For patients who have had multiple surgeries, it is particularly important to obtain a good history, paying special attention to complications and to why the pain has not been alleviated. Similarly, the physical examination of this patient needs to be meticulous to understand what is contributing the most to her pain. The most concerning aspect of this patient’s history was the presence of wound drainage, so it is critical to evaluate for erythema, warmth, swelling, and drainage. Laboratory studies such as white blood cell count (WBC), sedimentation rate, and C-reactive protein should be obtained. MRI scanning can also be helpful in the assessment because, in addition to characteristics of the rotator cuff tear, findings ofeffusion, bone edema, soft-tissue edema, or debris in the joint are indicative of infection. Typically, patients who have a previous infection after rotator cuff repair will have little tendon tissue remaining to repair [1]. If there are concerns about infection, the patient should undergo repeat arthroscopic debridement , removal of anchor and suture material if possible, and intravenous antibiotics. Re-repair should not be attempted at the time of debridement, and decisions about re-repair can be made after the infection is cleared. However, in our experience, typically in patients with infections after rotator cuff surgery, the rotator cuff tendons cannot be repaired. Instead, it is imperative to diminish the patient’s stiffness.

The other confounding issue in this patient is the role of stiffness in her pain. It is important to establish whether the loss of motion is due to neural causes or not, so questioning the patient about paresthesias and neck pain is important. Similarly, the physical examination should include a complete neurological examination. The most commonly injured nerves with rotator cuff repair surgery are the axillary nerve and the musculocutaneous nerve. However, there can rarely be a brachial plexopathy caused by traction from arm suspension used in surgery or from regional anesthesia, specifically an interscalene block.

Evaluating the range of motion in a patient whose previous surgery has failed can be challenging if there is pain. If the patient has a large shrug sign [2] or a positive drop-arm sign [3], it is important to determine whether these are due to stiffness, pain, or mechanical problems with the shoulder. When testing range of motion, typically the patient is asked to perform the motions actively, such as elevation in flexion or abduction in the plane of the scapula. Often, patients will attempt to elevate their arms, which they proceed to do in the plane of their body with their thumbs down. This maneuver will undoubtedly produce less-than-normal motion, especially if the patient has secondary gain. In some patients, range of motion can be tested better with the patients supine because gravity is eliminated. If the patient’s active and passive motions are similarly diminished, then frozen shoulder, osteoarthritis, fixed glenohumeral dislocation, or avascular necrosis should be considered.

If the patient has not had a distal clavicle excision, the acromioclavicular (AC) joint should also be tested for signs of inflammation. Local tenderness of the AC joint is the most important test but the active-compression test, cross-arm adduction stress test, and the arm-extension test have all been shown to be helpful in localizing pain to the AC joint [4, 5].

Lastly, the patient should be examined to determine the extent of rotator cuff injury. The supraspinatus can be tested with resisted abduction and the infraspinatus with resisted external rotation at the side. The patient should be tested for an external rotation lag sign, which is performed with the arm at the side and the elbow bent 90°. The arm is externally rotated to the maximum and then internally rotated a few degrees. The patient is asked to hold the arm in that position, and if they cannot, the arm falls into internal rotation, then the patient has involvement of the supraspinatus and infraspinatus tendons .

The integrity of the subscapularis tendon should also be evaluated in several ways. The first is increased external rotation of the shoulder on the side with the subscapularis tear. The second is the liftoff test, in which the patient is asked to lift the arm off the back; inability to do this is a positive test [6]. A liftoff lag sign is also very helpful for determining integrity of the subscapularis tendon [7]. In this test, the patient places the arm in internal rotation up the back, and the examiner then lifts the hand off the back and asks the patient to hold it there. A positive test is when the arm falls to the back or the elbow falls into extension, and this means that the subscapularis tendon also is irreparable.

This series of examinations should give the examiner a good sense of what can be achieved with an individual patient. Imaging should be performed, beginning with plain radiographs, including a true anterior-posterior radiograph (i.e., Grashey view) [8] of the shoulder, an anterior-posterior view in internal rotation, and an axillary view. A scapular Y view does not add much to the evaluation because it is neither reliable nor reproducible and has been shown to have no effect on the clinical result [9]. A plain radiograph with superior humeral head subluxation or with erosion of the acromion and greater tuberosity suggests that the rotator cuff tear is long-standing and irreparable [10]. It is important to look for spurs consistent with osteoarthritis and for joint space narrowing, because cartilage loss may be contributing to the patient’s pain and limited motion.

MRI can be helpful to confirm the integrity of the tendons and, in the case of rotator cuff tears, the size of the tear. Multiple studies have confirmed that the best predictor of rotator cuff surgery is the size of the rotator cuff tear [11–13]. Muscle atrophy has also been shown to be a negative prognostic sign for success of rotator cuff surgery [11] (Fig. 21.3). MRI also can demonstrate edema in the bone or soft tissues, which might suggest infection; however, edema in the muscles can be seen after acute injury or with Parsonage-Turner syndrome [14].

In this patient, blood work would be helpful to rule out infection. Although the WBC is usually normal, if the sedimentation rate and C-reactive protein level are elevated, then aspiration of the shoulder could be considered. Long-standing infections of the glenohumeral joint are often accompanied by glenohumeral arthritis due to destruction of the cartilage by the infection. In a chronically infected shoulder, MRI can also show edema in the humeral head, debris in the joint, lymph node enlargement, joint effusion, and diffuse muscle edema [15].

Plain radiographs of our patient revealed a small spur on the humeral head but no glenohumeral joint-space narrowing. There was narrowing of the humeral head to acromial distance of 3 mm. Her WBC count, C-reactive protein level, and sedimentation rate were normal, and an aspiration of her shoulder was negative for cultures over 15 days. Her MRI showed tendon retraction of the supraspinatus and infraspinatus to the glenoid with atrophy and grade III–IV fatty infiltration of the muscles (grade III begins equal fat and muscle and grade IV begins more fat than muscle) using the system of Goutallier [16, 17]. The MRI showed no signs of chronic infection and only mild osteoarthritis of the glenohumeral joint.

The patient was treated without surgery using a program that included changing her expectations, avoiding things that aggravated her shoulder, cryotherapy, daily range-of-motion exercises, judicious use of NSAIDs, and occasional intra-articular cortisone injections.

She regained function of the shoulder for activities of daily living but was weak when using the arm overhead and away from her body. She retired and settled with her employer for work-related injuries. Ten years later, at last follow-up, she had nearly full range of motion and little pain. She was anticipating shoulder replacement in the future but her shoulder was not painful enough to warrant surgery.

Decision making for any patient with a rotator cuff tear, whether recurrent or not, involves many factors. The major variables shown to determine the success or failure of rotator cuff surgery are the size of the rotator cuff tear and the patient’s age [18–20]. Another factor that should be considered is whether the patient has pain; it is difficult to rationalize an operation with limited success when the patient has no pain. Whether it is the dominant side can be a consideration, with less compelling need for operating on a nondominant shoulder. The degree to which the patient has limitation of function, including activities of daily living, also is a factor. The goals of the patient in terms of returning to work or sport should be considered but the patient should have realistic expectations of what can be accomplished with further surgery. The health of the patient is an important consideration; in patients for whom an operation is risky such as patients with American Society of Anesthesiologists (ASA) classification of 3 or 4, the risk of death for better shoulder function requires special counseling and considerations.