Diagnostic Injections About the Shoulder

Edward McFarland, MD; Johnathan Bernard, MD, MPH; Eric Dein, MD; Alex Johnson, MD


J Am Acad Orthop Surg. 2017;25(12):799-807. 

In This Article

Rotator Cuff Abnormalities and the Subacromial Space

Presentation and Evaluation

Our understanding of the causes of rotator cuff tears continues to evolve, influencing our interpretation of the success or failure of injections for this condition.[2] Rotator cuff syndromes are thought to have both intrinsic causes (ie, tendon degeneration over time) and extrinsic causes (ie, "impingement" against other structures). The exact cause of this pain is unknown, however, so the interpretation of an injection that relieves pain is conjectural. In the shoulder, the tissue with the most pain fibers is the bursa, but pain fibers are also found in the biceps tendon, transverse humeral ligament, and rotator cuff tendons.[3]

The patient history alone may not accurately identify the source of shoulder pain, which is classically attributed to a rotator cuff syndrome. Typically, pain associated with rotator cuff disease occurs in the anterior and lateral shoulder. However, pain associated with other shoulder conditions, such as stiff shoulder, arthritis, and biceps pathologies, also can radiate into these areas on the deltoid.

Similarly, findings from a physical examination for rotator cuff syndrome can be difficult to interpret and may not establish the source of the pain. Studies of physical examinations of shoulders with symptomatic rotator cuff disease have shown that the most useful findings are weakness to manual muscle testing with resisted abduction or weakness to resisted external rotation with the arm at the side.[4] Other physical examination results that may help diagnose rotator cuff abnormalities are the drop arm sign, the external rotation lag sign, and a painful arc. Neer and Hawkins impingement signs have been inexact in the diagnosis of rotator cuff disease.[5] However, Park et al[6] found that patients who had weakness in external rotation and a painful arc had a 91% chance of having a rotator cuff tear.

Imaging for rotator cuff syndromes can rule out some pathologies and identify other symptomatic abnormalities. Plain radiography should be used to evaluate the osseous structures, the glenohumeral joint, and the subacromial space. Full-thickness supraspinatus tears are found in 90% of patients in which the distance between the humeral head and the undersurface of the acromion is <7 mm.[7]

Arthrography with or without CT, ultrasonography, or MRI can help evaluate the status of the rotator cuff. However, the presence of a rotator cuff tear does not necessarily mean it is the source of pain. According to a literature review by Reilly et al,[8] 26% of asymptomatic shoulders in patients with a mean age of 44 years exhibited full- or partial-thickness rotator cuff tears on MRI.

Because physical examination and imaging studies often fail to confirm one diagnosis over another, injections into the subacromial space can help differentiate pain at this site from pain originating from other locations. The role of diagnostic injections for impingement was first championed by Neer,[9] who described injection of a local anesthetic into the subacromial space (Figure 1). When the pain diminishes with the injection, the Neer test is thought to confirm the diagnosis of rotator cuff disease. The use of this test to diagnose rotator cuff syndromes is supported by the histologic studies of Soifer et al,[3] showing the largest concentration of nerve pain fibers in the shoulder occurs in the subacromial bursa.

Figure 1.

Illustration of a sagittal view of the shoulder showing injections into the subacromial space from posterior and lateral directions. A = rotator cuff, B = bursa. (Copyright Louis Okafor, MD, Baltimore, MD.)

Many clinicians have sought to verify the diagnostic value of the Neer test and its ability to predict outcomes. In a study of 153 cases of impingement syndrome, Park et al[10] showed a 67% reduction in pain after subacromial injection with lidocaine. In a study of 208 subjects, Cadogan et al[11] reported that a full-thickness tear of the supraspinatus was associated with a positive subacromial injection (odds ratio, 5.02). A study of 153 patients by Oh et al[12] showed a correlation between reduction in pain after subacromial injection and reduction in pain early in the postoperative period after rotator cuff repair.

In contrast, Kirkley at al[13] found a poor correlation between a positive Neer test and reduction in pain after a subacromial decompression. One factor thought to account for variable results with subacromial injections is that the injection sometimes does not reach the subacromial space[14–18] (Table 1).

Technique Summary

Three sites are used for subacromial injections: anterior, lateral, and posterior (Figure 2). When given an injection, the patient should sit up straight on the side of the table with the arm hanging unsupported, which opens the subacromial space. In some patients, the arm may be more relaxed when it is placed in the lap; this prevents muscle forces from closing down the subacromial space, which makes it difficult to deliver the medicine to the bursa. In certain patients, the use of a topical anesthetic spray may help diminish anxiety and pain.

Figure 2.

Photograph of the shoulder showing posterior injection into the subacromial space.

The anterior approach is the most challenging because of the presence of the coracoacromial ligament anteriorly and anterior acromial spurs in some patients. In contrast, the lateral injection site allows easy access to the subacromial space and provides flexibility for addressing both anterior and posterior abnormalities. The edge of the acromion can be palpated easily, and the soft spot between it and the greater tuberosity can be palpated in most patients. When this approach is used, the needle should be aimed just below the lateral and inferior border. When the needle is in the subacromial space, the injected anesthetic should flow easily, without resistance.

When pain is primarily posterior, injection into the posterior site can be used occasionally. However, because abnormalities are typically found in the anterior part of the shoulder, the needle should be inserted far enough to cover this area as well.

The choice of anesthetic depends on the goals of the injection; a short-acting lidocaine, a longer-acting bupivacaine, or a combination can be used. After the anesthetic is injected, a few minutes are needed to determine a positive response (ie, pain relief). When simultaneous injection of local anesthetic and corticosteroid relieves pain in the short term, the cortisone presumably was delivered to the site of the pathology.

In a study of 26 patients who received injections in the subacromial space for impingement, Skedros and Pitts[19] reported that considerable pain relief occurred within 10 minutes in 42% of patients and after 10 minutes in an additional 35% of patients. The reduction in pain from a subacromial injection is believed to reflect a bursal source of pain in patients with rotator cuff syndromes. However, in patients with full-thickness rotator cuff tears, some relief may occur because of infiltration of the anesthetic into the glenohumeral space.

The use of ultrasonography to verify the location of the subacromial injection depends on the experience and comfort of the person administering the injection. In a systematic review of four cadaver studies and nine clinical studies, Aly et al[20] showed no difference in accuracy of injections into the subacromial space when performed with or without ultrasonography guidance.