- Email: [email protected]
Hemiplegic Shoulder Pain Syndrome: Interrater Reliability of Physical Diagnosis Signs
294
BRIEF REPORT
Hemiplegic Shoulder Pain Syndrome: Interrater Reliability of Physical Diagnosis Signs Alexander W. Dromerick, MD, Ashok Kumar, MD, Oksana Volshteyn, MD, Dorothy F. Edwards, PhD ABSTRACT. Dromerick AW, Kumar A, Volshteyn O, Edwards DF. Hemiplegic shoulder pain syndrome: interrater reliability of physical diagnosis signs. Arch Phys Med Rehabil 2006;87:294-5. Objective: To determine the interrater reliability of shoulder physical diagnosis signs in the acute stroke rehabilitation setting. Design: Prospective inception cohort. Setting: Academic inpatient stroke rehabilitation service. Participants: People admitted to stroke rehabilitation service. Interventions: Not applicable. Main Outcome Measures: The Neer impingement test, Speed test, acromioclavicular shear test, Rowe shoulder score, and palpation. Results: Two examiners evaluated 46 consecutively admitted participants at 18.9⫾14.1 days after stroke. Percentage agreement was 78% or higher on all tests. The statistic was in the fair to excellent interrater reliability range on all tests except those involving the acromioclavicular joint, for which findings were infrequent. Conclusions: Most of the shoulder physical diagnosis signs used in this study have sufficient interrater reliability for use in future clinical studies of hemiplegic shoulder pain. Key Words: Cerebrovascular accident; Joint diseases; Physical examination; Rehabilitation; Shoulder; Shoulder pain; Stroke. © 2006 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation HE HEMIPLEGIC SHOULDER PAIN syndrome (HSPS) has been reported to occur in 5% to 84% of patients after T a stroke. Despite the high frequency of HSPS, the literature 1,2
is mixed regarding the epidemiology, risk factors, and etiology of HSPS. One step in clarifying HSPS is to standardize the soft tissue examination of the shoulder. A structured evaluation using musculoskeletal examination techniques with known interrater reliability in persons with stroke would allow systematic and valid cataloging of many of the clinical phenomena of HSPS.
From the Division of Physical Medicine and Rehabilitation and Department of Neurology (Dromerick, Kumar, Volshteyn, Edwards), Program in Occupational Therapy (Dromerick, Edwards), and Program in Physical Therapy (Dromerick), Washington University School of Medicine, St. Louis, MO. Supported by the Missouri Arthritis Foundation, the National Institute for Neurologic Diseases and Stroke (grant no. NS41261), and the James S. McDonnell Foundation (grant no. 21002032). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. Correspondence to Alexander W. Dromerick, MD, National Rehabilitation Hospital, Georgetown University School of Medicine, 102 Irving St NW, Washington, DC 20010, e-mail: [email protected]. Reprints are not available from the author. 0003-9993/06/8702-10145$32.00/0 doi:10.1016/j.apmr.2005.10.004
Arch Phys Med Rehabil Vol 87, February 2006
The frequency and patterns of soft tissue involvement observed then allows the objective classification of patient subgroups and correlation of clinical findings with radiographic or laboratory data. We determined the interrater reliability of well-known musculoskeletal physical examination techniques in persons undergoing acute stroke rehabilitation. We chose these measures because they are widely used, had not been previously evaluated in stroke, and because they could be unambiguously defined. To our knowledge, no systematic comparison to musculoskeletal imaging of these clinical tests has been made in stroke patients. METHODS This prospective study was performed in an academic stroke rehabilitation program and was approved by the Washington University Human Studies Committee. All persons admitted for stroke rehabilitation during the 3-month study period were included, except those who refused, were unable to follow 1-step commands, or could not provide informed consent. Each consenting subject was examined within 2 weeks of rehabilitation admission. To determine interrater reliability, the examination was performed independently by 2 physiatrists unaware of the other’s findings. The observers had previously examined several patients together to ensure consistent technique. The affected shoulder was assessed for specific sites of tenderness.3 The bicipital tendon was palpated between the lesser and greater tuberosities of the humeral head. The supraspinatus tendon was palpated anteriorly over its insertion at the greater tuberosity of the humerus, with the arm at 30° of shoulder extension. The subacromial area was examined by palpating the gap between the acromial process and head of the humerus on the superolateral aspect of shoulder. Pain in the subacromial region is traditionally attributed to inflammation of the subacromial bursa. The acromioclavicular joint (ACJ), coracoid process, and surrounding soft tissues were also examined for any localized or diffuse tenderness. Four physical diagnosis maneuvers were tested. In the Neer impingement test,4 the subject’s arm was elevated through forward flexion by the examiner passively bringing the greater tuberosity against the anteroinferior border of the acromion. An abnormal finding is the provocation of pain; such provocation is said to be indicative of impingement of the rotator cuff or injury to the supraspinatus muscle. In the Speed test,5 the patient attempts shoulder flexion against resistance; during this maneuver, the forearm is supinated and the elbow is extended. Pain in the bicipital groove is said to be indicative of bicipital tendon involvement. The acromioclavicular shear test6 is said to indicate pathology at the ACJ. This maneuver was performed with the subject sitting while the examiner cupped his or her hands anteriorly on the clavicle and posteriorly on the spine of scapula. Squeezing the heel of the hands together elicits pain in the presence of ACJ inflammation. Finally, the Rowe shoulder score was used to demonstrate multidirectional instability in the shoulder.5 Here, the examiner held the head of the humerus by placing 1 hand over the shoulder so that the index and the middle fingers sat over the anterior aspect of the humeral head and the thumb on the posterior aspect of the humeral head.
295
HEMIPLEGIC SHOULDER PAIN SYNDROME, Dromerick
The examiner then exerted anterior and posterior force to elicit instability in either direction. For inferior instability, traction was applied vertically to elicit a sulcus sign. Statistical analyses were performed using SASa for Windows. Interrater reliability was described both by the percentage of agreement between observers and the statistic.7,8 For interpreting the degree of interobserver agreement, the statistic is considered poor if less than 0.4, fair between 0.4 and 0.6, good between 0.6 and 0.8, and excellent if greater than 0.8.7 The result can be deceptively low in cases where there is a high rate of interobserver agreement, but where the occurrence of the condition is infrequent. RESULTS Two examiners completed initial evaluations of all 46 subjects who met entry criteria. The mean time from stroke onset to evaluation was 18.9⫾14.1 days. Five other patients were admitted for stroke rehabilitation during the study period but were not included because of severe aphasia, refusal, or transfer for complications. The dominant hand was involved in 50%; all but 2 subjects were right handed. Prior stroke was reported by 28% of subjects, and 22% reported preexisting shoulder pain on the affected side. No participant reported disabling shoulder pain before stroke. Using the National Institutes of Health Stroke Scale neglect item,9 71% of subjects were without neglect. Weakness of shoulder flexion, extension, or abduction was found in 85% of subjects. Hypertonicity on passive shoulder flexion and abduction was found in 65% of subjects, using the Modified Ashworth Scale.10 Frequencies of the various findings are listed in table 1. The results of the interrater reliability evaluations for the localization of tenderness to palpation and for the physical diagnosis maneuvers are shown in table 2. The percentage of agreement between raters was high for all assessments. With respect to localized tenderness, the statistics ranged from .18 to .84. The statistic revealed moderate to substantial interrater reliability in all tests except those involving the ACJ. For all the physical diagnosis maneuvers, the percentage agreement was greater than 80% and the statistics were in the fair to good range. DISCUSSION Our results show that most of the physical diagnosis signs we evaluated have sufficient interrater reliability to be used in
Table 1: Participant Characteristics (Nⴝ46) Participant Characteristics
Age (y) Sex Male Female Prior stroke Stroke location Left Right Prior nondisabling shoulder pain Self-reported pain
Values
57.30⫾25.20 (33–78) 24 (52) 22 (48) 13 (28) 24 (52) 22 (48) 10 (22) 17 (37)
NOTE. Values are mean ⫾ standard deviation (range) or n (%).
Table 2: Interobserver Agreement for Pain on Palpation and Physical Diagnosis Signs Interobserver Agreement
Percentage Agreement
88 84 78 87
.84 (.65–1.0) .64 (.42–.86) .56 (.32–.80) .18 (–.23 to .59)
86
.78 (.60–.96)
83 84
.72 (.54–.91) .62 (.28–.95)
80
.48 (.09–.87)
Pain on palpation (frequency) Subacromial (26%) Bicipital (54%) Rotator cuff (50%) ACJ tenderness (7%) Physical diagnosis signs (frequency) Neer impingement test (50%) Speed test (9%) Rowe shoulder score (15%) ACJ shear (11%) NOTE. Values are percent or
(95% confidence intervals).
future studies. An exception may be the tests that evaluate pathology at the ACJ. Because the statistic evaluates the rate of agreement beyond chance, even occasional disagreements between raters can cause the value of to plummet in rarely occurring conditions.8 However, the high rate of interobserver agreement suggests that even the ACJ tests might require only small refinements to improve reliability. CONCLUSIONS This demonstrated reliability now facilitates longitudinal natural history studies, validation of physical findings compared with imaging results, and better delineation of the pathophysiology of HSPS. References 1. Dromerick AW, Reding MJ. Medical and neurological complications during stroke rehabilitation. Stroke 1994;25:358-61. 2. Roy CW. Shoulder pain in hemiplegia: a literature review. Clin Rehabil 1988;2:35-44. 3. Hoppenfeld S. Physical examination of the shoulder: physical examination of the spine and extremities. East Norwalk: Appleton-Century-Crofts; 1976. p 1-34. 4. Neer CS, Welsh RP. The shoulder in sports. Orthop Clin North Am 1997;8:583-91. 5. Rowe CR. Dislocation of the shoulder. In: Rowe CR, editor. The shoulder. Edinburgh: Churchill Livingstone; 1988. 6. Davies GI, Gould IA, Larson RL. Functional examination of the shoulder girdle. Phys Sports Med 1981;9:82-104. 7. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1997;33:159-74. 8. Tooth LJ, Ottenbacher KJ. The statistic in rehabilitation research: an examination. Arch Phys Med Rehabil 2004;85:1371-6. 9. Brott T, Adams HP, Olinger CP, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989;20: 864-70. 10. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther 1987;67:206-7. Supplier a. Version 9.12; SAS Institute Inc, 100 SAS Campus Dr, Cary, NC 27513.
Arch Phys Med Rehabil Vol 87, February 2006
BRIEF REPORT
Hemiplegic Shoulder Pain Syndrome: Interrater Reliability of Physical Diagnosis Signs Alexander W. Dromerick, MD, Ashok Kumar, MD, Oksana Volshteyn, MD, Dorothy F. Edwards, PhD ABSTRACT. Dromerick AW, Kumar A, Volshteyn O, Edwards DF. Hemiplegic shoulder pain syndrome: interrater reliability of physical diagnosis signs. Arch Phys Med Rehabil 2006;87:294-5. Objective: To determine the interrater reliability of shoulder physical diagnosis signs in the acute stroke rehabilitation setting. Design: Prospective inception cohort. Setting: Academic inpatient stroke rehabilitation service. Participants: People admitted to stroke rehabilitation service. Interventions: Not applicable. Main Outcome Measures: The Neer impingement test, Speed test, acromioclavicular shear test, Rowe shoulder score, and palpation. Results: Two examiners evaluated 46 consecutively admitted participants at 18.9⫾14.1 days after stroke. Percentage agreement was 78% or higher on all tests. The statistic was in the fair to excellent interrater reliability range on all tests except those involving the acromioclavicular joint, for which findings were infrequent. Conclusions: Most of the shoulder physical diagnosis signs used in this study have sufficient interrater reliability for use in future clinical studies of hemiplegic shoulder pain. Key Words: Cerebrovascular accident; Joint diseases; Physical examination; Rehabilitation; Shoulder; Shoulder pain; Stroke. © 2006 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation HE HEMIPLEGIC SHOULDER PAIN syndrome (HSPS) has been reported to occur in 5% to 84% of patients after T a stroke. Despite the high frequency of HSPS, the literature 1,2
is mixed regarding the epidemiology, risk factors, and etiology of HSPS. One step in clarifying HSPS is to standardize the soft tissue examination of the shoulder. A structured evaluation using musculoskeletal examination techniques with known interrater reliability in persons with stroke would allow systematic and valid cataloging of many of the clinical phenomena of HSPS.
From the Division of Physical Medicine and Rehabilitation and Department of Neurology (Dromerick, Kumar, Volshteyn, Edwards), Program in Occupational Therapy (Dromerick, Edwards), and Program in Physical Therapy (Dromerick), Washington University School of Medicine, St. Louis, MO. Supported by the Missouri Arthritis Foundation, the National Institute for Neurologic Diseases and Stroke (grant no. NS41261), and the James S. McDonnell Foundation (grant no. 21002032). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. Correspondence to Alexander W. Dromerick, MD, National Rehabilitation Hospital, Georgetown University School of Medicine, 102 Irving St NW, Washington, DC 20010, e-mail: [email protected]. Reprints are not available from the author. 0003-9993/06/8702-10145$32.00/0 doi:10.1016/j.apmr.2005.10.004
Arch Phys Med Rehabil Vol 87, February 2006
The frequency and patterns of soft tissue involvement observed then allows the objective classification of patient subgroups and correlation of clinical findings with radiographic or laboratory data. We determined the interrater reliability of well-known musculoskeletal physical examination techniques in persons undergoing acute stroke rehabilitation. We chose these measures because they are widely used, had not been previously evaluated in stroke, and because they could be unambiguously defined. To our knowledge, no systematic comparison to musculoskeletal imaging of these clinical tests has been made in stroke patients. METHODS This prospective study was performed in an academic stroke rehabilitation program and was approved by the Washington University Human Studies Committee. All persons admitted for stroke rehabilitation during the 3-month study period were included, except those who refused, were unable to follow 1-step commands, or could not provide informed consent. Each consenting subject was examined within 2 weeks of rehabilitation admission. To determine interrater reliability, the examination was performed independently by 2 physiatrists unaware of the other’s findings. The observers had previously examined several patients together to ensure consistent technique. The affected shoulder was assessed for specific sites of tenderness.3 The bicipital tendon was palpated between the lesser and greater tuberosities of the humeral head. The supraspinatus tendon was palpated anteriorly over its insertion at the greater tuberosity of the humerus, with the arm at 30° of shoulder extension. The subacromial area was examined by palpating the gap between the acromial process and head of the humerus on the superolateral aspect of shoulder. Pain in the subacromial region is traditionally attributed to inflammation of the subacromial bursa. The acromioclavicular joint (ACJ), coracoid process, and surrounding soft tissues were also examined for any localized or diffuse tenderness. Four physical diagnosis maneuvers were tested. In the Neer impingement test,4 the subject’s arm was elevated through forward flexion by the examiner passively bringing the greater tuberosity against the anteroinferior border of the acromion. An abnormal finding is the provocation of pain; such provocation is said to be indicative of impingement of the rotator cuff or injury to the supraspinatus muscle. In the Speed test,5 the patient attempts shoulder flexion against resistance; during this maneuver, the forearm is supinated and the elbow is extended. Pain in the bicipital groove is said to be indicative of bicipital tendon involvement. The acromioclavicular shear test6 is said to indicate pathology at the ACJ. This maneuver was performed with the subject sitting while the examiner cupped his or her hands anteriorly on the clavicle and posteriorly on the spine of scapula. Squeezing the heel of the hands together elicits pain in the presence of ACJ inflammation. Finally, the Rowe shoulder score was used to demonstrate multidirectional instability in the shoulder.5 Here, the examiner held the head of the humerus by placing 1 hand over the shoulder so that the index and the middle fingers sat over the anterior aspect of the humeral head and the thumb on the posterior aspect of the humeral head.
295
HEMIPLEGIC SHOULDER PAIN SYNDROME, Dromerick
The examiner then exerted anterior and posterior force to elicit instability in either direction. For inferior instability, traction was applied vertically to elicit a sulcus sign. Statistical analyses were performed using SASa for Windows. Interrater reliability was described both by the percentage of agreement between observers and the statistic.7,8 For interpreting the degree of interobserver agreement, the statistic is considered poor if less than 0.4, fair between 0.4 and 0.6, good between 0.6 and 0.8, and excellent if greater than 0.8.7 The result can be deceptively low in cases where there is a high rate of interobserver agreement, but where the occurrence of the condition is infrequent. RESULTS Two examiners completed initial evaluations of all 46 subjects who met entry criteria. The mean time from stroke onset to evaluation was 18.9⫾14.1 days. Five other patients were admitted for stroke rehabilitation during the study period but were not included because of severe aphasia, refusal, or transfer for complications. The dominant hand was involved in 50%; all but 2 subjects were right handed. Prior stroke was reported by 28% of subjects, and 22% reported preexisting shoulder pain on the affected side. No participant reported disabling shoulder pain before stroke. Using the National Institutes of Health Stroke Scale neglect item,9 71% of subjects were without neglect. Weakness of shoulder flexion, extension, or abduction was found in 85% of subjects. Hypertonicity on passive shoulder flexion and abduction was found in 65% of subjects, using the Modified Ashworth Scale.10 Frequencies of the various findings are listed in table 1. The results of the interrater reliability evaluations for the localization of tenderness to palpation and for the physical diagnosis maneuvers are shown in table 2. The percentage of agreement between raters was high for all assessments. With respect to localized tenderness, the statistics ranged from .18 to .84. The statistic revealed moderate to substantial interrater reliability in all tests except those involving the ACJ. For all the physical diagnosis maneuvers, the percentage agreement was greater than 80% and the statistics were in the fair to good range. DISCUSSION Our results show that most of the physical diagnosis signs we evaluated have sufficient interrater reliability to be used in
Table 1: Participant Characteristics (Nⴝ46) Participant Characteristics
Age (y) Sex Male Female Prior stroke Stroke location Left Right Prior nondisabling shoulder pain Self-reported pain
Values
57.30⫾25.20 (33–78) 24 (52) 22 (48) 13 (28) 24 (52) 22 (48) 10 (22) 17 (37)
NOTE. Values are mean ⫾ standard deviation (range) or n (%).
Table 2: Interobserver Agreement for Pain on Palpation and Physical Diagnosis Signs Interobserver Agreement
Percentage Agreement
88 84 78 87
.84 (.65–1.0) .64 (.42–.86) .56 (.32–.80) .18 (–.23 to .59)
86
.78 (.60–.96)
83 84
.72 (.54–.91) .62 (.28–.95)
80
.48 (.09–.87)
Pain on palpation (frequency) Subacromial (26%) Bicipital (54%) Rotator cuff (50%) ACJ tenderness (7%) Physical diagnosis signs (frequency) Neer impingement test (50%) Speed test (9%) Rowe shoulder score (15%) ACJ shear (11%) NOTE. Values are percent or
(95% confidence intervals).
future studies. An exception may be the tests that evaluate pathology at the ACJ. Because the statistic evaluates the rate of agreement beyond chance, even occasional disagreements between raters can cause the value of to plummet in rarely occurring conditions.8 However, the high rate of interobserver agreement suggests that even the ACJ tests might require only small refinements to improve reliability. CONCLUSIONS This demonstrated reliability now facilitates longitudinal natural history studies, validation of physical findings compared with imaging results, and better delineation of the pathophysiology of HSPS. References 1. Dromerick AW, Reding MJ. Medical and neurological complications during stroke rehabilitation. Stroke 1994;25:358-61. 2. Roy CW. Shoulder pain in hemiplegia: a literature review. Clin Rehabil 1988;2:35-44. 3. Hoppenfeld S. Physical examination of the shoulder: physical examination of the spine and extremities. East Norwalk: Appleton-Century-Crofts; 1976. p 1-34. 4. Neer CS, Welsh RP. The shoulder in sports. Orthop Clin North Am 1997;8:583-91. 5. Rowe CR. Dislocation of the shoulder. In: Rowe CR, editor. The shoulder. Edinburgh: Churchill Livingstone; 1988. 6. Davies GI, Gould IA, Larson RL. Functional examination of the shoulder girdle. Phys Sports Med 1981;9:82-104. 7. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1997;33:159-74. 8. Tooth LJ, Ottenbacher KJ. The statistic in rehabilitation research: an examination. Arch Phys Med Rehabil 2004;85:1371-6. 9. Brott T, Adams HP, Olinger CP, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989;20: 864-70. 10. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther 1987;67:206-7. Supplier a. Version 9.12; SAS Institute Inc, 100 SAS Campus Dr, Cary, NC 27513.
Arch Phys Med Rehabil Vol 87, February 2006