Rehabilitation Program
Physical Therapy
Although studies have shown the efficacy of physical therapy, no current evidence has suggested that physical therapy alone improves function in the treatment of FS. [2, 27, 28, 29] However, physical therapy associated with an intra-articular injection of corticosteroid improves function and ROM more rapidly than does intra-articular corticosteroid injection alone. [2, 3]
Therapeutic exercises
Although many therapeutic exercises are described, few have been evaluated in the treatment of FS. Therapeutic exercises that have been studied include articular stretching and pulley therapy. [4, 5, 6] Passive articular stretching exercises improve ROM. The superiority of supervised versus home exercise programs has yet to be demonstrated. [7]
In an assessment of nonsurgical management of 75 patients with FS, Russell and colleagues found that individuals who participated in a hospital-based group exercise class achieved greater improvement in shoulder symptoms than did individuals who participated in individual physical therapy or in home exercises alone. It was also found that patients in the group exercise and individual physical therapy programs showed greater improvement in range of motion and in their Hospital Anxiety and Disability Scale (HADS) anxiety scores than did the home-exercise patients. [30]
A prospective, randomized, controlled trial by Mueller et al indicated that in patients with type 2 diabetes, mild to moderate shoulder symptoms can be significantly improved with a progressive, active movement program. Patients who underwent movement intervention demonstrated a 12.7-point improvement on the Shoulder Pain and Disability Index (SPADI) score over those who underwent a wellness program, with the movement intervention group still scoring higher at 9-month follow-up. However, although after completing the program, the shoulder movement patients had greater active shoulder flexion than did the wellness intervention group (by 7.2 degrees), flexion was comparable between the two groups at later follow-ups. [31]
Manual therapy
Data from 2 studies support the use of manual therapy to improve ROM in the short term. One study showed that passive mobilization in the end-range position of the glenohumeral joint was more effective for improving ROM and function than was passive mobilization in the pain-free zone. [32] However, the overall difference between the interventions was small. In addition, patients appeared to achieve greatest improvement in ROM when treatment was administered early. [33]
In summary, the findings indicated that patients with FS improve with physical therapy regardless of when it is administered after the onset of FS but that they achieve the greatest improvement in their ROM when treatment is administered early.
Physical modalities
Many electroanalgesic and thermoanalgesic modalities are often used in physical therapy. Prospective, randomized, placebo-controlled clinical trials have demonstrated the inefficacy of bipolar interferential electrotherapy, TENS (transcutaneous electrical nerve stimulation) pulsed ultrasound therapy, laser therapy, and magnetotherapy with electromagnetic fields in the treatment of painful shoulder disorders. Indeed, a systematic review of randomized clinical trials has demonstrated that little evidence supports the use of common physiotherapeutic modalities to treat shoulder pain. (Nonetheless, clinical observations suggest that electroanalgesia and heat can provide some temporary relief to patients with severe shoulder pain and that they can do so without side effects.)
In summary, physical therapy alone has not been shown to improve function. However, when associated with an intra-articular corticosteroid injection, physical therapy improves ROM and function more rapidly than does intra-articular corticosteroid injection alone. A study by Klç et al indicated that in patients with FS, combining physical therapy with a suprascapular nerve block leads to greater improvements in pain and functional status than does physical therapy alone. [34] The effectiveness of physical modalities has not been demonstrated. However, therapeutic exercises and manual therapy do improve ROM (see above).
Given the paucity of evidence supporting the effectiveness of physical therapy, the present authors believe that clinicians should be most conservative in designing a physical therapy program for patients with FS. An intermittently supervised, home-based articular stretching and strength-maintenance program can be combined with manual therapy, the appropriate analgesic medications, and the judicious use of electroanalgesia and warm packs. This approach generally suffices to provide adequate pain relief in most patients while the disease runs its usual, favorable course. (For additional information, see Home Exercises for the Stiff Shoulder, on the University of Washington School of Medicine Web site.
Occupational Therapy
Patients with severe FS may benefit from a referral to an occupational therapist for assistance and instruction in performing activities of daily living (ADLs). The occupational therapist helps the patient learn how to use adaptive equipment and suggest home and workplace modifications that may be necessary and beneficial for completing professional activities and routine daily tasks (eg, dressing, bathing, grooming). (See also Patient Education.) However, the effectiveness of these interventions has yet to be demonstrated. Researchers who performed a systematic review concluded that evidence supporting the effectiveness of rehabilitation in the workplace is lacking. [35]
Medical Issues/Complications
Certain authors, including the present authors, believe that idiopathic FS should be considered a distinct clinical entity. The prognosis and therapeutic approach for idiopathic FS differ from those of posttraumatic shoulder stiffening resulting from a fracture or a surgical insult to the glenohumeral joint.
Surgical Intervention
Duplay, the first person to describe the syndrome of FS, in 1872, proposed treating this condition with manipulation of the glenohumeral joint, with the patient under general anesthesia. Although some orthopedic surgeons continue to practice this technique, the benefits of this approach have not been demonstrated in controlled clinical trials.
In 1972, Hazelman noted that, in the context of a prospective controlled clinical trial, patients who had been treated with manipulation of the glenohumeral joint fared no better than did patients who received only a series of steroid injections and physical therapy. [36] Data from a long-term prospective study by Binder and colleagues seemed to support the notion that subjects who are treated with vigorous shoulder manipulation fare worse than do those who are not treated at all. [19]
A 2009 study by Jacobs et al also found no evidence that manipulation provides a better treatment outcome in FS. [37] The investigators divided 53 patients suffering from idiopathic FS into 2 groups, with one undergoing shoulder manipulation under anesthetic and the other being treated with intra-articular steroid injections and distention (see Other Treatment). After 2 years of follow-up, the investigators found no statistical difference between outcome measures in the 2 groups.
Various improvements in surgical techniques, such as the advent of controlled capsular release by using arthroscopic access to the anterior glenohumeral joint capsule and the coracohumeral ligament, appear to offer promising treatments. However, the effectiveness of these surgical techniques has yet to be demonstrated in controlled clinical trials.
Considering the favorable prognosis for patients with idiopathic FS, surgical intervention should probably be reserved for rare patients whose condition does not respond to maximal conservative modalities implemented over a sufficient period of time. [8]
Some authors propose early surgical intervention in patients with FS because of the patients' high risk of permanent contracture of the glenohumeral joint. However, this approach remains controversial because patients with diabetes often have recurring pain and contracture in the postoperative rehabilitation phase and because their rate of preoperative complications is generally increased.
Other Treatment
Studies of intra-articular and intrabursal injections, glenohumeral distention arthrography, and nerve blocks have been conducted, as described in the following text.
Intra-articular and intrabursal injections
Intra-articular corticosteroid injections
Intra-articular injections of steroid derivatives are the second most common medical intervention for treating painful joint conditions. (The most common intervention is the administration of nonsteroidal anti-inflammatory drugs [NSAIDs]).
A blinded, randomized, controlled study by Yoon and colleagues indicated that high- and low-dose intra-articular corticosteroid injections for FS are equally effective. The study involved 53 patients with primary adhesive capsulitis in the freezing stage. Twenty patients received a single ultrasonographically guided intra-articular injection of 40 mg of triamcinolone acetonide, while another 20 received an ultrasonographically guided intra-articular injection of 20 mg of the drug and 13 received a placebo. The injections were followed by participation in a home exercise program, with posttreatment follow-up at 1, 3, 6, and 12 weeks. The investigators found that there was greater improvement in shoulder pain and passive range of motion in the high- and low-dose groups than in the placebo patients but also that the level of improvement was about the same in both corticosteroid groups. [38]
The rationale for injection of an intra-articular steroid derivative is to allow direct delivery of a modest dose of concentrated drug with analgesic and anti-inflammatory properties to the targeted site of pathology. [39] Researchers at the Manchester Rheumatology Service investigated the accuracy of a variety of joint injections that were administered by using anatomic landmarks to the upper and lower extremities. Their study demonstrated inaccurate placement of the drug in 65% of 108 joints injected.
A study by Eustace and colleagues showed results similar to those above, demonstrating that 68% of shoulder injections performed by expert specialist physicians without radiologic guidance failed to hit their target. [40] Overall, when radiologic guidance is not used, the accuracy of these injections is poor. When anatomic landmarks are used, the rate of misplacement is as high as 58% in the glenohumeral joint and as high as 69% in the subacromiodeltoidian bursae. Accuracy of the injection is associated with improved clinical results. [40, 41, 42]
If the rationale for using an intra-articular injection of steroid derivative is to deliver targeted therapy, the target in FS is the glenohumeral joint. The work by Eustace and coauthors suggests that targeted therapy that does not hit its target is not ineffective. [40] In general, however, physicians who are less experienced in joint injection than the expert rheumatologists in that study should avoid injecting shoulders without radiologic guidance. The results of their study are impressive, even for physicians who are experienced in the field of joint injections.
Eustace's study also puts an interesting twist on the interpretation of data from placebo-controlled clinical trials that failed to demonstrate favorable results with the steroid injection into the shoulder to treat FS. Some investigators did use radiologic guidance, but most employed anatomic landmarks. [2, 43, 44, 45, 46] Because some studies involved radiologic confirmation of location of the injectate, most of the injections may have missed their target. The present authors believe that no conclusions can be drawn from studies using anatomic landmarks, given the inaccuracy of the intervention.
On the other hand, a study by Ranalletta et al found that in patients with FS, a single intra-articular corticosteroid injection performed without imaging control, prior to the start of a physical therapy program, produced faster pain relief and quicker improvement in shoulder function and motion than did treatment with oral nonsteroidal anti-inflammatory drugs (NSAIDs) and physical therapy. The study found that patients in the corticosteroid group experienced faster improvements in pain, function, and motion during the first 8 weeks posttreatment than did the NSAID group. However, by final follow-up, at 12 weeks, there was no significant difference between the two groups regarding these factors. [47]
From the current literature, we know that intra-articular corticosteroid injections or distention arthrography with corticosteroids or sodium chloride solution (discussed below) considerably improve FS in the short term. However, in the long term, they are not superior to placebo but they are superior to supervised physical therapy. [2, 48] Whether their effectiveness is the result of distention arthrography, the intra-articular corticosteroid, or their combination is unclear because most studies used corticosteroids.
Intrabursal corticosteroid injections
Many authors have shown that the pathology of FS is in the extra-articular structures, such as the coracohumeral ligament, the interval of the rotator cuff, the subacromial space, and the articular capsule. [9, 10, 11] From a pathophysiologic point of view, a rationale supports the use of corticosteroid injection in the subacromiodeltoidian bursa to treat FS.
A comparison of intra-articular and intrabursal injections with lidocaine alone or with lidocaine and a corticosteroid showed similar decreases in pain and increases in ROM in all groups. [6] However, a significant short-term decrease in pain was observed in the lidocaine-plus-corticosteroid group compared with the lidocaine-only group. Data from another study supported these findings, showing that patients whose condition did not improve with distention arthrography benefited from an intra-articular bursal injection of corticosteroid. [49]
Because of the inaccuracy of injections given by using anatomic landmarks, the present authors recommend that intra-articular and intrabursal injections should be administered by using fluoroscopic guidance. When the needle is accurately positioned, intra-articular corticosteroid injection and distention arthrography are effective in the short them. However, in the long term, they are not superior to placebo. This observation is not surprising given the favorable evolution of this disease. An intrabursal injection of corticosteroid can be used as a complementary treatment to intra-articular injection and distention arthrography or as treatment in patients whose condition does not improve with an intra-articular corticosteroid injection or with distention arthrography.
Many issues remain unclear. Factors to be resolved are whether distention is necessary, whether an intra-articular corticosteroid injection without distention (or vice versa) is enough, how many injections are needed, the stage of disease at which injections should be administered, the most effective corticosteroid, and the most effective dosage.
Intra-articular injection of sodium hyaluronate
Sodium hyaluronate has a metabolic effect on the articular cartilage, synovial tissues, and liquid. Few studies on the effect of sodium hyaluronate have been reported. [50, 51, 52, 53] In one controlled study, sodium hyaluronate was as effective as an intra-articular corticosteroid injection or physical therapy in improving function, but it was less effective than the comparators in improving ROM. [50] In another randomized controlled trial involving 70 patients with FS, the addition of intra-articular hyaluronate injections to physical therapy produced no added benefit [53] . Although additional studies are needed before conclusions about efficacy can be drawn, intra-articular injections of sodium hyaluronate may be an alternative treatment for FS, mainly in patients in whom corticosteroid injections are contraindicated.
Glenohumeral distention arthrography
One controlled study showed no benefit to distention arthrography over intra-articular corticosteroid injection without distention. [54] Another controlled study showed a significant increase in ROM and a significant decrease in the use of analgesics after distention arthrography, compared with the use of intra-articular corticosteroid injection alone. [55]
The benefit of performing distention arthrography until the capsule ruptures must be demonstrated. [6, 56, 57, 58]
One uncontrolled study that was designed to examine the optimal number of distention arthrography procedures that should be performed showed that 2 procedures administered within 3 weeks, when combined with home exercises, significantly improved function. However, a third procedure offered no benefit. [59]
Another unknown factor is the stage at which infiltration should be performed. To the authors' knowledge, only 1 uncontrolled study on this question has been completed. [44] The researchers concluded that distention arthrography should be done in the second stage of disease that is not progressing, despite the patient's participation in physical therapy.
Nerve blocks
Suprascapular nerve block
The suprascapular nerve block is a simple procedure but is not well known by most clinicians.
Dangoisse's technique for suprascapular nerve block was modified to render it steroid free and accessible to most physicians who practice musculoskeletal medicine in ambulatory-care facilities and private offices. A 3.75-cm, 25-gauge needle is directed in the plane of the scapula toward the center of the floor of the supraspinous fossa. The needle-insertion point is 2 cm above the bisection point of the upper border of the spine of the scapula. After aspiration is performed to rule out intravascular needle placement, 10 mL of bupivacaine 0.5 is slowly injected into the floor of the supraspinous fossa to fill the fascial contents of this fossa and to produce an indirect suprascapular nerve block. The highly concentrated bupivacaine bathes the suprascapular nerve as it enters the fossa through the suprascapular notch.
Since 1996, thousands of bupivacaine suprascapular nerve blocks have been performed by using the above-described technique. No notable complications, other than vasovagal episodes and tenderness at the site of injection, have occurred. Most specialists in musculoskeletal medicine who treat shoulder pain in an outpatient context can safely and inexpensively perform modified indirect suprascapular nerve blocks without radiologic guidance if they are adequately trained.
A randomized controlled trial was performed to compare the effectiveness of a single suprascapular nerve block with that of series of intra-articular corticosteroid injections given by using anatomic landmarks without fluoroscopic guidance. [60] Pain decreased and ROM increased more rapidly and more completely with the nerve block than they did with the intra-articular injections.
A randomized, double-blind, placebo-controlled study by Shanahan et al indicated that a suprascapular nerve block, when added to proven therapies, can lead to, on average, an approximately 6-month decrease in the duration of FS. In patients with FS who received a suprascapular nerve block, it took a mean period of 5.4 months for the condition to resolve, compared with 11.2 months for the placebo group. At 9 months, the total Shoulder Pain and Disability Index (SPADI) score was 39.2% for the placebo patients (down from a baseline of 67.2%), and 11.5% for those who received the nerve block (down from a baseline of 63.1%). In addition to suprascapular injections (using bupivacaine plus methylprednisolone, in the non-placebo group), patients received an injection of triamcinolone and local anesthetic into the glenohumeral joint and participated in a physical therapist–directed exercise program. [61]
A double-blind, placebo-controlled, randomized trial showed the superiority of 3 bupivacaine suprascapular nerve blocks. [62] At 1 month, a 64% decrease in pain was reported in the bupivacaine group compared with 13% in the placebo group. Although functional improvement was superior with bupivacaine, the difference was not statistically significant. The same authors preformed a double-blind, placebo-controlled, randomized trial larger than the previous study by using the same protocol (unpublished data). They observed no statistically significant difference between bupivacaine and placebo in terms of pain and function.
Some studies have shown the effectiveness of bupivacaine suprascapular nerve blocks to treat chronic tendinopathy of the rotator cuff, as well as to treat painful shoulders in patients with rheumatoid arthritis. [63, 64, 65] However, studies are scant, and the results are contradictory regarding the effectiveness of bupivacaine suprascapular nerve blocks in managing FS. Additional studies are needed before conclusions can be drawn.
Stellate block
An uncontrolled study of the combined effect of electropuncture, stellate block, and suprascapular nerve block showed that the combination improved pain control and increased ROM more than electropuncture or nerve block alone. [66]
