In my clinical practice, I evaluate active range of motion of both shoulders even before I perform the common semeiological tests. The assessment can be made with the patient standing or sitting. Personally I prefer the latter position since the compensatory movements of the spine and pelvis are reduced. Commonly I write down in the clinical record the degrees of flexion, abduction, and external and internal rotations. The external rotation is performed starting from the position with the limb along the side or in abduction to 90° (in both cases the elbows are flexed at 90°), while the internal rotation is performed with the arm moved to the back (the vertebral level reached by the thumb is reported). The motion degrees are registered with a goniometer or by performing the movements behind a wall where a graduated circle hangs. The decision of performing the active range of motion evaluation at the beginning of the visit is motivated by the need to transcribe in the clinical record the information that is not affected by the pain which could appear/increase during the execution of some tests. If patient agrees, I take some photos of the extreme degrees reached. This documentation may be useful for future medical lax disputes (Fig. 7).

In all cases, I observe also the active range of motion of the joint between the scapula and the toracic surface. Any dyskinesias are clearly visible on the return from the position of maximum flexion. The movement of the scapula can also be reduced or altered as the result of trauma, adhesions, or neoformations.

This phase ends with the assessment of the passive mobility while patient is first sitting and then supine. This allows the examiner to appreciate any stiffness, intra-articular or pathological snaps.

The main flexors of the shoulder are the deltoid (anterior portion) (axillary nerve, C5-C6), the pectoralis major (clavicular portion; lateral pectoral nerve, C5-C6), and the coracobrachialis and biceps (musculocutaneous nerve, C5-C7). The evaluation is performed by placing a resistance to about 90° of flexion.

This movement is possible through the action of the lateral portion of the deltoid (axillary nerve, C5-C6) and the supraspinatus muscle (suprascapular nerve, C5). The resistance is applied when the limb is abducted to 90°, slightly anteponed of about 20° and in full pronation.

This movement is realized through the infraspinatus (suprascapular nerve, C5-C6) and teres minor muscles (axillary nerve, C5-C6). The posterior portion of the deltoid participates only minimally (axillary nerve, C5-C6). The assessment can be performed with the limb at the side or abducted to 90°. In both cases, the elbow is flexed to 90°.

The main internal rotator muscles are the subscapularis (upper subscapular, C5; and lower subscapular nerves, C5-C6) and the pectoralis major (lateral pectoral nerve, C5-C7; medial pectoral, C8-T1). The resistance is applied to the volar surface of the forearm while the arm is at the side, the elbow is flexed and in neutral rotation.

Any muscular deficit must be quantified and reported in the medical card. The classification of muscle strength we use is shown in Table 2.

Soft tissues, such as tendons, capsule, and ligaments, can be rubbed or compressed by the surrounding bony epiphysis or apophysis. The resulting symptoms give rise to the so-called impingement syndrome. Among these, the subacromia one is the most common. In this case, because of the acquired narrowness of the subacromial space, the upper cuff tendons touch the antero-inferior margin of the acromion during the arc of movement between 60° and 120° of flexion.

Patients with subacromial impingement, without rotator cuff tear, are usually older than 40 years old. Younger patients often have posterior capsular contracture or lesser forms of instability that simulate the clinical picture of the subacromial impingement syndrome. Male subjects engaged in manual labor (carpenters, painters, masons, builders, etc.) are statistically more involved. Pain is localized at the front, under the anterior acromial margin and radiates along the anterior and lateral surface of the arm. The appearance or intensification of pain during flexion of the arm is a distinctive feature, in particular, during the range of motion between 60° and 120° of flexion (painful arc). Nocturnal pain is not constant, but it is much more common if the impingement has caused a degeneration of the bursal side of cuff tendons. Shoulder range of motion is generally complete but painful during flexion (painful arc). However, mobility will be reduced if the impingement is associated with adhesive capsulitis or with a cuff tear, even partial one.

There are three tests described in the literature to reveal this form of impingement: Neer’s test, Neer’s sign, and Hawkins’ tests.

Neer’s test consists of injecting about 10 cc of xylocaine 2 % in the subacromial space. Generally, the anesthetic is injected through the path that corresponds to the posterior arthroscopic portal (1.5 cm inferiorly and medially to the posterolateral corner of the acromion). The tip of the needle is directed superiorly and addressed to the anterolateral margin of the acromion. After the infiltration, patient is observed in the next 3 h and then asked to make movements that would normally cause pain. The test is positive when the pain decreases by at least 80 %. However, a positive test indicates that the structures present in the subacromial space are suffering, but the suffering is not necessarily caused by impingement.

Personally, I do not ever run this test because it will be positive even in the case of a subacromial bursitis without impingement, of an acromioclavicular joint arthropathy and of rotator cuff tears. Moreover, some patients, although rarely, may develop an allergic reaction to the anesthetic; so I consider taking any risks to the patient useless.

Neer’s sign [5] is positive when the examiner, passively flexing the shoulder, causes anterior or deltoid pain over movement between 60° and 120°. Pain is much more intense, and the sign most sensitive, if the examiner, in addition to flexing the limb, exerts a downward pressure of the anterior margin of the acromion [6] (Fig. 8a).

Hawkins’ test [7] consists of flexing the arm to 90°, while the elbow is flexed, and then performing an internal rotation and an adduction of the shoulder. In this position, and in case of a pathological narrowness of the acromio-humeral distance, a painful impingement occurs between the upper portion of rotator cuff tendons and the acromion (Fig. 8b).

Results of a prospective study showed that the sensitivity of Neer’s sign of and Hawkins’ tests for the diagnosis of subacromial bursitis is, respectively, 75 % and 92 %, while that for rotator cuff tear is 85 % and 88 % [8].

In rare cases, subscapularis tendon can be compressed between the lesser tuberosity and the apex of the coracoid (subcoracoid impingement). When it occurs, there is always a combination of anatomical constitutional factors (long or with too much inclination coracoid process; restricted coraco-glenoid space); acquired (calcific tendinitis of subscapularis tendon) or iatrogenic (e.g., results of surgery for joint stabilization according to Latarjet) [9, 10]. Usually patients feel pain during flexion and internal rotation of the limb (driving a car, writing on a board). There are three tests commonly used to detect this form of friction (Fig. 9):

The first is that of passive flexion–internal rotation that arouses or increases pain, because in this position the distance between the apex of the coracoid and the lesser tuberosity becomes smaller (Fig. 9a).

The anesthetic infiltration test (5 cc xylocaine 2 %) is the most sensitive. The xylocaine is introduced just laterally to the coracoid apex. The test is positive if, after infiltration, flexion and internal rotation do not cause more pain. The effect of the anesthetic is exhausted in 3–24 h (Fig. 9b).

The last test is the “reduced motion in flexion–adduction–internal rotation”. With the hand of the affected side, patient is not able to touch the spine of the controlateral scapula; the same movement they can do easily with the hand of the healthy side (Fig. 9c).

Recently, after a dynamic arthroscopic evaluation performed on 16 patients with chronic pain caused by flexion and internal rotation of the shoulder, Gerber and Sebesta [11] showed an internal impingement between biceps long head, upper glenohumeral ligament, coracohumeral ligament, and superior glenoid labrum. This occurs when limb is flexed more than 90°; if the degree of flexion is inferior, the impingement may involve the subscapularis tendon, the superior glenoid labrum, and the bony margin of the glenoid.

In patients involved in throwing activities, pain caused by abduction and full external rotation of the shoulder should raise the suspicion of an impingement between the deep portion of the supraspinatus tendon and the postero superior glenoid joint line [12].

Among the tests designed to detect a lesion of the subscapularis tendon, the lift–off test is the best known. The patients individually moved the back of the hand to the lumbar region (L2-L5). This may not be possible when shoulder is markedly painful or stiff (adhesive capsulitis). After that, they are invited to shift his hand from the trunk. If it fails, the test is considered positive [13]. Usually I prefer to run the internal rotation lag sign (IRLS) [14]. The examiner grasps the patient’s hand by the thumb and moves it back toward the lumbar region (L2-L5), keeping it departed from the trunk. He/she then invites the patient to independently maintain this position (Fig. 10a). If the back of the hand “falls” on the lumbar spine, the test is considered positive (Fig. 10b). The IRLS results more sensitive than lift-off, as it can detect even partial tears [15].

In 1996, Gerber et al. [16] described a test to be performed in case of marked painful stiffness in internal rotation. The test was called belly press test, and it is performed with the arm along the side, the shoulder in internal rotation, and the elbow flexed to 90°. The patient presses the abdomen with the palm of his hand while keeping the limb internally rotated. The examiner assesses the resistance offered when trying to deviate the hand from the abdomen of the patient. The test is considered positive in case of evident weakness compared to the controlateral limb, or when patient, in an attempt to oppose the force of the examiner, extends the shoulder. Napoleon’s sign (by analogy with the position taken by the famous historical character) is a variant of the belly press test. Barth et al. [17] consider the test as positive, when patient, while pressing his hand on the abdomen, does not keep the wrist (radio -carpal joint) in line with the forearm, but flexed it to 30–60° (intermediate deficit) or to 90° (severe deficiency) (Fig. 10c, d). For this test, the evaluation is performed by comparison. The lesion of the subscapularis tendon is also suspected when there is an increase in the passive external rotation compared to the healthy side [18].

Beaudreuil et al. [19] performed a systematic review of the literature. The authors noted that the “lift off” test has a high specificity (range 85–100 %) and a lower sensitivity (range 59–62 %). Hertel et al. [15] believe that the IRLS has a specificity of 96 % and a sensitivity of 97 %.

The bear–hug test is the test from which I get the most reliable information [17] (Fig. 10e, f). The palm of the hand of the involved side is placed on the opposite shoulder. Fingers are extended and the elbow is placed in front of the body. The examiner grasps the patient’s hand and tries to lift it by applying to the forearm a perpendicular force in external rotation. The patient should oppose this attempt to deviate the hand. The test is considered as positive when patient fails to keep the palm of the hand on his shoulder. Barth et al. [17] believe that the bear-hug test is not very sensitive (60 %); however, it is definitely more sensitive than the lift off (18 %), belly-press (40 %), and Napoleon tests (25 %). In contrast, the specificity of the four tests is very high: bear-hug test: 92 %; lift off: 100 %; Napoleon and belly-press tests: 98 %. On the basis of intraoperative findings, the authors argue that the positivity of the bear-hug and belly-press tests suggests a lesion of at least 30 % of the subscapularis tendon, while a positive Napoleon test indicates that more than 50 % of the subscapularis is injured. In addition, the lift off test becomes positive when at least 75 % of the tendon is lesioned. In conclusion, the bear-hug test increases the chance to appreciate even the smallest of subscapularis tears, but only the execution of all four tests helps to predict the size of the tear.

Jobe and Moynes [20] observed that the supraspinatus’ function can be appreciated singly during the elevation of the arm, abducted to 90° (in the scapular plane), anteponed of 20°, and completely internally rotated (Jobe test or Empty can test). According to Jobe and Bradley [21], the test is positive when there is a strength deficit secondary to the supraspinatus tendon tear or if it causes pain as a result of a subacromial impingement (Fig. 11a). However, test interpretation is controversial [15–22]. EMG studies showed that the activity of the supraspinatus is the same whether the limb is placed in internal or 45° of external rotation (Full can test, Fig. 11b); however, in external rotation the test causes less pain [22]. Itoi et al. [23] found that muscle weakness is the only parameter to be considered, because, in contrast to pain, it does not reduce the specificity of the test. In addition, they observed that both tests are equivalent in terms of accuracy. Therefore, considering that Full can test causes less pain, it should be preferred to Jobe’s one. Our data indicate that Jobe’s test has a high sensitivity and low specificity. This means that if the test is carried out on a group of both healthy subjects and patients, it is able to identify all true positives; however, it will be positive even in some healthy subjects.

Itoi et al. [23] believe that Jobe test has a 77 % of sensitivity for the strength, while the specificity, positive and negative predictive value, and accuracy are 68 %, 44 %, 90 %, and 70 % respectively. The same values for the Full can test are: 77, 74, 49, 91, and 75 %.

Data of a clinical and MRI study [24] indicate that the Jobe and Full can tests specificity is higher if we exclude partial tears.

Gillooly et al. [25] prefer to perform Jobe test with the arm abducted to 90° in the coronal plane and in internal rotation so that, with the elbow flexed to 90°, the fingers point downwards and thumb is directed medially. This test, called Lateral Jobe, is positive when the examiner, exerting against resistance a force directed downwards and applied on the arm, feels a weakness. According to the creators, the “Lateral Jobe” and “Jobe” tests have 81 % and 58 % of sensitivity, respectively, while the specificity is nearly equal: 89 % and 88 %.

When there is a lesion of the supraspinatus tendon, the strength in abduction may be also compromised. In order to evaluate it, the examiner stands behind the patient. With one hand stops the scapula rotation and with the other one evaluates the strength over movement between 45° and 90° of abduction (Fig. 11c). Data of our unpublished study indicate that the sensitivity, specificity, and positive and negative predictive value of the test are 74 %, 91 %, 75 %, and 88 %, respectively.

Wolf and Agrawal [26] described a method to appreciate cuff tear with a transdeltoideus palpation (Rent test). The examiner stands behind the patient. With one hand he holds the elbow flexed to 90°, and with the index finger of the other hand palpates the subacromial region. Then he extends the arm and executes internal and external rotation movements. If there is a tear, the index palpates the greater tuberosity and an “empty” area, corresponding to the failure tendon insertion on the tuberosity (Fig. 11d, e). The sensitivity of the positive predictive values for this test was 95 %, while the specificity and negative predictive value were 96 %. These values are questionable because, as one might guess, the test can be performed only in slim patients with a poor representation of the deltoid muscle.

If there is a suspicion that the patient has posterior-superior cuff pathology, it is good practice to perform, after the joint function evaluation, tests that do not cause pain and, therefore, do not affect their validity. Among these, the external rotation lag sign (ERLS) and the drop sign [15] are the most used. During the execution of both tests, the examiner stands behind the patient, who is sitting. The ERLS (for supra and infraspinatus) consist of supporting (from the elbow) the limb under examination in slight abduction and flexion (20°) and close to maximum external rotation (the maximum external rotation would result in a physiological elastic return after releasing the arm). The examiner should check that patient does not rotate the trunk. Then, the patient is asked to maintain independently this position. The test is positive when at least 10° of external rotation are lost (Fig. 12a, b). The interpretation of this test can be difficult when the passive external rotation is reduced (capsulitis) or increased (subscapularis tear). In these cases, it is possible to obtain, respectively, false negatives and false positives.