Abstract
Background:
Diagnosing superior labrum anterior and posterior (SLAP) lesions through physical examination remains challenging. The dynamic labral shear test (DLST) has been shown to have likelihood ratios (LRs) of 31.6 and 1.1 for diagnosing SLAP lesions.
Purpose:
To determine the clinical utility of the DLST for diagnosing SLAP lesions.
Study Design:
Cohort study (diagnosis); Level of evidence, 2.
Methods:
This prospective, consecutive case series included 774 patients who underwent diagnostic arthroscopy and a preoperative DLST between 2007 and 2013. Patients were divided into 3 groups: 610 control patients with no SLAP lesion but with other abnormalities, 9 patients with isolated SLAP lesion (ISL), and 155 patients with concomitant SLAP lesion (CSL), who had a SLAP lesion and another shoulder abnormality. We determined sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), odds ratio (OR), and diagnostic accuracy (DA) of the DLST with and without other tests.
Results:
The DLST was positive for 242 of 610 controls (40%), 7 of 9 patients (78%) in the ISL group, and 88 of 155 patients (57%) in the CSL group. In the ISL group, the DLST had a sensitivity of 78%, specificity of 51%, PPV of 2%, NPV of 100%, OR of 3.58, and DA of 51%. In comparison, the ORs were 1.09 for the active compression test, 1.30 for the lift-off test, and 1.53 for the relocation test, which were not significantly different from each other. For diagnosing a SLAP lesion existing in a joint with other associated injury, the DLST had a sensitivity of 57%, specificity of 52%, PPV of 23%, NPV of 83%, OR of 1.4, and DA of 53%. Combining all 4 tests did not improve the OR for detecting ISLs or CSLs.
Conclusion:
The DLST is sensitive but not specific for detecting ISLs. With an OR of 3.58, the DLST is useful for diagnosing ISLs. However, in patients who have CSLs, the DLST is not as useful for diagnosing SLAP lesions.
The superior labrum anterior and posterior (SLAP) lesion was first described by Andrews et al 1 in 1985, and the acronym SLAP was coined by Snyder et al 33 in 1990. Snyder et al 33 originally described 4 variants of the SLAP lesion, and since then additional variants have been described. 23 The most common are type I (associated with degenerative changes in the glenohumeral joint) and type II (frequently seen in athletes or after injury). 22
Despite the plethora of literature regarding SLAP lesions, their clinical diagnosis remains challenging for a number of reasons. First, although many physical examination tests are available to assist in diagnosing SLAP lesions,3,12,19,28,34 the diagnostic value of many of these tests is inconsistent and ambiguous.4,6,13,14,18,24,31 Second, SLAP lesions most commonly occur concomitantly with other shoulder injuries (eg, glenohumeral instability, rotator cuff tears, biceps tendon ruptures). Third, SLAP lesions have no specific associated pain pattern, and patients may be unable to accurately describe the location of their pain or provide a precise history. 2 Therefore, clinicians must rely on a combination of history, physical examination, and magnetic resonance imaging findings to make a preoperative diagnosis. Because of this uncertainty, diagnostic arthroscopy is necessary to positively identify a SLAP lesion.
The dynamic labral shear test (DLST), which was first described in 2007, is used for diagnosing SLAP lesions (Cheung EV, O’Driscoll SW. “The Dynamic Labral Shear Test for Superior Labral Anterior Posterior Tears of the Shoulder.” Presented at American Academy of Orthopaedic Surgeons, 2007). Two studies have evaluated the clinical utility of the DLST for diagnosing labral abnormality; one study suggested that the DLST was not diagnostic (odds ratio = 1.1), 5 and the other concluded that it was diagnostic (odds ratio = 31.6). 21
The purposes of our study were (1) to assess the clinical utility (ie, sensitivity, specificity, positive predictive value, negative predictive value, odds ratio, and diagnostic accuracy) of the DLST for detecting isolated SLAP lesions (ISLs); (2) to determine the clinical utility of the DLST in the shoulders of patients who have SLAP lesions with coexisting abnormalities; and (3) to determine whether using the DLST with other common physical examination tests for SLAP lesions increases the clinical utility of the test. Drawing on our clinical experience, we hypothesized that the clinical utility of the DLST (1) would be low for diagnosing ISLs, (2) would be even lower in the presence of concomitant shoulder abnormalities, and (3) would not improve in combination with other tests.
Methods
Institutional review board approval was obtained for this retrospective consecutive case series of patients studied prospectively between 2007 and 2013. For inclusion, patients must have undergone a physical examination consisting of 4 tests and subsequent diagnostic arthroscopy, which was performed by a single surgeon. Patients with previous shoulder surgeries were excluded. All 1371 patients who received arthroscopy at our institution underwent this protocol, and 774 met the criteria.
All study patients had a standardized preoperative visit within 4 weeks before surgery. During this evaluation, patients completed a detailed standardized questionnaire, underwent a standardized physical examination, and had preoperative shoulder radiography. All patients were examined by the senior author (E.G.M.) or his physician assistant. The examination included commonly used tests to evaluate SLAP lesions, as well as measures of shoulder range of motion, strength, and laxity.
The DLST was performed in each patient as described by Cheung and O’Driscoll (“The Dynamic Labral Shear Test for Superior Labral Anterior Posterior Tears of the Shoulder.” Presented at American Academy of Orthopaedic Surgeons, 2007) and Kibler et al. 21 The test was performed with the examiner behind the standing patient, holding the wrist of the patient with 1 hand and applying an anteriorly directed force on the proximal humerus near the joint line with the other hand (Figure 1).
Dynamic labral shear test. The examiner stands behind the patient, holding the patient’s wrist with 1 hand and applying an anteriorly directed force on the proximal humerus near the joint line with the other hand.
The patient was instructed to relax and to not resist the motion in any manner. The patient’s arm was then elevated in the plane of the body from the side to maximal abduction (Figure 2).
Dynamic labral shear test. The patient’s arm is elevated from the side to 150°. A test is considered positive when the patient reports pain or the examiner feels a click in the patient’s posterior shoulder between 90° and 120° of elevation.
A test was considered positive when the patient reported pain or the examiner felt a click in the patient’s posterior joint line between 90° and 120° of elevation.
In addition, we performed the active compression test, 28 the relocation test, 12 and the resisted lift-off test. 10 These tests were chosen because they are widely reported in the literature and normally are well-tolerated. The active compression test was performed as previously described 28 with the patient standing; the elbow extended; and the arm in 90° of forward flexion, in 10° of horizontal adduction, and internally rotated with the thumb pointed down. The examiner applied downward pressure at the patient’s wrist and asked the patient to resist the downward force and to report any pain or click “inside the shoulder.” The patient was then asked to externally rotate the arm with the palm pointed up and once again to resist the force of the examiner’s downward pressure. The test was positive if there was pain in the thumb-down position that was reduced or eliminated in the palm-up position.
The relocation test was performed as previously described 12 with the patient supine on the examination table with the elbow flexed 90° and the arm abducted 90°. The arm was then gradually externally rotated until the patient reported pain in the shoulder. The examiner then placed a posteriorly directed force on the proximal humerus near the shoulder, pushing the humeral head posteriorly. The test was considered positive if the patient’s pain was relieved by the translation of the humeral head posteriorly.
The resisted lift-off test was performed in a manner similar to the original lift-off test described by Gerber and Krushell. 9 With the arm in extension, the patient was asked to internally rotate the arm up the back with the palm facing posteriorly and lifted off the back. The examiner pushed against the patient’s hand and asked the patient to resist the pressure. 25 The test was considered positive if the maneuver elicited pain.
All patients underwent diagnostic arthroscopy by the senior author, who was not blinded to the potential diagnosis. (The senior author performs only shoulder surgery and has been in practice for 26 years.) With the patient in a lateral decubitus position, the arthroscope was inserted through a posterior portal. All patients had a supplementary anterior portal through which the labrum was probed by means of a nerve hook. SLAP lesions were classified via the system described by Snyder et al. 33 Lesions of type II through IV were considered positive for this study. A type I lesion was considered to be a degenerative lesion only and therefore a negative examination. All patients underwent a systematic examination of the entire glenohumeral joint for any rotator cuff, ligamentous, or labral abnormalities. The biceps tendon was pulled into the joint to detect any proximal groove injuries in all cases.
Measures of diagnostic accuracy, sensitivity, specificity, odds ratio, negative and positive predictive value, and positive and negative likelihood ratio were calculated from a standard 2 × 2 table for each physical examination and combination of examinations. Statistical differences between odds ratios were calculated from a z test of the log odds ratio, and an alpha level of .05 was chosen.
Results
Characteristics of the study population are presented in Table 1. Of the 774 patients, 9 had an isolated SLAP lesion, and 155 had a SLAP lesion with a concomitant abnormality. The remaining 610 patients without a SLAP lesion were considered the control group.
Characteristics of 774 Patients Who Underwent Diagnostic Shoulder Arthroscopy by the Senior Author Between 2007 and 2013 a
CSL, concomitant superior labrum anterior and posterior lesion; ISL, isolated superior labrum anterior and posterior lesion.
Expressed as mean ± SD.
Of the 765 patients in the control and concomitant SLAP lesion (CSL) groups, 582 (76%) had an intra-articular lesion, of whom 467 (80%) had a supraspinatus tear (263 partial, 200 full, and 4 massive); 111 (19%) had infraspinatus tears (63 partial, 48 full); 213 (36%) had subscapularis tears (159 partial, 52 full, and 2 massive); 103 (18%) had Bankart lesions; and 98 (17%) had Hill-Sachs lesions. Three hundred seventy-four (64%) patients had humeral head osteoarthritis, and 305 (52%) had glenoid osteoarthritis. All patients in the CSL group had other associated diagnoses: instability, 155 (100%); acromioclavicular arthritis, 68 (44%); biceps tear, 20 (13%); and other diagnoses, 67 (44%).
The DLST was positive for 242 of 610 patients (40%) in the control group, 7 of 9 patients (78%) in the ISL group, and 88 of 155 patients (57%) in the CSL group. The results of the 4 physical examination tests in patients in the ISL group are summarized in Table 2. The most sensitive tests were the DLST (78%) and the active compression test (78%), followed by the lift-off test (56%) and the relocation test (33%). The most specific test was the relocation test (78%), followed by the DLST (51%), the lift-off test (50%), and the active compression test (24%). The likelihood ratio was highest for the DLST (3.54), followed by the relocation test (1.74), the lift-off test (1.30), and the active compression test (1.08). There was no increase in the odds ratios for these tests when they were used together in various combinations.
Accuracy of Examination Tests for the Diagnosis of Isolated SLAP Lesions a
CI, confidence interval; DA, diagnostic accuracy; DLST, dynamic labral shear test; +LR, positive likelihood ratio; –LR, negative likelihood ratio; NPV, negative predictive value; PPV, positive predictive value; SLAP, superior labrum anterior and posterior.
The results of the 4 physical examination tests in patients who had coexisting abnormalities are summarized in Table 3. All tests had less clinical utility in this group compared with the ISL group. When the 4 tests were statistically evaluated in different combinations, no increase was found in the likelihood ratio for any of the combinations versus the use of the DLST alone.
Accuracy of Examination Tests for the Diagnosis of SLAP Lesions With Concomitant Shoulder Abnormalities a
CI, confidence interval; DA, diagnostic accuracy; DLST, dynamic labral shear test; +LR, positive likelihood ratio; –LR, negative likelihood ratio; NPV, negative predictive value; PPV, positive predictive value; SLAP, superior labrum anterior and posterior.
Discussion
This study shows that the DLST has clinical utility for patients with ISLs but not for patients with CSLs. Despite its relatively high likelihood ratio for detecting ISLs, the DLST is sensitive but not specific for that lesion. This property would cause the lesion to be overdiagnosed in patients suspected of having ISLs. In patients with concomitant shoulder abnormalities (ie, most of the patients in our study and in other reported studies), the DLST does not perform as well as it does in patients with ISLs. Moreover, combining the DLST with other physical examination tests for SLAP lesions did not improve its diagnostic accuracy.
The results of our study differ from those of the 2 previous studies of this test. The technique that we used to perform the DLST was similar to the original description by Cheung and O’Driscoll (“The Dynamic Labral Shear Test for Superior Labral Anterior Posterior Tears of the Shoulder.” Presented at American Academy of Orthopaedic Surgeons, 2007) and Kibler et al. 21 In the study by Cook et al, 5 the patient was sitting or standing and the examiner stood in front of the patient. The patient’s arm was externally rotated 90° and then brought into 90° of abduction. The examiner maintained the external rotation and horizontal abduction and applied an anterior shear load while abducting the arm from 90° to 120°. Cheung and O’Driscoll and Kibler et al 21 described externally rotating the arm to tightness when putting the shoulder through a range of motion, whereas Cook et al 5 described maintaining 90° of external rotation throughout the range of motion.
Our results also differed from those reported by Kibler et al. 21 In their study, the patient population included a control group that did not undergo surgical exploration. However, their mean (±SD) patient age of 43.2 ± 12.6 years was similar to the mean ages in our cohorts (controls, 46.4 ± 16.7 years; ISL group, 33.9 ± 15.9 years; and CSL group, 42.5 ± 15.3 years), and the most common diagnoses in their study of 101 patients were rotator cuff tears (55%), SLAP tears (34%), and biceps injuries (23%). In that study, the clinical effectiveness of the test in patients with ISLs compared with patients with CSLs was not evaluated. Our study also differs from that of Kibler et al 21 because they studied an active patient population with a potentially greater likelihood of having SLAP lesions.
Many factors may have influenced the results of our study. There were few ISLs in this cohort, which could contribute to type II error. Our patient population may be older with more diverse shoulder lesions compared with other study populations. This is supported by the finding that the mean ages of the patients in our 3 groups were statistically different from each other (P = .026). In comparison, the mean (±SD) ages were 43.2 ± 12.6 years in the study by Kibler et al 21 and 45.3 ± 15.3 years in the study by Cook et al. 5 The results of our study might have been different if we had targeted a group of athletic individuals who performed overhead sports. Although some of our patients injured themselves while playing sports, activity level was not a parameter we measured. Similarly, an ideal control group would be patients without shoulder pain who underwent arthroscopic surgery, but that is not ethically possible.
Second, our results may have been influenced by the clinician who performed the examinations. Although the physician assistant in this study works with the senior author and was taught how to perform the test and interpret the results, there was otherwise no interobserver or intraobserver reliability testing performed. However, none of the previously published studies on the use of this test have reported interobserver or intraobserver reliability of the DLST.
Third, the surgeon was not blinded to the preoperative diagnoses of the patients. Although the surgeon did not normally recall the results of the preoperative examination at the time of surgery, inherent bias was present when the surgeon was not blinded to the procedure, the preoperative diagnosis, or the knowledge of what procedure was to be performed on the posting sheets and consents for surgery. Another source of bias was that all intraoperative findings were determined by 1 surgeon. However, a study has shown that intraobserver reliability for classifying SLAP lesions with arthroscopy is quite high when experienced shoulder surgeons, such as the surgeon in this study, make the diagnosis arthroscopically. 17 In that study, experienced shoulder surgeons had an intraobserver reliability of 0.670 and an interobserver reliability of 0.804.
The tests used in this study reflect the preferences of the senior author after studying the literature on this subject (Table 4).13,14 Many other physical examination tests are purported to provide a high degree of clinical utility for SLAP lesions, but those that have only 1 supporting study were not used by the senior author. The inability of these 4 tests to accurately diagnose SLAP lesions, as single tests or in combination, has been confirmed by several studies.11,13,26,27 Only 1 study, by Oh et al, 29 found that combining “major tests” and “minor tests” increases the accuracy of the physical examination for SLAP lesions, but their proposed schema has not been widely accepted or used in clinical practice. Also, this study did not include variables such as patient history, presence of pain, or the role of imaging in supporting the diagnosis. Kibler and Sciascia 20 pointed out that making the diagnosis of a SLAP lesion should include patient history, examination, imaging, and arthroscopic assessment when treating patients with suspected SLAP lesions.
Review of Diagnostic Accuracy for Shoulder Biceps and SLAP Lesions as Reported in the Literature a
DA, diagnostic accuracy; DLST, dynamic labral shear test; +LR, positive likelihood ratio; –LR, negative likelihood ratio; NPV, negative predictive value; NR, not reported; PPV, positive predictive value; SLAP, superior labrum anterior and posterior.
According to the system of Snyder et al. 33
The present study has several strengths. First, this is the largest cohort of patients evaluated prospectively by 2 experienced examiners. Second, the use of only 1 examiner who is an experienced shoulder surgeon would be expected to provide high intraobserver reliability for classifying SLAP lesions at the time of surgery, as suggested by Jia et al. 17 Third, arthroscopy was used as the gold standard, and the diagnosis was not confirmed with history or advanced imaging alone.
This study shows that the DLST has some clinical utility in making the diagnosis of ISLs. However, the high sensitivity and low specificity suggest that the DLST cannot reliably distinguish SLAP lesions from other shoulder abnormalities, such that the results of the DLST must be interpreted with caution. The utility of the DLST did not increase when it was combined with other commonly used tests for SLAP lesions, especially when other abnormalities were present. Further study of the DLST in larger patient groups is warranted.
Footnotes
The authors declared that they have no conflicts of interest in the authorship and publication of this contribution.