Available online at

www.sciencedirect.com Joint Bone Spine 76 (2009) 15e19

Review

Contribution of clinical tests to the diagnosis of rotator cuff disease: A systematic literature review Johann Beaudreuil a,*, Re´my Nizard b, Thierry Thomas c, Mireille Peyre d, Jean Pierre Liotard e, Pascal Boileau f, Thierry Marc g, Catherine Dromard h, Elisabeth Steyer i, Thomas Bardin a, Philippe Orcel a, Gilles Walch e a

Service de Rhumatologie, Hoˆpital Lariboisie`re, 2 rue Ambroise Pare´, 75475 Paris cedex 10, France Service de Chirurgie Orthope´dique et Traumatologique, Hoˆpital Lariboisie`re, 75475 Paris cedex 10, France c Service de Rhumatologie, CHU de Saint-Etienne, 42055 Saint-Etienne, France d Service de Re´e´ducation, Hoˆpital national de Saint-Maurice, 94410 Saint-Maurice, France e Centre Orthope´dique Santy, 69008 Lyon, France f Service de Chirurgie Orthope´dique et Traumatologique, Hoˆpital de l’Archet, 06202 cedex 3 BP 3079, France g Centre de Re´e´ducation Spe´cialise´e, 34090 Montpellier, France h Cabinet me´dical, 91400 Saclay, France i Cabinet me´dical, 57525 Talange, France b

Accepted 23 April 2008 Available online 6 December 2008

Abstract Objective: To evaluate the diagnostic performance of clinical tests for degenerative rotator cuff disease, based on a systematic literature review. Methods: We searched Medline, Embase, and Pascal Biomed until the first half of 2006 inclusive for articles that reported at least the sensitivity and specificity of clinical tests for rotator cuff disease. Predictive values and accuracy were recorded where available. The results were discussed and validated. Results: We selected nine studies, of which three investigated tests for subacromial impingement syndrome and seven tests for rotator cuff tendinopathy. The Neer and Hawkins tests had good sensitivity but low specificity for subacromial impingement syndrome. For diagnosing tears of the supraspinatus or infraspinatus, the Jobe sign and the full can test showed similar performance characteristics to the Patte test and resisted external rotation with the elbow at the side flexed at 90 . For diagnosing tendinopathies with or without tears, active unresisted external rotation for the infraspinatus and the lift off test for the subscapularis were specific but lacked sensitivity. In one study, limitation of the range of active unresisted internal rotation was sensitive and specific for subscapularis tendon disease. The palm up test performed poorly for diagnosing long head of biceps disease. Conclusions: Data on the diagnostic performance of clinical tests for rotator cuff tendon disease are fragmentary. However objective data exist to support the usefulness of some of these tests. Further studies are needed. Ó 2008 Elsevier Masson SAS. All rights reserved. Keywords: Clinical tests; Rotator cuff disease; Diagnostic value; Systematic review

1. Introduction

* Corresponding author. Tel.: þ33 149 956 308; fax: þ33 149 958 631. E-mail address: [email protected] (J. Beaudreuil).

Degenerative rotator cuff disease is the leading cause of shoulder pain [1]. Clinical testing is the recommended firstline investigation for diagnosing degenerative rotator cuff disease [2]. Some clinical tests are designed to detect subacromial impingement, whereas others are intended for

1297-319X/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.jbspin.2008.04.015

16

J. Beaudreuil et al. / Joint Bone Spine 76 (2009) 15e19

identifying tendon and muscle dysfunction. Subacromial impingement is a cause of rotator cuff disease, whereas tendon and muscle dysfunction are consequences. The diagnostic performance of clinical tests can be assessed based on sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and ability to classify patients accurately according to the type of tendinopathy. Knowledge of these performance characteristics is crucial to the practice and interpretation of clinical tests conducted in patients with shoulder pain. The goal of this study was to review objective data on the diagnostic performance of clinical tests for rotator cuff disease. To this end, we conducted a systematic literature review.

Table 1 The nine selected studies. n of Patients

Authors and tests

Tests for subacromial impingement Calis et al., 2000 [3] 120 Litaker et al., 2000 [4] 448 McDonald et al., 2000 [5] 85 Tests for rotator cuff disease Noe¨l et al., 1989 [6] Walch, 1993 [7] Leroux et al., 1995 [8] Hertel et al., 1996 [9] Ito€ı et al., 1999 [10] Litaker et al., 2000 [4] Ardic et al., 2006 [11]

227 439 55 100 136 448 58

Age (years)

Recruitment

52 57 40

MS S S

52

S S S S S S M

51 51 43 57 55

S, surgical; M, medical; MS, medical and surgical.

2. Methods 3.2. Clinical tests for subacromial impingement We searched Medline, Embase, and Pascal Biomed using the following keywords: rotator cuff, shoulder pain, diagnostic value, sensitivity, specificity, predictive value, and diagnostic accuracy; and their French translations coiffe des rotateurs, e´paule douloureuse, valeur diagnostique, sensibilite´, spe´cificite´, valeur pre´dictive, and exactitude diagnostique. We used the literature search performed by the French High Health Authority in April 2005 for developing recommendations on the management of chronic shoulder pain without instability in adults [2]. We updated the findings to the first half of 2006 inclusive. Studies were selected based on the following criteria: patients with rotator cuff disease, number of patients known to be greater than 30, known reference criterion (subacromial impingement or tendon disease), well-described diagnostic method, knowledge of the prevalence of the abnormality in the overall study population, well-described clinical tests with clear definitions of positive and negative test results, and availability of sensitivity and specificity data. For each selected article, one of us recorded the sensitivity and specificity of each clinical test and, where available, the PPV, NPV, and accuracy. Data on subacromial impingement and data on rotator cuff lesions were presented separately. We grouped clinical tests that were similar in terms of intended use and performance characteristics and that were often evaluated in the same studies. The data from the literature review were discussed and validated by the other authors, who were experts in the fields of clinical rheumatology and scientific literature analysis. 3. Results 3.1. Selected studies We selected nine studies (Table 1). Three studies evaluated the detection of subacromial impingement [3e5]; the number of patients ranged from 85 to 448, mean age ranged from 40 to 57 years, and recruitment occurred chiefly in surgery departments. Sample size, age, and recruitment were comparable in the seven studies on rotator cuff lesions [4,6e11].

3.2.1. Painful arc test The painful arc during lateral arm elevation was evaluated in two studies comparatively to a positive subacromial injection test [3] or arthrographic evidence of a tendon tear [4]. These reference standards [12,13] are open to criticism. Subacromial injection of several milliliters of anesthetic may be followed by diffusion of the drug to the tendons and, in the event of tendon rupture, to the joints, limiting the specificity of the test results. Furthermore, subacromial impingement is only one of the many causative factors of tendon tears [14]. Therefore, tendon tears do not constitute a valid marker for subacromial impingement. These two studies [3,4] produced conflicting results in terms of sensitivity, specificity, PPV, and NPV (Table 2). Accuracy was poor in the study that evaluated this parameter [3]. Thus, the results fail to clarify the value of the painful arc test for diagnosing subacromial impingement syndrome. The discrepancies in the results may be ascribable to the differences between the two studies regarding both the reference standard and patient recruitment (medical and surgical in one study [3] and surgical only in the other [4]). Table 2 Performance characteristics of tests for subacromial impingement. Authors and tests Painful arc test Calis et al., 2000 [3] Litaker et al., 2000 [4]

Sensitivity Specificity PPV NPV Accuracy % % % % % 32 97

80 10

80 67

80 69

46

Neer test Calis et al. 2000 [3] 89 McDonald et al., 2000 [5] 75

30 47

76 36

52 83

72

Hawkins test Calis et al., 2000 [3] 91 McDonald et al., 2000 [5] 92

25 44

75 39

56 93

75

Calis et al., 2000 [3]: reference, subacromial injection; prevalence, 70%. Litaker et al., 2000 [4]: reference, rotator cuff tear/arthrography; prevalence, 67%. McDonald et al., 2000 [5]: reference, bursitis/arthroscopy; prevalence, 28%.

J. Beaudreuil et al. / Joint Bone Spine 76 (2009) 15e19

3.2.2. Neer and Hawkins tests Two studies evaluated the Neer and Hawkins tests (Table 2) [3,5]. For the Neer test, only anterolateral elevation was considered; the effects of subacromial injection were not assessed. The reference standard was the effect of subacromial injection [3] or findings by shoulder arthroscopy [5]. In the study that used arthroscopy as the reference [5], subacromial bursitis and rotator cuff tearing were evaluated separately and no difference was found. Therefore, we will discuss only the findings pertaining to subacromial bursitis, which is a more specific marker for subacromial impingement. Both tests were sensitive (Neer, 75e89%; and Hawkins, 91e92%). Specificity, in contrast, was low (Neer, 30e40%; and Hawkins, 25e44%) [3,5]. PPV and NPV differed between the two studies. Diagnostic accuracy was determined in one study [3] and showed that 72e75% of cases were correctly classified by either test. Combining the two tests produced similar performance characteristics [5]. 3.2.3. Yocum test Few data are available on the Yocum test. Sensitivity was estimated at 78% versus arthroscopic evidence of tendinopathy with or without tearing [8]. 3.3. Tests evaluating the rotator cuff tendons 3.3.1. Tests evaluating the supraspinatus tendon 3.3.1.1. Jobe test and full can test. The Jobe test is used for diagnosing tears in the supraspinatus tendon [15]. Two studies evaluated the Jobe test performed for this reason [6,10] (Table 3). When weakness was used to define a positive test, both studies indicated good sensitivity (77e95%), NPV (85e90%), and accuracy (70e85%), with somewhat lower specificity (65e68%). Using pain exacerbation to define a positive test

17

yielded less satisfactory performance characteristics [10]. PPV varied across studies [6]. When the reference standard was tendon damage with or without tearing, sensitivity was usually greater than specificity [4,8,9]. In two studies, a positive Jobe test correlated weakly with the extent of tendon tearing [7,8]. The full can test consists in evaluating the patient’s ability to resist downward pressure on the arms held at 90 elevation in the scapular plane and 45 external rotation. It was developed as an alternative to the Jobe test (or empty can test), as external rotation usually causes less pain than internal rotation [10]. A single study assessed the full can test (Table 3) [10]. The reference standard was supraspinatus tendon tearing by magnetic resonance imaging. Weakness was better than pain as a criterion for a positive test. The Jobe test and full can test assessed based on weakness showed comparable performance characteristics, although no head-to-head comparison was performed. 3.3.2. Tests evaluating the infraspinatus tendon 3.3.2.1. Patte test and resisted external rotation with the elbow at the side flexed at 90 . Several clinical tests involving external rotation of the arm are used to assess the infraspinatus tendon [4,7,9] (Table 3). The Patte test and resisted external rotation with the elbow at the side flexed at 90 were evaluated comparatively to presence of a partial or complete tear by arthrography [4,7]. Both tests proved useful. Sensitivity was 79% for the Patte test and 76% for external rotation. Specificity was somewhat lower (67 and 57%, respectively). 3.3.2.2. Lag signs in external rotation. Limited active external rotation with the arm abducted at 20 or at 90 (called the external rotation lag sign and drop sign, respectively) was evaluated in one study for detecting tendinopathy with or without tearing [9] (Table 4). Both tests showed high

Table 3 Performance characteristics of tests for rotator cuff disease. Authors and tests Supraspinatus tests Jobe test Noe¨l et al., 1989 [6] Ito€ı et al., 1999 [10] Full can test Ito€ı et al., 1999 [10]

Response Sensitivity Specificity PPV NPV Accuracy criterion % % % % %

Authors and tests Weakness 95 Pain 63 Weakness 77 Pain 66 Weakness 77

65 55 68 64 74

86 31 44 37 49

85 82 90 85 91

85 57 70 64 75

Infraspinatus tests Resisted external rotation with the elbow at the side flexed at 90 Litaker et al., 2000 [4] Weakness 76 57 78 54 Patte test Walch 1993 [7]

Table 4 Performance characteristics of tests for detecting rotator cuff disease with or without tearing.

Weakness 79

67

Noe¨l et al., 1989 [6]: reference, tear/arthrography; prevalence, 67%. Ito€ı et al., 1999 [10]: reference, full-thickness tear/MRI; prevalence, 24%. Litaker et al. [4]: reference, tear/arthrography; prevalence, 67%. Walch, 1993 [7]: reference, tear seen during surgery; prevalence, 40%.

Sensitivity %

Specificity %

PPV %

NPV %

Infraspinatus tests Weakness in active external rotation arm in 20 of abduction Hertel et al., 1996 [9] 70 100 100 56

Accuracy %

78



Weakness in active external rotation arm in 90 of abduction Hertel et al., 1996 [9] 21 100 100 32 Subscapularis tests Lift off test Walch, 1993 [7] Hertel et al., 1996 [9]

59 62

Weakness in active internal rotation Hertel et al., 1996 [9] 97

43

85 100 96

97

97

96

Hertel et al. [9]: reference, tendinopathy  tear at surgery; prevalence, 27% for the infraspinatus and 29% for the subscapularis. Walch, 1993 [7]: reference, tendinopathy  tear at surgery; prevalence, 29%.

J. Beaudreuil et al. / Joint Bone Spine 76 (2009) 15e19

18

Table 5 Performance characteristics of tests for detecting disease of the long head of biceps tendon. Authors and tests

Sensitivity %Specificity %PPV %NPV %Accuracy %

Palm up test Leroux et al., 1995 [8]63 Ardic et al., 2006 [11] 69

35 60

43 82

55 43

41

Leroux et al., 1995 [8]: reference, tendinopathy/arthroscopy; prevalence, 29%. Ardic et al., 2006 [11]: reference, tendinopathy/ultrasonography; prevalence, 61%.

specificity (100%), PPV (100%) and accuracy. However, sensitivity and NPV were low. 3.3.3. Tests for the subscapularis tendon Two tests involving internal rotation have been evaluated for detecting subscapularis tendon damage with or without tearing, namely, the lift off test and the internal rotation lag sign (Table 4) [7,9]. The lift off test showed high specificity (85e100%) and lower sensitivity (59e62%). The internal rotation lag sign was highly sensitive and specific in a single study [9], whose results need to be confirmed. 3.3.4. Tests for the long head of biceps Only the palm up test (also known as the Speed test or Gilcreest test) has been evaluated [8,11]. Its diagnostic performance was limited (Table 5). 4. Discussion Clinical testing for subacromial impingement and rotator cuff disease is a key step in the clinical assessment of the painful shoulder, together with history taking, inspection, palpation, and determination of active and passive motion ranges [2]. Clinical tests are most useful in patients without passive motion range limitation due to capsular or glenohumeral abnormalities. Passive motion range limitation may be associated with pain upon motion and with decreased resistance to loading, independently from tendon disease. Therefore, pain or weakness is difficult to ascribe to tendinopathy in patients with restricted passive motion. The tests discussed in this review were evaluated in patients with normal passive motion. Passive range of motion must be examined before using clinical tests for subacromial impingement or rotator cuff damage. In patients with suspected degenerative rotator cuff disease, clinical testing is performed as the first-line investigation for tendon damage, as well as to look for underlying subacromial impingement. The results may affect treatment decisions. Evidence of subacromial impingement may prompt a glucocorticoid injection into the subacromial bursa, together with rehabilitation aimed at dynamic centering of the humeral head. Weakness suggests tendon tearing, which requires close monitoring during medical treatment followed, in nonresponders, by surgical repair. Therefore, the diagnostic performance of clinical tests is of crucial importance.

However, we found that uncertainties remain regarding the validity of clinical tests. Of the many clinical tests for degenerative rotator cuff disease [2,16,17], some have not been evaluated and others have been investigated in only small numbers of studies [3e11] (Table 2). Among the most widely used tests for subacromial impingement, the Neer test seems particularly well-suited to the detection of anterolateral impingement involving the supraspinatus and infraspinatus tendons; and the Hawkins and Yocum tests to the detection of anteromedial impingement involving the supraspinatus, long head of biceps, and subscapularis tendons [18]. The Neer and Hawkins tests are sensitive for subacromial bursitis, which can be viewed as a marker for subacromial impingement, whether primary or secondary to tendon damage. However, both tests lack specificity. The Yocum test, which is sensitive, was usually evaluated as a tool for detecting tendinopathy of any cause rather than subacromial impingement. The reference standards used to evaluate the painful arc test are of doubtful relevance, and the performance data are conflicting. Thus, the use of tests for subacromial impingement rests on assumptions that have not been entirely validated. Their poor specificity may be related in part to the need for internally rotating the shoulder, which consistently causes pain in patients with shoulder disease [12,17,19,20]. Furthermore, the passive movements used for these tests may load the acromioclavicular joint, which may be the source of the symptoms. The Yocum test requires elevation and adduction of the shoulder, which also occur with the cross arm test used to detect symptomatic acromioclavicular disease [20]. Correlations between the results of tests for subacromial impingement and those of tests for acromioclavicular disease have not been investigated. Weakness during the Jobe test and full can test is sensitive for detecting supraspinatus tendon tears. Both tests are less specific than sensitive. Their performance characteristics are similar when weakness is used to evaluate the result. That pain without weakness indicates tendinopathy without tearing has been suggested but not validated. The Patte test and resisted external rotation with the elbow at the side flexed at 90 are sensitive for infraspinatus tendon tears; their specificity is lower than their sensitivity. In contrast, weakness during unresisted active external rotation is a specific sign that has good PPV for infraspinatus tendon damage. The lift off test is highly specific for subscapularis tendon lesions; sensitivity is lower. However, restricting the use of unresisted external rotation tests and of the lift off test only to the detection of infraspinatus and subscapularis tendon tearing or avulsion might improve sensitivity. Weakness of active internal rotation may be the most reliable marker for subscapularis tendinopathy, according to a single study. Tests for subscapularis tendinopathy often cause pain, in common with subacromial impingement tests, because they require marked internal rotation [2]. This factor limits their feasibility. The belly press test was recently suggested for evaluating the subscapularis tendon [21]. Although this test seems less painful, it needs to be evaluated [2]. No test performs well for detecting lesions to the long head of biceps. Inspection and palpation may show

J. Beaudreuil et al. / Joint Bone Spine 76 (2009) 15e19

evidence of tearing or dislocation out of the bicipital groove, with a decrease in muscle tone and a change, subtle at times, in the bulge produced by resisted contraction. 5. Conclusions In sum, the most extensively studied tests for subacromial impingement e Neer and Hawkins e are sensitive but lack specificity. Among tests for rotator cuff disease, when weakness is used to assess the test result, the Jobe test and the full can test show good sensitivity with somewhat lower specificity for the supraspinatus, and the same is true of the Patte test and external rotation lag test for the infraspinatus. Performance characteristics are poorer when pain is the response criterion. A number of tests can be used to detect tendinopathies with or without tearing: weakness of unresisted active external rotation for the infraspinatus and the lift off test and internal rotation lag test for the subscapularis. This last test was both sensitive and specific, in a single study. The other tests may lack sensitivity. Knowledge of these data is crucial, as clinical tests for the rotator cuff provide information on the nature of the lesions and therefore help to guide treatment decisions. Nevertheless, studies of the performance of clinical tests are few. Furthermore, the reference standards varied across studies and were sometimes open to criticism. In addition, the precise techniques and subjective interpretation required by clinical tests lead to substantial interobserver variability [22e25]. Thus, further evaluations are needed, as well as efforts to standardize the manner in which these tests are used, performed, and interpreted. This systematic review and the recommendations issued by the French High Health Authority in April 2005 [2] should help to achieve these goals. References [1] Roquelaure Y, Ha C, Leclerc A, et al. Epidemiologic surveillance of upper-extremity musculoskeletal disorders in the working population. Arthritis Rheum 2006;55:765e78. [2] Recommandations pour la pratique clinique. Modalite´s de prise en charge d’une e´paule douloureuse chronique non instable chez l’adulte. Available from: http://www.has-sante.fr; avril 2005 (accessed date 28.12.2008.). [3] Calis M, Akgu¨n K, Birtane M, et al. Diagnostic values of clinical diagnostic tests in subacromial impingement syndrome. Ann Rheum Dis 2000;59:44e7. [4] Litaker D, Pioro M, El Bilbeisi H, et al. Returning to the bedside: using the history and physical examination to identify rotator cuff tears. J Am Geriatr Soc 2000;48:1633e7.

19

[5] MacDonald PB, Clark P, Sutherland K. An analysis of the diagnostic accuracy of the Hawkins and Neer subacromial impingement signs. J Shoulder Elbow Surg 2000;9:299e301. [6] Noe¨l E, Walch G, Bochu M. La manoeuvre de Jobe. A propos de 227 cas. Rev Rhum 1989;56:803e4. [7] Walch G. L’examen programme´ de l’e´paule douloureuse chronique. In: Journe´e lyonnaise de l’e´paule. Lyon: clinique de chirurgie orthope´dique et traumatologique; 1993. p. 169e89. [8] Leroux JL, Thomas E, Bonnel F, et al. Diagnostic value of clinical tests for shoulder impingement syndrome. Rev Rhum Engl Ed 1995; 62:423e8. [9] Hertel R, Ballmer FT, Lombert SM, et al. Lag signs in the diagnosis of rotator cuff rupture. J Shoulder Elbow Surg 1996;5:307e13. [10] Itoi E, Kido T, Sano A, et al. Which is more useful, the ‘‘full can test’’ or the ‘‘empty can test’’, in detecting the torn supraspinatus tendon? Am J Sports Med 1999;27:65e8. [11] Ardic F, Kahraman Y, Kacar M, et al. Shoulder impingement syndrome. Relationships between clinical, functional, and radiologic findings. Am J Phys Med Rehabil 2006;85:53e60. [12] Neer CS. Impingement lesions. Clin Orthop Relat Res 1983;173:70e7. [13] Schaeverbecke T, Leroux JL. L’examen de l’e´paule de´ge´ne´rative. Rev Rhum 1996;63(Suppl. 1):15SPe21SP. [14] Bard H, Vandenbussche E, Cohen M, et al. Physiopathologie des tendinopathies de la coiffe des rotateurs. In: Rodineau J, Rolland E, editors. Pathologie intra- et pe´ritendineuse du membre supe´rieur des sportifs. Masson; 2006. p. 3e15. [15] Jobe FW, Jobe CM. Painful athletic injuries of the shoulder. Clin Orthop Relat Res 1983;173:117e24. [16] Tennent TD, Beach WR, Meyers JF. A review of the special tests associated with shoulder examination. Part I: the rotator cuff tests. Am J Sports Med 2003;31:154e60. [17] De Le´cluse J. Description princeps et modifications des tests cliniques. In: Rodineau J, Rolland E, editors. Pathologie intra- et pe´ritendineuse du membre supe´rieur des sportifs. Masson; 2006. p. 31e6. [18] Peyre M, Perrier D, Groc M, et al. Souffrance de la coiffe des rotateurs: recueil et analyse des donne´es cliniques-De´marche diagnostique. In: Rodineau J, Rolland E, editors. Pathologie intra- et pe´ritendineuse du membre supe´rieur des sportifs. Masson; 2006. p. 19e30. [19] Hawkins RJ, Kennedy JC. Impingement syndrome in athletes. Am J Sports Med 1980;8:151e8. [20] Yocum LA. Assessing the shoulder. History, physical examination, differential diagnosis, and special tests used. Clin Sports Med 1983;2: 281e9. [21] Gerber C, Hersche O, Farron A. Isolated rupture of the subscapularis tendon. J Bone Joint Surg Am 1996;78:1015e23. [22] de Winter AF, Jans MP, Scholten RJ, et al. Diagnostic classification of shoulder disorders: interobserver agreement and determinants of disagreement. Ann Rheum Dis 1999;58:272e7. [23] Norregaard J, Krogsgaard MR, Lorenzen T, et al. Diagnosing patients with longstanding shoulder joint pain. Ann Rheum Dis 2002;61:646e9. [24] Bertilson BC, Grunnesjo M, Strender LE. Reliability of clinical tests in the assessment of patients with neck/shoulder problems-impact of history. Spine 2003;28:2222e31. [25] Tamalet B, Ghozlane L, De Le´cluse J. Convergences et divergences au sein d’une e´quipe. In: Rodineau J, Rolland E, editors. Pathologie intra- et pe´ritendineuse du membre supe´rieur des sportifs. Masson; 2006. p. 37e9.