and 0.52, respectively) (Table 2). progress to independent walker (MFAC 5 to 7) are 6.7%, 14.8%, and 50.3%, respectively. MFAC pro¯le Correlation FAC pro¯le is a matrix table representing the dis- tribution of discharge MFAC of subjects according Both admission MFAC and admission MBI had to admission category of MFAC (Table 3). The left- strong correlations with discharge MFAC hand column of this pro¯le lists the admission (r ¼ 0:84, P < 0:0001 and r ¼ 0:78, P < 0:0001, Table 3. MFAC pro¯le of subjects (n ¼ 2;722). Discharge MFAC (%) 3 Dependent 4 Assisted 5 Supervised 6 Indoor 7 Outdoor walker walker 1 Lyer 2 Sitter walker walker walker Admission MFAC 1 Lyer 74.8 12.5 6.7 5.1 0.8 0.1 0 2 Sitter 2.8 39.0 23.9 22.2 10.1 2.0 0 3 Dependent Walker 0.4 4.2 34.4 34.9 19.4 6.5 0.2 4 Assisted Walker 0.2 0.7 1.4 43.4 39.5 13.0 1.8 5 Supervised Walker 0 0.5 0 0.5 48.7 42.3 8.0 6 Indoor Walker 00 0 0 0 77.5 22.5 7 Outdoor Walker 0 0 0 0 0 0 100

146 B. P. H. Chung Table 4. Spearman's rank correlation functional gain (16.6þ/À11.7) than those who scored of 36 `and functional gain of (27.6þ/ coe±cients (rs) of functional score at À23.3). When considering the functional gain, admission and discharge. present study still showed that the best MRMI gains (6.6) and best MBI gain belonged to the Discharge patients with moderate disability, i.e., admission MFAC MRMI MBI category of MFAC 3, but not the most severely disabled patients, i.e., MFAC 1 and 2. Admission 0.84* 0.82* 0.78* MFAC 0.78* 0.78* 0.94* The present study also showed that MFAC and MBI MBI at admission had similar correlation to MFAC, MRMI, and MBI of patients with stroke Notes: *p < 0:0001. upon discharging from inpatient rehabilitation program. The admission MFAC had a strong cor- Hong Kong Physiother. J. 2018.38:141-147. Downloaded from www.worldscientific.com respectively), discharge MRMI (r ¼ 0:82, P < relation (rs ¼ 0:84, P < 0:0001) with discharge by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. 0:0001 and r ¼ 0:78, P < 0:0001, respectively) and MFAC. This ¯nding echoed with previous studies discharge MBI (r ¼ 0:78, P < 0:0001 and r ¼ 0:94, which found that functional score at admission is P < 0:0001, respectively) (Table 4). positively correlated with the functional score at discharge.13,14 However, how to apply that strong Discussion correlation between admission and discharge MFAC and in clinical situation is a challenge. MFAC is easy to be scored by clinicians and to be Hence, the present study developed the MFAC understood by patients and their families. To our pro¯le which makes use of admission mobility level best knowledge, this is the ¯rst study to demon- to estimate various possibility of discharge mobil- strate that admission MFAC could be a strati¯- ity level of patients with stroke. For instance, the cation tool for patients with stroke in inpatient chance of patients with admission categories of rehabilitation. The present study showed that the MFAC 1 and 2 progress to any kind of walker groups of patients who had the greatest bene¯t (MFAC 3 to 7) is 12.7% and 58.2%, respectively. from stroke rehabilitation program were those with Therefore, the sitter has 4.5 times of chance to walk admission categories of MFAC 3 to 5 in terms of again upon discharge than the lyer. The chance of MRMI e±ciency and those with admission cate- patients with admission categories of MFAC 3 to 5 gories of MFAC 4 to 5 in terms of MBI e±ciency. progress to independent walker (MFAC 5 to 7) is The present study echoes with Louie's study7 that 6.7%, 14.8% and 50.3%, respectively. The possible patients with stroke with moderate disability on mobility level of patient upon discharge (discharge admission presented the best rate of improvement MFAC) from stroke rehabilitation program is in functional outcomes on discharge. Under re- useful information for discharge planning with source constraints, stroke rehabilitation teams patients and their family. have to stratify patients with stroke as soon as possible so as to allocate suitable resources, such as The present study had its limitations. First, therapy sessions or LOS in hospital, to patients sampling bias may exist due to all subjects which with stroke. Therefore, according to the result of were recruited from only one local rehabilitation present study, the rehabilitation team could allo- hospital. Further studies are suggested to verify cate more resources to patients with moderate the pattern of MFAC pro¯le and functional out- disability on admission, and formulate early dis- comes strati¯ed by admission MFAC of patients charge plan for severe and mild disability. How- with stroke in di®erent setting, di®erent phases, ever, our ¯nding was a little di®erent from previous and di®erent countries. studies by Gagnon and colleagues5 and Lin.8 Gagnon and colleagues5 showed that the best total Conclusion (41.6) and motor-FIM (38.9) gains were observed in most severely disabled patients with stroke This study showed that patients on admission (IRF-CMG classi¯cation system: 114). In addition, with moderate disability in terms of MFAC had Lin8 showed that patients with stroke admission the greatest mobility gain and basic ADL gain FIM total scores of ! 73 were scored lower from inpatient stroke rehabilitation. Admission

Hong Kong Physiother. J. 2018.38:141-147. Downloaded from www.worldscientific.com MFAC could be a strati¯cation tool of patients Strati¯cation of stroke rehabilitation 147 by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. with stroke in inpatient rehabilitation. 6. Nakao S, Takata S, Uemura H, et al. Relationship Con°ict of Interest between Barthel Index scores during the acute phase of rehabilitation and subsequent ADL in The author declares that he has no ¯nancial stroke patients. J Med Invest 2010;57(1–2):81–8. a±liations (including research funding) or in- volvement with any commercial organization that 7. Louie SW, Wong SK, Wong CM. Pro¯les of func- has a direct ¯nancial interest in any matter tional outcomes in stroke rehabilitation for Chinese described in this manuscript. The author has no population: A cluster analysis. NeuroRehabilita- other ¯nancial or non¯nancial con°icts of interest tion 2009;25(2):129–35. related to any matter in this study. 8. Lin JH. In°uence of admission functional status on Funding/Support functional gain and e±ciency of rehabilitation in ¯rst time stroke patients. Kaohsiung J Med Sci There were no sources of funding or grant support 2001;17(6):312–8. for this study. The author would like to thank the colleagues of the Physiotherapy Department of Tai 9. Chau MWR, Chan SP, Wong YW, et al. Reliability Po Hospital for their support and contributions. and validity of the modi¯ed functional ambulation classi¯cation in patients with hip fracture. Hong Author Contributions Kong Physiother J 2013;31(1):41–4. Bryan Ping Ho Chung contributed to the design, 10. Chung BPH. The e®ectiveness of robotic-assisted data collection, data analysis and manuscript gait training in stroke rehabilitation: A retrospec- writing of the study. tive matched control study. Hong Kong Physiother J 2017;36:10–6. References 11. Chung BPH. E®ect of di®erent combination of 1. World Health Organization. Recommendations on physiotherapy treatment approaches on functional stroke prevention, diagnosis, and therapy: Report outcomes of patients with stroke: A retrospective of the WHO task force on stroke and other cere- analysis. Hong Kong Physiother J 2014;32 brovascular disorders. Stroke 1989;20:1407–31. (1):21–7. 2. Hospital Authority. Hospital Authority Statistical 12. Park CS, An SH. Reliability and validity of the Report 2014;2012–13. modi¯ed functional ambulation category scale in patients with hemiparalysis. J Phys Ther Sci 3. Langhorne P, Coupar F, Pollock A. Motor recovery 2016;28(8):2264–7. after stroke: A systematic review. Lancet Neurol 2009;8:741–54. 13. Joa KL, Han TR, Pyun SB, et al. Inpatient stroke rehabilitation outcomes in Korea derived from the 4. Bohannon RW, Horton MG, Wikholm JB. Impor- Korean brain rehabilitation centers' online data- tance of four variables of walking to stroke base system for the years 2007 to 2011. J Korean patients. Int J Rehabil Res 1991;14(3):246–50. Med Sci 2015;30(5):644–50. 5. Gagnon D, Nadeau S, Tam V. Clinical and ad- 14. Inouye M, Kishi K, Ikeda Y. Prediction of func- ministrative outcomes during publicly-funded in- tional outcome after stroke rehabilitation. Am J patient stroke rehabilitation based on a case-mix Phys Med Rehabil 2000;79(6):513–8. group classi¯cation model. J Rehabil Med 2005;37 (1):45–52. 15. Holden MK, Gill KM, Magliozzi MR, Nathan J, Piehl-Baker L. Clinical gait assessment in the neurologically impaired. Reliability and meaning- fulness. Phys Ther 1984;64(1):35–40. 16. Lennon S, Johnson L. The Modi¯ed Rivermead Mobility Index: Validity and reliability. Disabil Rehabil 2000;22:833–9. 17. Shah S, Vanclay F, Cooper B. Improving the sen- sitivity of the Barthel Index for stroke rehabilita- tion. J Clin Epidemiol 1989;42:703–9.

Research Paper Hong Kong Physiotherapy Journal Vol. 38, No. 2 (2018) 149–160 DOI: 10.1142/S1013702518500130 Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com Hong Kong Physiotherapy Journal by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. http://www.worldscientific.com/worldscinet/hkpj Comparative e®ectiveness of transverse oscillatory pressure and cervical traction in the management of cervical radiculopathy: A randomized controlled study Adesola Ojo Ojoawo* and Ayodele Damilare Olabode Department of Medical Rehabilitation, Faculty of Basic Medical Sciences College of Health Sciences, Obafemi Awolowo University, Ile Ife, Nigeria *[email protected] Received 14 September 2016; Accepted 24 October 2017; Published 14 August 2018 Background: Radiating neck pain is one of the major symptoms of cervical radiculopathy (CR). Objective: This study compared the e®ects of cervical traction (CT) and transverse oscillatory pressure (TOP) in management of CR. Methods: Seventy-¯ve participants with unilateral radiating neck pain were randomly allocated into three groups, 25 (14 males, 11 females) for CT, 25 (15 males and 10 females) for TOP and 25 (11 males and 14 females) control (Cnt) group. All participants received massage, cryotherapy and active exercises three times in a week for six weeks. CT was administered to CT group, TOP to TOP group while the third group served as control. Pain intensity (PI) and neck functional disability (NFD) were assessed pretreatment, 3rd and 6th week of intervention. Data were analyzed using descriptive and inferential statistics. Results: There was a signi¯cant reduction in PI and NFD between pretreatment and 6th week in all the groups (p < 0:05). The e®ect size of PI (F ¼ 7:533, p < 0:001) and disability index (F ¼ 37:888, p < 0:001) in CT group were signi¯cantly lower than that of TOP group at 3rd week. PI of TOP was signi¯cantly (p < 0:05) lower than that of CT and Cnt groups at the 6th week. Conclusion: TOP reduces the PI and disability of patients with CR faster compared to CT. Keywords: Cervical traction; transverse oscillatory pressure (TOP); cryotherapy; neck disability index; visual analogue scale. *Corresponding author. Copyright@2018, Hong Kong Physiotherapy Association. Published by World Scienti¯c Publishing Co Pte Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 149

Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com 150 A. O. Ojoawo & A. D. Olabode With respect to the management of CR, Cost- by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. ello17 observed that conservative treatment is more Introduction e®ective than surgical options. Conservative treatment for CR typically includes therapeutic Radiating neck pain, one of the major symptoms of exercise (range of motion, strengthening), manual cervical radiculopathy (CR), though less common therapy (muscle energy techniques, non-thrust than non-radiating neck pain has constituted an mobilization, manipulation), modalities (cryother- important cause of disability; therefore, it is im- apy, traction), massage therapy, medication and perative to discover the best way to manage it.1–6 cervical collar.18–21 From the empirical observa- In addition to neck pain, other most common tions of Maitland,22 transverse oscillatory pressure complaint of individuals with CR are paresthesia (TOP) which is one of the manipulative techniques and radicular pain and while sensory manifestation was recommended for unilaterally distributed can be dermatomal, the expression of pain may be symptoms of cervical origin. TOP, originated by myotomal.7 Patterns of dematomal pain is com- Nwuga,23 although one of the frequently used mon at C4 level followed by C6 and then C7 and manipulative techniques by physiotherapists, has scapular pain may occur in 51.6%, pain at peri- been claimed to be e®ective in amelioration of pain scapular region and in the upper limb, as well as intensity (PI) especially radiating pain in cervical, neurological signs such as numbness, weakness and thoracic and lumbar regions.23,24 It involves mo- loss of re°exes in the a®ected nerve root distribu- bilization of the spinous process of the vertebrae in tion.8 Painful range of motion and reduced tendon the region of the spine that had mechanical pain.24 re°ex are typically found on the course of exami- This technique was reported to be useful when pain nation with more than 10% having upper limb has a unilateral distribution, whether localized to weakness and one-third may present with de- the neck or referred to the upper limb.24 creased sensation, and muscular atrophy may be present in less than 2%.9 The major causes of CR Cervical traction (CT) consists of administering are discogenic and spondylitis combined, which a distracting force to the neck in order to separate form 68%, while 22% of cases were from interver- the cervical segments and relieve compression of tebral disc.9 nerve roots by intervertebral disks. Various tech- niques (supine versus sitting; intermittent versus The annual incidence report of CR was 83.2 per sustained; motorized or hydraulic versus an over- 100 000 and an increased prevalence in the ¯fth the-door pulley with weights) and durations decade of life among the general population. Ap- (minutes versus up to an hour) have been recom- proximately 14–71% of adults experience neck pain mended for the management of CR.25 According to at some points in their lifetime and the one-year Shirai,26 CT increases blood °ow to neck muscles prevalence rate for neck pain in adults ranges from 2 min after it is applied. A systematic review by 16% to 75%.10 Study has found a prevalence of Graham et al.27 also reported that there was neck pain of 53.6% among learners in the Gauteng moderate evidence to support the use of mechani- Province, South Africa.11 Also, researchers have cal intermittent CT in the management of cervical documented that neck pain was found to be com- disorder. Ojoawo et al28 in their study reported mon among Nigerian university undergraduate and that CT is e®ective in the management of CR. CT a®ects females than males.12 In the south western in addition to other exercises is the major treat- part of Nigeria, it has been documented that the ment technique in many facilities in Nigeria phys- leading work-related musculoskeletal disorder was iotherapy clinic to manage CR but there are low back pain, followed by neck pain among nurses paucity of data on its e±cacy in Nigerian envi- and physiotherapist which is an indication that ronment. More so, TOP though e®ective was not a neck pain is very prominent among musculoskele- common practice in Nigerian physiotherapy clinic tal pain in Nigeria.13,14 because of the skill required. The question is, that does TOP and CT have the same result in the The etiology of neck pain though multifactorial management of CR? CT requires kit which in some and poorly understood has been linked to factors facilities may not be readily available and TOP like poor posture, depression, anxiety, aging, acute demands special skills which were not known by all injury and occupational or sporting activities.15,16 physiotherapists. If TOP cannot be applied, will This leads to altered joint mechanics, muscle CT give the desired result? The purpose of the structure or function resulting into mechanical neck pain. Researchers also reported that the most common cause of mechanical neck pain is zygo- physeal joint locking and muscle strain.16

Comparative e®ectiveness of TOP and CT in cervical radiculopathy 151 Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com study was to compare the e®ects of TOP and CT in S ¼ within group standard deviation, d ¼ by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. the management of CR. expected di®erence between means within the group. With respect to the study of Ojoawo et al.,30 Methods S ¼ 2:98 in one of the groups and d was 2.01. The equation sample size is 22 (2.98)2/(2.01)2 þ 1 ¼ 39 The participants for this study were 75 (40 males, for each group. The total number for the three 35 females) individuals referred for physiotherapy groups should be 116. The study lasted for a period treatment at the Obafemi Awolowo University of two years and three months, the total number of Teaching Hospital Complex, Ile-Ife, Nigeria, with patients with CR ful¯lled the requirement for that CR in either right or left upper limb. They were period was 87 and all of them were considered for recently diagnosed patients from the Orthopaedic the study. However, only 75 (86.2%) patients were Clinic of the same hospital. All patients' reported able to participate in the study. neck pain that radiated distally down the right or left arm to the elbow. Fifty patients had pain radiating to right while 25 patients reported radiation toward the left Sample Size Determination upper limb. Their pain started 6–8 weeks before the commencement of the study. None of the Sample size equation 22S 2=d2 þ 1 for calculation patients could remember any pathology that pre- of study with comparison of three groups according cipitated his/her complaint. There were various to Dallal29 is adopted for the study. descriptions of pains with a greater percentage which described the pain as a deep ache in the neck and a peppery sensation into either of the arms Enrolment Assessed for eligibility (n=87) Excluded (n= 12) Not meeting inclusion criteria (n=8) Declined to participate (n= 2) Other reasons (n= 2) Randomized (n= 75) ALLOCATION Cervical Traction Group (n= 25) Transverse Oscillatory Pressure TOP Group (n= Exercise and Massage only group Received cervical traction, exercise, ice 25) Received TOP, exercise, ice therapy and (n=25) Received exercise, ice therapy and massage massage therapy and massage Follow-Up Lost to follow-up (give reasons) (n=1) Lost to follow-up (give reasons) (n= 0) Lost to follow-up (give reasons) (n= 2) Discontinued intervention (n=1) Discontinued intervention (n=0) Discontinued intervention (n= 2) Analyzed (n= 24) Analysis Analyzed (n= 23) Excluded from analysis (n=1) Excluded from analysis (n= 2) Analyzed (n=25) Excluded from analysis) (n= 0) Fig. 1. Consort diagram of random allocation of subjects in to three groups.

Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com 152 A. O. Ojoawo & A. D. Olabode commonly anchored by \\no pain\" (score of 0) and by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. \\pain as bad as it could be\" or \\worst imaginable that was aggravated by activities at work or even pain\" (score of 10; 100-mm scale).34 Each partici- at rest. Fifteen patients reported numbness of some pant was asked to point to the number corre- ¯ngers in addition to the radiating pain. Patients sponding to the PI, which was recorded. with unilateral radiating neck pain that was not of mechanical origin and patients with recent major The NDI utilized in the study was in English trauma or fracture of the cervical spine. Patients Language. NDI is a commonly utilized outcome whose primary complaint was that of headaches or measure to capture perceived disability in patients facial pain associated with unilateral radiating with neck pain.35 The NDI contains 10 items: seven neck pain, and any patient who had received related to activities of daily living, two related to manual therapy of the cervical region in the past pain and one related to concentration.36 Each item three months were excluded from the study. Each is scored from 0 to 5, and the total score is participant's blood pressure (mmHg), height (m) expressed as a percentage, with higher scores cor- and weight (kg) were measured. Active range of responding to greater disability.36 The NDI has motion of the neck-elicited pain, especially lateral demonstrated moderate test–retest reliability and °exion toward the side of radiculopathy. Skin has been shown to be a valid health outcome rolling test according to Bansevicius and Pareja31 measure in a patient population with CR. West- and posterior–anterior pressure according to away et al.37 identi¯ed the minimum detectable Egwu32 to the cervical region provoked pain be- change as 5 (10% points) in a group of 31 patients tween the fourth cervical and seventh cervical with neck pain. Stratford and colleagues38 identi- vertebrae in all patients. Spurling's distraction and ¯ed the minimal detectable change also to be 5 Valsalva tests were carried out according to Konin (10% points) in a group of 48 patients with neck et al.33 and were found positive. Individuals found pain and arm pain. suitable for the study were randomly allocated into three groups. Seventy ¯ve pieces of paper with in- Interventions scription of CT, TOP and control group (Cnt) on 25 each were put in an opaque envelope. Each Exercise therapy participant was asked to pick one and such patient was allocated as inscribed into the paper. Each During each appointment, participants in the three participant was randomly allocated into CT, TOP groups underwent exercises. All participants per- and Cnt group. CT group had 24 participants (13 formed cervical spine retraction, rotation in each males 11 females) TOP group with 25 participants direction especially away from the direction of (15 males and 10 females) while Cnt group had 23 pain, extension and side-bending stretching exer- (11male and 12 females) as shown in Fig. 1. cises on the side that is not painful. Stretching exercises were applied with the aim of releasing any Each participant in the three Groups was trea- contracture that might set in by virtue of unilateral ted two times per week. The maximum experi- °exion or side rotation away from the pain. The mental treatment period for a participant was six goal of exercises was to improve endurance of the weeks, after which the treatment time was cervical deep neck muscles to cause the muscles to estimated from the patient's record. stop exhibiting painful response and to improve strength according to Ylinen et al.39 Although This amounted to 12 treatment sessions for each Ylinen et al.39 examined the general musculature of participant in the groups. Present PI and neck the neck, deep neck muscles are part of the general functional disability (NFD) were measured using neck muscles. Any cervical range of motion that visual analogue scale (VAS) and neck disability aggravated pain was avoided during the exercise. index (NDI), respectively. In a case of high PI, cryotherapy was applied ¯rst to relieve the pain which was followed by exercise. Outcome Measures The exercises also included passive stretching and isometric exercises to the posterior neck muscles. Patients were introduced to a 10-point VAS with Isometric exercise was administered according to instructions not to over or underestimate the pain. Kisner and Colby40 to the posterior neck muscles VAS is a continuous scale comprised of a horizon- for 10 s in 10 rounds, in which the contraction was tal or vertical line, usually 10 cm (100 mm) in length, anchored by two verbal descriptors, one for each symptom extreme. The scale is most

Comparative e®ectiveness of TOP and CT in cervical radiculopathy 153 Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com against resistance of the physiotherapist's hand. of pain when pressure is applied on the spinous by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. Ice chips were packed in a towel and applied to the process toward the side of pain. This will restore cervical region for 7 min. This was done to suppress the joint play and mobilize the articular surfaces on muscle spasm caused by the pain. The isometric the painful side.24 The oscillation is done rhyth- exercise, cryotherapy and stretching served as mically with low amplitudes for a period of 20 s. baseline intervention for all the groups. This was repeated three times with a rest period of 2 min for a session per day.24 Cervical traction Treatment was administered twice per week for The CT group was given CT using the \\over the 6 weeks and cumulated to 12 treatment sessions. door\" CT for 15 min in addition to exercise, ice Each session was followed with isometric exercise therapy and massage. A strap was a±xed under in order to improve the strength of the muscles; the chin of the patient. This chinstrap was then cryotherapy was applied to suppress the muscle connected to a water bag that was held aloft over a spasm and stretching of the neck muscles with the doorway via pulleys that were utilized. The water aim of preventing and releasing any muscular bag was loaded in kilogramme to 10% of the contracture that may be setting in. Patient re- patient's total body weight according to Akinbo sponse was assessed after each second treatment et al.41 The 10% weight administered was the session using VAS and NDI. Patients were advised minimum weight; this was increased gradually not to involve in any other intervention without according to the tolerance of each patient to the consulting the corresponding author of this paper. extent to which there was a desired pull at the cervical region. Treatment was administered twice Data Analysis per week for 6 weeks making 12 treatment sessions. Each session was followed with isometric exercise, SPSS version 17.0 (SPSS Inc., Chicago, IL, USA) cryotherapy and stretching as baseline treatment was used to analyze the data. Descriptive and in- to improve the strength release contracture and ferential statistics were used to summarize the re- suppress muscle spasm. Patient response was sult. Analysis of variance (ANOVA) was used to assessed after each second treatment session using compare each of the physical characteristics of the VAS and NDI. participants in the three groups. Repeated mea- sures ANOVA was used to compare the pretreat- Transverse oscillatory pressure ment, 3rd week and 6th week values of PI and NDI of participants in CT, TOP and Cnt within the TOP was administered with the patient lying groups. Repeated measures ANOVA was also used prone on a couch with the forehead placed on the to compare pre-treatment, 3rd week and 6th week backs of her ¯ngers. Standing on the side of the values of the outcome measures among the groups. patient, the therapist placed the pad of the thumbs Post hoc analysis using Tukey's highest signi¯cant against the left side (or the right side depending on di®erence was carried out to examine which vari- the location of the pain) of the spinous process of ables were signi¯cantly di®erent from each other. the vertebrae to be moved. The ¯ngers are spread An alpha level of 0.05 was set as level of signi¯cant. out on the neck and the upper thoracic region. With respect to the subjects that could not com- Pressure is directed horizontally through the plete the six weeks, the principle of intention to thumbs to the side of the spinous process. TOP is treat was applied and the last observation carried executed by a pressure-relaxed sequence on the forward (LOCF) method was used for the analysis. spinous process. Movement is initiated from the trunk and transmitted down the arm to the Results thumbs.24 Treatment is a®ected by a push–relax sequence on the spinous process using the thumbs Figure 1 is the consort °owchart of the partici- to produce an oscillatory movement. Transverse pants. Eighty-seven were recruited for the study pressure was directed toward the side of pain on but twelve participants did not meet the inclusion the cervical vertebrae. The rationale of performing criterial. Seventy-¯ve were randomly allocated to transverse pressure toward painful side is in order three groups but a candidate dropped out from CT to rotate the body of vertebrae away from the side group and two participants were not able to complete the study in the Cnt group. Based on the

Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com 154 A. O. Ojoawo & A. D. Olabode shown in Table 3. The mean di®erence between the by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. pretreatment and 3rd week PI for CT group was premise that those who did not complete the study 1.04 and between the 3rd and sixth week was 3.33 had report of their PI and disability, and were able (F ¼ 7:355; p < 0:001). The di®erence in disability to complete three weeks of treatment, they were between the pretreatment and 3rd week was 12.63, also included in the study. Indicating that 75 par- and between the 3rd week and 6th week was 9.50. ticipants were analyzed for the study. The di®erence of TOP on PI and disability was greater than that of CT at the 3rd week and 6th Physical characteristics of all the participants week except the PI at 6th week which was less were shown in Table 1. There were no signi¯cant (F ¼ 23:156, p < 0:001). The observed power using di®erence (p > 0:05) among the physical char- post hoc power analysis was 1 indicating that acteristics of participants in all the groups. Infer- though the response rate of the subjects was ring that the three groups' physical parameters 86.2%, the sample size was enough to give a reliable were comparable and that the results obtained e®ect size. from the study were as a result of the intervention not based on variation in the groups' physical In Table 4, the repeated measures ANOVA with characteristics. Post Hoc Turkey Highest Signi¯cance Di®erence comparing the mean values of PI among the three In Table 2, the repeated measures ANOVA groups' pretreatment, 3rd week and 6th week is comparing the e®ect of TOP, CT and exercise only shown. There was no signi¯cant di®erence on the PI and neck disability of participants at pre- (p > 0:05) among the mean values of pretreatment treatment, 3rd week and 6th week of treatment is pain intensity in the three groups. It can be in- presented. There was a signi¯cant di®erence ferred from this that the pretreatment PI among (p < 0:05) among the pretreatment, 3rd week and the three groups was comparable, therefore, any 6th week of PI and neck disability of participants result obtained from the study is due to the in all the groups. The mean di®erence among the outcome mea- sures in the three groups and the observed power is Table 1. Physical characteristics of participants (N ¼ 75). CT group n ¼ 25 TOP group n ¼ 25 Crt group n ¼ 25 Variables Mean Æ SD Mean Æ SD Mean Æ SD F P Age (yrs) 51.38 Æ 6.545 55.67 Æ 5.35 59.50 Æ 2.646 3.01 0.08 Weight (kg) 73.13 Æ 13.010 73.00 Æ 5.36 71.25 Æ 5.377 0.05 0.95 Height (m) 1.63 Æ 0.12 1.66 Æ 0.10 1.65 Æ 0.026 0.09 0.92 BMI (Kg/m2) 27.99 Æ 7.96 26.83 Æ 4.437 26.02 Æ 2.041 0.15 0.86 Sex: M F 14 15 11 11 10 14 Note: BMI ¼ Body Mass Index. Table 2. Repeated measures ANOVA comparing the mean values of PI and neck dis- ability of the three groups (N ¼ 75). Pretreatment WK3 WK6 Mean Æ SD Variables Mean Æ SD Mean Æ SD F P CT Grp PI 6.87 Æ 0.99 5.83 Æ 1.64 2.50 Æ 0.53 7.533 0.001** ND 42.13 Æ 16.86 29.50 Æ 17.88 20.00 Æ 17.82 37.881 0.001** TOP Grp PI 7.63 Æ 2.98 4.83 Æ 0.75 2.66 Æ 0.81 23.156 0.001** ND 58.66 Æ 8.91 39.00 Æ 17.46 16.33 Æ 9.75 40.352 0.001** Cntr Grp PI 7.00 Æ 0.81 6.25 Æ 0.95 3.75 Æ 0.53 24.540 0.001** ND 55.32 Æ 11.30 33.82 Æ 1.67 21.50 Æ 5.00 34.40 0.001** Note: **signi¯cant at P < 0:001, WK ¼ week, Grp ¼ Group; Cntr ¼ Control.

Comparative e®ectiveness of TOP and CT in cervical radiculopathy 155 Table 3. Observed power and test between subjects' e®ects within the three groups (N ¼ 75). Groups PnDif Wk1–3 PnDif Wk3–6 F P DisDif Wk1–3 DisDif Wk3–6 F P ObP CT 1.04 3.33 7.533 0.001 12.63 9.50 37.881 0.001 1.000 TOP 2.80 22.67 40.352 0.001 1.000 Cntr 0.78 2.17 23.156 0.001 19.66 12.32 34.400 0.001 1.000 2.50 24.540 0.001 21.50 Note: PnDifWk ¼ Pain di®erence between weeks 1 and 3, Dis ¼ Disability, ObP ¼ Observed Power. Table 4. Repeated measures ANOVA with post hoc comparing the mean values of PI of the three groups (N ¼ 75). Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com CT Grp TOP Grp Cntr Grp by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. Mean Æ SD Mean Æ SD Mean Æ SD F P Pretreatment 6.87 Æ 0.99a 7.63 Æ 2.98a 7.00 Æ 0.81a 0.688 0.518 WK3 5.83 Æ 1.64a 4.83 Æ 0.75b 6.25 Æ 0.95a 7.800 0.005* WK6 2.50 þ 0.53c 1.66 Æ 0.81d 3.75 Æ 0.53e 10.121 0.002* Notes: Superscript letters a–e in the table mean that a mean mode of the same superscript letters indicates no signi¯cant di®erence but mean mode with di®erent superscript letters indicates signi¯cant di®erence. intervention, and not of variation from the pre- TOP at the 6th week (2p ¼ 1:6) followed by that of treatment PI. In the CT and Cnt groups, there was CT and Cnt at 6th week (2p ¼ 1:2). The magni- no signi¯cant di®erence between pretreatment PI tude of e®ect size between TOP and CT group at and 3rd week (p > 0:05) but there exists a signi¯cant the third and 6th week was 0.6 and 0.4, respec- di®erence (F ¼ 7:80, p < 0:05) in the pretreatment tively. An indication is that TOP reduces PI faster and 3rd week in TOP group inferring that PI may be than CT. ameliorated after 2nd week of intervention of TOP. Considering the time e®ect, TOP proves to be faster Table 6 shows the repeated measures ANOVA in relieving PI. Nonetheless, there was a signi¯cant and post hoc comparison of the disability index di®erence (p < 0:05) in the 6th week PI among the among the three groups and Table 7 shows the three groups with TOP group having the minimum magnitude of the e®ect size. There was a signi¯cant values, interpreting that TOP can relieve PI quicker di®erence among (F ¼ 4:08; p < 0:05) the three and more than either CT or exercise only. groups at third week, there was a signi¯cant dif- ference between CT and TOP, and Cnt and TOP Table 5 revealed the magnitude of the e®ect size but not between CT and Cnt in the 6th week. The in the three groups using partial Eta square. The e®ect size of TOP and Cnt at the 6th week was the highest magnitude is between the Cnt group and highest among the others (2p ¼ 0:4). Table 5. Magnitude of e®ect size using partial Eta square for PI among the three groups for 3rd and 6th weeks. Variables M1 M2 M1 À M2 SD1 SD2 SD1 þ SD2 PETA (2pÞ CT and TOP 3rd WK 5.83 4.83 1.00 1.64 0.75 2.39 0.40 2.59 0.20 CT and Cnt 3rd WK 5.83 6.25 0.42 1.64 0.95 1.70 0.80 1.34 0.60 TOP and Cnt 3rd WK 4.83 6.25 1.42 0.75 0.95 1.06 1.20 1.34 1.60 CT and TOP 6th WK 2.50 1.66 0.84 0.53 0.81 CT and Cnt 6th WK 2.50 3.75 1.25 0.53 0.53 TOP and Cnt 6th WK 3.75 1.66 2.09 0.81 0.53

156 A. O. Ojoawo & A. D. Olabode Table 6. Repeated measures ANOVA with post hoc comparing the mean values of neck disability index of the three groups (N ¼ 75). CT Grp TOP Grp Cntr Grp F P Mean Æ SD Mean Æ SD Mean Æ SD Pretreatment 42.13 Æ 16.86a 58.66 Æ 8.91a 55.32 Æ 11.30a 0.125 0.115 WK3 29.50 Æ 17.88b 39.00 Æ 17.46d 33.82 Æ 1.67f 4.870 0.009 WK6 20.00 Æ 17.82c 16.33 Æ 9.75e 21.50 Æ 5.00c 0.118 0.889 Note: Superscript letters a–f in the table mean that a mean mode of the same su- perscript letters indicates no signi¯cant di®erence but mean mode with di®erent superscript letters indicates signi¯cant di®erence. Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com Table 7. Magnitude of e®ect size using partial Eta square for disability among the three groups for 3rd and 6th Weeks. by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. Variables M1 M2 M1 À M2 SD1 SD2 SD1 þ SD2 PETA (2pÞ CT and TOP 3rd WK 39.00 29.50 9.50 17.88 17.46 35.34 0.30 CT and Cnt 3rd WK 33.82 29.50 4.32 17.88 1.67 19.55 0.20 TOP and Cnt 3rd WK 39.00 33.82 5.18 17.46 1.67 19.13 0.30 CT and TOP 6th WK 20.00 16.33 3.67 17.82 9.75 27.57 0.10 CT and Cnt 6th WK 20.00 21.50 1.50 17.82 5.00 22.82 0.10 TOP and Cnt 6th WK 21.50 16.33 5.17 9.75 5.00 14.75 0.40 Notes: M ¼ mean value of disability index; WK ¼ Week; SD ¼ standard deviation; PETA ¼ partial eta. Discussion mobility of the region.24 Also, individual motion segments are thought to be capable of buckling, This study compared the therapeutic e®ect of CT thereby producing relatively large vertebral and TOP in the management of CR. The study motions that achieve a new position of stable revealed that the pretreatment PI, neck disability equilibrium.44 The manipulative impulse provides assessment and other physical characteristics of su±cient energy to restore a buckled segment to a participants in the CT group, TOP and Cnt group lower energy level, thus reducing mechanical stress did not show any signi¯cant di®erence. This is an or strain on soft and hard spinal tissues.45 Giles43 indication that the baseline parameters of the proposed that spinal manipulation activates all participants in all the groups were comparable and known mechanosensitive, somatosensory receptors any result obtained from the study was due to the because they all possess mechanical thresholds intervention. Considering the comparative e®ect of lower than the peak force delivered during a ma- CT and TOP, it was observed that there was a nipulation and the receptor types are responsive to signi¯cant reduction in PI at 3rd week in TOP dynamic and/or static components of a mechanical group than that of CT group. Meanwhile, the mean stimulus. These may be reasons while TOP relieves di®erence between the 3rd week and 6th week of the PI of patient faster. CT group PI was more than that of TOP. This may be inferred that TOP may relieve PI faster but Elnaggar et al.46 in their study reported that CT CT may relieve the PI for longer period of time. methods had a signi¯cant e®ect on neck and arm It has to be emphasized that the technique of TOP pain reduction, a signi¯cant improvement in nerve is the application of oscillatory pressure directly to function and a signi¯cant increase in neck mobility. the speci¯c region where there is pain.24 This may Our ¯nding was also in consistence with the work alter segmental biomechanics by releasing trapped of Voltonen et al.47 who concluded that traction meniscoids, releasing adhesions or by diminishing relieves muscle spasm and signi¯cantly decreases distortion in the intervertebral disc42,43 and re- electrical activity in the muscles and produces re- stored joint play which immediately increases the laxation, which leads to systematic relief of pain.47

Comparative e®ectiveness of TOP and CT in cervical radiculopathy 157 Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com Krause et al.48 found that traction has been shown Researchers have reported that TOP to the spinal by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. to separate the vertebrae, stretch the cervical joint region has both neurological and mechanical capsules, stretch neck muscles and open the e®ects.56,57 foramina. These may be the reasons while CT relieves PI of patients with CR longer. Paris57 mentioned further that mobilization technique stretches tissues by taking them into the It was observed that there was a signi¯cant area of plastic deformation of stress–strain curve. improvement between the pretreatment PI and the TOP has been found as one of the techniques that six-week PI and disability index in CT group. This stretches cervical connective tissues and joint inferred that CT was e®ective in the management capsules to a reasonable point on the stress–strain of CR. Our ¯ndings were in line with the study of curve to produce a salvo of bene¯cial neuro inhib- Borman et al.49 and Cleland et al.20 Borman et al.49 itory and mechanical e®ects.56,57 and Cleland et al.20 in their independent studies using intermittent CT documented that the ap- Considering the participants in the cnt group of plication of CT produced a desired result in the the study, it was revealed that exercise with mas- management of CR. In addition, Rhee et al.50 and sage and cryotherapy also reduced the PI and Swezey et al.25 reported that the application of CT disability of participants. This is an indication that at home was found to decrease radicular symp- pain of CR and NDI can be ameliorated when toms. Levine et al.51 documented further that CT treated with a combination of exercise, massage is most bene¯cial when acute muscular pain has and cryotherapy. The improvement of outcome subsided and should not be used in patients who measures in cnt group is in line with the observa- have signs of myelopathy. CT was known in theory tion of Radhakrishnan et al.9 Studies have exam- to distract the neural foramen and decompresses ined the e®ect of isometric exercise on the the a®ected nerve root.52 Evidence also revealed contracting body part as well as on the contralat- that continuous CT decreases the pressure within eral and a distant body part to the contracting one the vertebral disks and stretches muscles and and a±rmed that the hypoalgesic e®ect of iso- ligaments of the cervical region thereby unloads metric exercise was multisegmental and not iso- the structure of the spine.53 It is probable that lated to the contracting muscle.58,59 Moreover, traction has an important role in breaking the the pain-reducing e®ects of isometric exercise on \\circle of pain\" in CR caused by a herniated disk. the contralateral and distant body parts were This cycle begins when nerve roots are compressed similar in magnitude to the local body part. These by a herniated disk, causing entrapment within the results suggest that a central widespread inhibitory intervertebral foramina. The irritated nerve pro- mechanism is activated by static muscle contrac- duces a re°ex response to the patient's cervical tions. As discussed by Kosek and Lundberg,60 these muscles, causing those muscles to contract. That central mechanisms may include increased secre- contraction further narrows the foramina and the tion of b-endorphins, attention mechanisms, acti- neck pain is increased. Traction helps to relieve the vation of di®use noxious inhibitory controls or an in°ammatory reaction of nerve roots by improving interaction of the cardiovascular and pain regula- the circulation and reducing the tissues swelling. tory systems. Gentle alteration of stretching and relaxation of the neck soft tissue structures prevents the for- The e®ect of kneading massage in this study is mation of adhesions of the dural sleeve.54 These are explained by a researcher who documented that additional factors why CT can reduce the PI of massage has traditionally been used to relieve pain cervical radicular patients longer than TOP. in producing short-lived analgesia by activating the \\pain gate\" mechanism.61 Cutaneous mechan- The study observed in addition that TOP group oreceptors are stimulated by touch and transmit reported signi¯cant decrease in the outcome mea- information within large nerve ¯bers to the spinal sures when the pretreatment mean values were cord.62 These impulses block the passage of painful compared with the posttreatment values. This is stimuli entering the same spinal segment along an indication that TOP has a signi¯cant thera- small, slowly conducting neurons.63 Massage is a peutic value in the management of CR. Research- potent mechanical stimulus and a particularly ef- ers reported that manipulation may provide fective trigger for the pain gate process which can short-term bene¯t in the treatment of neck pain, reinforce a naturally occurring discomfort, cause cervicogenic headaches51 and radicular symptoms.55 much greater release of opiates and achieve more profound pain suppression.64 The contribution of

158 A. O. Ojoawo & A. D. Olabode Con°ict of Interest cryotherapy in the relief of pain has been reported Authors did not have any con°ict of interest on the in studies noting that cryotherapy may be most study. e®ective when combined with exercise.65 Adequate cooling can reduce pain, spasm and neural inhibi- Funding/Support tion, thereby allowing for earlier and more ag- gressive exercises. Cryotherapy can increase pain The study did not enjoy funding from any orga- tolerance and pain threshold and decrease nerve nization or government. conduction velocity.66 Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com Limitations Authors Contributions by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. There are some limitations of this study, one of Ojoawo A.O conceived the idea of the study, ana- them is that the physiotherapist who treated the lyzed the data, interpreted and did the ¯nal write- patients was not blinded to the group allocation up including the proof reading to make it suitable and the assessors were not totally independent of for publication. Olabode A.D collected the the intervention. The reason is that the hospital data, supplied the literature and did the skeletal administration where the study was carried out write-up. did not permit blinding. The study also did not assess the range of motion and strength of the References cervical muscles. The researchers in the proposal did not consider the variables as part of the 1. Rubinstein SM, Pool JJ, van Tulder MW, Ripha- objectives of the study, though it may be an gen II, de Vet HC. A systematic review of the di- omission which has been noted for subsequent agnostic accuracy of provocative tests of the neck study, but the opinion is that ones the pain and for diagnosing cervical radiculopathy. Eur Spine J disability have been addressed, other variables 2007;16:307–19. will fall in line. Researchers were aware of the results of intermittent traction using traction 2. Childs JD, Cleland JA, Elliott JM, et al. Neck pain: machine in the management of CR, but lack of Clinical practice guidelines linked to the interna- funds was a major constrain why this could not tional classi¯cation of functioning, disability, and be used. More so, facilities in Nigeria with CT health from the orthopedic section of the American machine are very limited; it is then imperative Physical Therapy Association. J Orthop Sports that a research is carried out on what is com- Phys Ther 2008;38:A1–34. monly available in the environment of practice. The less favorable results of CT could be due to 3. Tsakitzidis G, Remmen R, Peremans L, et al. Non- continuous traction adopted in the present study. speci¯c neck pain: Diagnosis and treatment. KCE The results may be di®erent if intermittent CT reports. 2009:119C. was given to the patients. Graham et al.21 in the systematic review concluded that intermittent 4. Manchikanti L, Boswell MV, Singh V, et al. Com- traction is better than continuous traction for prehensive evidence-based guidelines for interven- mechanical neck disorders. The short- or long- tional techniques in the management of chronic time follow-up could not be concluded because of spinal pain. Pain Physician 2009;12:699–802. some logistic problems. 5. Da®ner SD, Hilibrand AS, Hanscom BS, Brislin Conclusion BT, Vaccaro AR, Albert TJ. Impact of neck and arm pain on overall health status. Spine It can be concluded from the study that combi- 2003;28:2030–5. nation of exercise, massage and cryotherapy reduces PI and disability of patient with CR in 6. Haldeman S, Carroll L, Cassidy JD, Schubert J, the 6th week of intervention but addition of Nygren A. The bone and joint decade 2000–2010. TOP proved better and may be better than Task force on neck pain and its associated dis- inclusion of CT. orders: Executive summary. Spine 2008;33:S5–7. 7. Slipman CW, Plastaras CT, Palmitier RA. Symp- tom provocation of °uroscopically guided cervical nerve root stimulation. Are dynatomal maps iden- tical to dermatomal maps? Spine 1998;23 (20):2235–42. 8. Murphy D, Hurwitz E, Gregory A. A nonsurgical approach to the management of patients with

Comparative e®ectiveness of TOP and CT in cervical radiculopathy 159 Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com cervical radiculopathy: A prospective observational 23. Nwuga VCB. Techniques of spinal manipulation. by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. cohort study. J Manipulative Physiol Ther 2006;29 In: Nwuga VC, ed. Manipulation of the Spine. (4):279–87. Baltimore, MD: William and Wilkins, 1975:53. 9. Radhakrishnan K, Litchy WJ, O'Fallon WM, Kurland LT. Epidemiology of cervical radiculo- 24. Nwuga VCB. Techniques of spinal manual therapy. pathy: A population-based study from Rochester, In: Manual Treatment of Back Pain. 2nd ed. Minnesota, 1976 through 1990. Brain 1994;117: Baltimore, MD: William and Wilkins, 2007:115. 325–35. 10. Fejer R, Kyvik KO, Hartvigsen J. The prevalence 25. Swezey RL, Swezey AM, Warner K. E±cacy of of neck pain in the world population: A systematic home cervical traction therapy. Am J Phys Med critical review of the literature. Eur Spine J Rehabil 1999;78(1):30–2. 2006;15:834–48. 11. Mafanya C, Rhoda A. Predictors of neck pain 26. Shirai Y. Intermittent cervical traction in subjects among South African youth. Afr J Phys Health with neck and shoulder pain — Analysis of a blood Educ Recreat Dance 2011;18(3):237–42. °ow volume and EMG signals. Sogo-Riha 12. Ayanniyi O, Mbada CE, Iroko OP. Neck pain oc- 1995;23:25–30. currence and characteristics in nigerian University Undergraduates. TAF Prev Med Bull 2010;9 27. Graham N, Gross AR, Goldsmith C. Mechanical (3):167–74. traction for mechanical neck disorders: A systemic 13. Tinubu B, Mbada CE, Oyeyemi AL, Fabunmi AA. review. J Rehab Med 2006;38(3):145–52. Work-related musculoskeletal disorders among nurses in Ibadan, South-west Nigeria: A cross-sec- 28. Ojoawo AO, Olabode A, Esan O, Badru A, Odejide tional survey. BMC Musculoskelet Disord S, Arilewola B. Therapeutic e±cacy of cervical 2010;11:12. traction in the management of cervical radiculo- 14. Adegoke BO, Akodu Ak, Oyeyemi AL. Work- pathy: A control trial. Rwanda J Health Sci 2013;2 related musculoskeletal disorders among Nigerian (2):25–29. physiotherapists. BMC Musculoskelet Disord. 2008;9:112. doi: 10.1186/1471-2474-9-112. 29. Dallal GE. Sample size calculations simpli¯ed. In: 15. Binder AI. Cervical spondylosis and neck pain. The Little Handbook of Statistical Practice. Bos- Br Med J 2007;334(7592):527–31. ton, MA: Jean Mayer USDA Human Nutrition 16. Peterson DH, Bergman TF. Chiropractic Tech- Research Center on Aging at Tufts University, nique: Principles and Procedures. United States of 2012. Available at: The UCLA Department of America: Mosby, 2002. Statistics website. Downloaded on June 25, 2016. 17. Costello M. Treatment of a patient with cervical radiculopathy using thoracic spine thrust manipu- 30. Ojoawo AO, Olabode A, Esan O, Badru A, Odejide lation, soft tissue mobilization, and exercise. J Man S, Arilewola B. Transverse oscillatory pressure in Manip Ther 2008;16:129–35. management of cervical radiculopathy: A rando- 18. Wainner RS, Gill H. Diagnosis and nonoperative mised controlled study. Hong Kong Physiother J management of cervical radiculopathy. J Orthop 2016;34:19–26. Sports Phys Ther 2000;30:728–44. 19. Cleland JA, Fritz JM, Whitman JM, Heath R. 31. Bansevicius D, Pareja JA. The \\skin roll\" test: A Predictors of short-term outcome in people with a diagnostic test for cervicogenic headache? Funct clinical diagnosis of cervical radiculopathy. Phys Neurol 1998;13:125–33. Ther 2007;87:1619–32. 20. Cleland JA, Whitman JM, Fritz JM, Palmer JA. 32. Egwu MO. Relative therapeutic e±cacy of some Manual physical therapy, cervical traction, and vertebral mobilization techniques in the manage- strengthening exercises in patients with cervical ment of unilateral cervical spondylosis: A compar- radiculopathy: A case series. J Orthop Sports Phys ative study. J Phys Ther Sci 2008;20:103–8. Ther 2005;35:802–11. 21. Waldrop MA. Diagnosis and treatment of cervical 33. Konin JG, Wiksten DL, Isear Jr JA, Brader H. radiculopathy using a clinical prediction rule and a Cervical spine. Special test for orthopedic exami- multimodal intervention approach: A case series. J nation. 3rd ed. Thorofare, NJ: Slack, 2006:13–7. Orthop Sports Phys Ther 2006;36:152–9. 22. Maitland GD. Vertebral Manipulation. 5th ed. 34. Hawker GA, Mian S, Kendzerska T, French M. Boston: Butterworth-Hienemann, 2003:201–12. Measures of adult pain. Arthritis Care Res 2011;63: s240–52. 35. Pietrobon R, Coeytaux RR, Carey TS, Richardson WJ, DeVellis RF. Standard scales for measurement of functional outcome for cervical pain or dys- function: A systematic review. Spine 2002;5: 515–22. 36. Vernon H, Mior S. The neck disability index: A study of reliability and validity. J Manipul Physiol Ther 1991;14:409–515. 37. Westaway MD, Stratford PW, Binkley JM. The patient-speci¯c functional scale: Validation of its

Hong Kong Physiother. J. 2018.38:149-160. Downloaded from www.worldscientific.com 160 A. O. Ojoawo & A. D. Olabode 52. Eubanks JD. Cervical radiculopathy: Nonoperative by 27.58.229.138 on 05/28/22. Re-use and distribution is strictly not permitted, except for Open Access articles. management of neck pain and radicular symptoms. use in persons with neck dysfunction. J Orthop Am Fam Physician 2010;81(1):33–40. Sports Phys Ther 1998;27:331–38. 38. Stratford PW, Riddle DL, Binkley JM, Spadoni G, 53. Saunders DH. Use of spinal traction in the treat- Westaway MD, Pad¯eld B. Using the neck dis- ment of neck and back conditions. Clin Orthop ability index to make decisions concerning indi- 1983;179:31–8. vidual patients. Physiother Can 1999;51:107–12. 39. Ylinen J, Takala EP, Nykanen M. Active neck 54. Bland JH. Disorders of the cervical spine: Diagnosis muscle training in the treatment of chronic neck and medical management. Philadelphia, PA: WB pain in women: A randomized controlled trial. Saunders, 1994; Bleakley C, McDonough S, JAMA 2003;289:2509–16. MacAuley D, The use of ice in the treatment of 40. Kisner C, Colby LA. Resisted exercise for impaired acute soft-tissue injury. Am J Sports Med 2004;32 muscle performance. In: Therapeutic Exercise (1):251–61. doi: 10.1177/0363546503260757. Foundation and Techniques. 5th ed. Philadelphia: F. A. Davis Company, 2007:169. 55. Haneline M. Chiropractic manipulation in the 41. Akinbo SR, Noronha CC, Okanlawon AO, Danesi presence of acute cervical intervertebral disc her- MA. E®ects of di®erent cervical traction weights on niation. Dyn Chiropract 1999;17(25). neck pain and mobility. Niger Postgrad Med J 2006;13(3):230–5. 56. Egwu MO, Alabi M, Nwuga VCB. E®ect of vertical 42. Farfan HF. The scienti¯c basis of manipulation oscillatory pressure on neck pain and some cardio- procedures. In: Buchanan WW, Kahn MF, Laine vascular variables. Physiotherapy 2003;89:666–74. V, et al. eds. Clinics in Rheumatic Diseases. Lon- don: W. B. Saunders Company, 1980:159–77. 57. Paris SV. Mobilization of spine. Phys, Ther 43. Giles LGF. Anatomical Basis of Low Back Pain. 1979;58:988–95. Baltimore: Williams & Wilkins, 1989. 44. Wilder DG, Pope MH, Frymoyer JW. The biome- 58. Lannersten L, Kosek E. Dysfunction of endogenous chanics of lumbar disc herniation and the e®ect of pain inhibition during exercise with painful muscles overload and instability. J Spinal Disord in patients with shoulder myalgia and ¯bromyalgia. 1988;1:16–32. Pain 2010;151:77–86. 45. Triano J. The mechanics of spinal manipulation. In: Herzog W, ed. Clinical Biomechanics of Spinal 59. Staud R, Robinson M, Price D. Isometric exercise Manipulation. New York: Churchill Livingstone, has opposite e®ects on central pain mechanisms in 2001:92–190. ¯bromyalgia patients compared to normal controls. 46. Elnaggar IM, Elhabashy HR, Abd El-Menam EM. Pain 2005;118:176–84. In°uence of spinal traction in treatment of cervical radiculopathy Egypt. J Neurol Psychiat Neurosurg 60. Kosek E, Lundberg L. Segmental and pluriseg- 2009;46(2):455–60. mental modulation of pressure pain thresholds 47. Voltonen EJ, Moller K, Wiljasob M, Arate B. during static muscle contractions in healthy indi- Comparative radiographic study of intermittent viduals. Eur J Pain 2003;7:251–8. and continuous traction on elongation of cervical spine. J Ann Med Intern 1996;57:143–6. 61. Jacobs M. Massage for the relief of pain: Anatom- 48. Krause M, Refshauge KM, Desen M, Boland R. ical and physiological considerations. Phys Ther Lumbar spine traction: Evaluation of e®ects and Rev 1960;40:93–8. recommended application for treatment. Man Ther 2000;5:72–81. 62. Watson J. Pain mechanisms: A review: 1. Char- 49. Borman P, Keskin D, Ekici B, Bodur H. The e±- acteristics of the peripheral receptors. Aust J cacy of intermittent cervical traction in patents Physiother 1981;27:135–43. with chronic neck pain. Clin Rheumatol 2008;27 (10):1249–53. doi: 10.1007/s10067-008-0895-z. 63. Goats GC. Massage — the scienti¯c basis of an 50. Rhee JM, Yoon T, Riew KD. Cervical radiculo- ancient art: Part 2. Physiological and therapeutic pathy. J Am Acad Orthop Surg 2007;15(8):486–94. e®ects. Br J Sports Med 1994;28(3):153–6. 51. Levine MJ, Albert TJ, Smith MD. Cervical radi- culopathy: Diagnosis and nonoperative manage- 64. Nordschow M, Bierman W. In°uence of manual ment. J Am Acad Orthop Surg 1996;4(6):305–16. massage on muscle relaxation: E®ect on trunk °exion. Phys Ther 1962;42:653. 65. Knight KL, Brucker JB, Stoneman PD. Muscle injury management with cryotherapy. Athletic Ther Today 2000;5:26–30. 66. Alga°y AA, George KP. The e®ect of cryotherapy on nerve conduction velocity, pain threshold and pain tolerance. Br J Sports Med 2007;41:365–9. Available at: http://dx.doi.org/10.1136/bjsm. 2006.031237.