The Painful Shoulder: A Practical Approach

Emerg Med 37(2):20-32, 2005

What set off the pain? Is it attributable to bursitis, tendinitis, arthritis, a torn rotator cuff, or something more unusual? Does the patient need diagnostic imaging, a specialist referral, physical therapy? The author walks through the processes of examining, assessing, and treating a painful shoulder.

By Allan V. Prochazka, MD, MSc

Persistent shoulder pain is a very common problem in adult patients. More than 9% of men and 12% of women over age 15 will be affected at some time during their lives. A recent survey in the Netherlands found that the prevalence of shoulder pain was 21% in adults; in those aged 60 to 64 years, it increases to more than 40%. Most studies tend to show a higher prevalence among women than men.

Shoulder complaints are associated with marked interference with daily functioning and are a common presenting complaint in acute care settings. In certain occupations that involve a lot of pushing and pulling with the arms, shoulder pain is a frequent work-related injury. Many recreational activities, such as canoeing or baseball, are also associated with shoulder injuries. Given the great mobility of the shoulder joint, many activities, such as combing one's hair, reaching for something in a cabinet, or putting on a shirt can be problematic in the presence of shoulder pain.

In many cases of shoulder pain, there is an obvious trigger to the pain, such as a recent fall on an outstretched arm. However, older patients often present with an insidious onset of pain without any obvious recent injury.

Consider the following case: A 67-year-old man presents with a two- to three-month history of left shoulder pain. The pain is localized mostly over the deltoid region, worsens with overhead movement, and bothers the patient at night. The patient states that he has tried ibuprofen and also glucosamine/chondroitin without much relief. There is no history of recent trauma, although the patient reports that lately he has been working in his yard. He has no history of prior shoulder problems but does have some osteoarthritis in both knees.
 

EXTENSIVE DIFFERENTIAL DIAGNOSIS

We will revisit the above patient at various points in this article, but first the differential diagnosis for shoulder pain, which is extensive and includes both acute and chronic conditions that directly involve the shoulder (see box below), as well as referred pain, will be discussed. Studies of large numbers of adults with persistent shoulder pain have found that about 60% will have subacromial bursitis or supraspinatus tendinitis; 12%, adhesive capsulitis; 10%, supraspinatus tear or rupture; 7%, acromioclavicular (AC) arthritis; 5%, bicipital tendinitis; and 7%, other causes. To complicate matters, more than one condition can occur simultaneously.

There is often significant overlap in the symptoms with these different conditions, so that an older patient with a rotator cuff tear may be hard to differentiate from one with tendinitis. In addition, the literature can be confusing because different authors will give different names to the same entity.

Making a correct diagnosis and formulating an effective treatment plan are dependent on an understanding of the anatomy of the shoulder. The shoulder comprises three joints (the glenohumeral, AC, and sternoclavicular) and four gliding surfaces (the scapulothoracic articulation, the subacromial bursa, the rotator cuff, and the long head of the biceps).

Standard symptom analysis is the best starting point for the patient history, focusing on the location of the pain, its duration, precipitating factors, aggravating and alleviating factors, radiation, and associated symptoms. It is also important to ask about trauma, work conditions, and hobbies. Often, the trigger for shoulder pain is related to an interest in woodworking, for example, or to overhead lifting on the job.

Age is a key factor in evaluating shoulder pain. In individuals who are under 30 years of age, shoulder instability is the most common problem. Between ages 30 and 45, instability continues to be a problem, but impingement and subacromial bursitis, partial rotator cuff tears, and degenerative joint disease become more common. In the older patient, adhesive capsulitis, rotator cuff tears, and degenerative joint disease predominate. There are some typical features that are often seen with shoulder pain that help narrow the differential diagnosis (see table below).

Isolated glenohumeral arthritis is a relatively uncommon cause of shoulder pain. Patients with osteoarthritis tend to have more difficulty in their knees, hips, and hands than in the shoulder. Some exceptions to this rule are patients with a history of prior significant trauma to the shoulder, who may develop glenohumeral arthritis later on, and patients with recurrent shoulder dislocations, who may develop arthritis in the joint.

The most common historical feature in patients with glenohumeral arthritis is a deep pain, present in all directions of motion, that may be referred to the deltoid region. Often, patients are able to compensate with scapular motion. True glenohumeral arthritis is most commonly seen in patients with rheumatoid arthritis and other types of inflammatory synovitis. In that setting, the other features of rheumatoid arthritis will usually be obvious. It would be quite rare for rheumatoid arthritis to present with shoulder pain as the only manifestation.

back to top

MOST COMMON CAUSES

As mentioned earlier, rotator cuff and supraspinatus tendinitis and subacromial bursitis are the most common causes of persistent shoulder pain in adults. The pain occurs mostly on lifting an arm over the head, reaching for a wallet in a back pocket, or combing one's hair. It is especially problematic at night and often disrupts sleep, because it is difficult to put the shoulder in a neutral position without strain during sleep. During the night, patients may sleep with their arm in a position of abduction and rotation, which can be painful. During the day, they are more likely to avoid such positions and so have less discomfort.

Older patients often have rotator cuff tendinitis in the absence of any overt trauma. The mechanism is thought to relate to a combination of repetitive low-grade injury and relative ischemia of the tendons. In the younger patient, there is usually a history of trauma—typically, a fall on an outstretched arm or excessive use of the shoulder during a sports activity such as pitching a baseball. The underlying trigger in most cases is impingement, which results from compression of the rotator cuff tendons between the greater tuberosity of the humeral head and the lateral edge of the acromion when the arm is abducted. Repetitive compression is thought to contribute to inflammation.

Rotator cuff tears may be full or partial. Acute full thickness tears in younger individuals (under age 40) are normally the result of repetitive use or trauma, such as a fall on an outstretched arm, but they are relatively uncommon. Elderly patients may have full tears that are more chronic or subacute in presentation and often the end result of impingement and tendinitis. There may be an initial injury, but the symptom complex is usually chronic lateral shoulder pain, sometimes with a catching sensation, and often with weakness on overhead lifting and nocturnal pain.

Bicipital tendinitis can occur by itself or in combination with rotator cuff tendinitis. The patient usually reports a more anterior location as the site of the pain than is typical with rotator cuff inflammation.

In the older patient, the clinician always needs to consider referred pain as a possible cause of persistent shoulder discomfort. Potential sources for referred shoulder pain include cervical spine disease, diaphragmatic irritation, coronary disease, pericardial disease, and intrathoracic neoplasms and inflammation. With referred pain, the patient will often have painless active and passive range of motion in the shoulder.

Acromioclavicular joint problems such as arthritis and separation are also common causes of shoulder pain. Typically, patients are able to identify the site of the pain with one finger over the AC joint. The joint will more commonly involve a history of direct trauma, such as being tackled while playing football, or it will result from an occupation that involves weight-bearing with the arm, such as painting.
 

OTHER ETIOLOGIES

Other possible causes of shoulder pain include adhesive capsulitis, thoracic outlet syndrome, and suprascapular nerve impingement.

Adhesive capsulitis. Also known as "frozen shoulder," adhesive capsulitis is a progressive painful shoulder condition that tends to involve the entire shoulder, producing stiffness, pain, and marked limitation in range of motion. The pain may radiate to the arm or neck. In the past, frozen shoulder was a common outcome after myocardial infarction or stroke due to prolonged bed rest and lack of physical therapy. Today, it can be the final result of any of the soft tissue conditions affecting the shoulder or the result of acute injuries that are not treated or are only partially treated. In adhesive capsulitis, shoulder X-rays are normal, the shoulder tendons are intact, and a shoulder arthrogram shows an adherent capsule with marked volume loss in the joint.

Thoracic outlet syndrome. In most cases, thoracic outlet syndrome presents with neck, shoulder, or arm pain. These symptoms result from compression of the brachial plexus and vasculature by a cervical rib, fibromuscular bands, or other anomalies. It most commonly results in paresthesias in the hand and vascular symptoms such as color changes and swelling. Typically, the symptoms are worse when the arm is raised above the head.

Thoracic outlet syndrome can sometimes be confusing. Typically, the pain changes with position and may be associated with numbness in a dermatomal distribution.

Suprascapular nerve impingement. Occasionally, irritation of the suprascapular nerve will cause weakness in the first 20 to 30 degrees of shoulder abduction and external rotation, along with pain. There is no sensory deficit. This condition can be caused by direct pressure on the nerve as it goes through the scapular groove, as in bricklayers who carry heavy loads on the shoulder, or by local problems such as ganglion cysts.
 

SYMPTOM PATTERNS

There are a number of common symptom patterns that help to limit the differential diagnosis of shoulder pain. Anterior shoulder pain is most commonly due to AC joint conditions, glenohumeral arthritis, subscapularis tendinitis, and long head of the biceps tendinitis. Posterior shoulder pain is relatively uncommon; when it does occur, it is most likely due to cervical disease or suprascapular nerve entrapment rather than intrinsic shoulder conditions.

Pain in the lateral deltoid area is the most common pattern. This pain pattern is often aggravated by reaching overhead; it is the classic pattern in impingement and rotator cuff tendinitis. Finally, poorly localized pain is more likely to be referred pain from the neck, thorax, or abdomen, or bone pain, fibromyalgia, or neuropathy.

To return to our patient at the beginning of this article, based on his history, the primary suspicion would be for an impingement syndrome causing supraspinatus tendinitis or subacromial bursitis. However, many older patients can sustain rotator cuff tears without major trauma. The possibility of referred pain should also be kept in mind.
 

SYSTEMATIC APPROACH TO THE PHYSICAL EXAM

The best approach to the physical examination is to be sure to systematically examine the patient. Mistakes are often made when the clinician takes shortcuts, such as not taking off the patient's shirt or not examining for cervical problems. A number of physical examination maneuvers have been tested for sensitivity and specificity, so reasonably solid recommendations can be made on the best tests.

Inspection of the patient's back, shoulder, and chest is the key first step for both acute and chronic shoulder pain. The patient's shirt must be completely removed. Inspection should begin with a comparison of both sides, looking especially for atrophy and other asymmetries. Some problems, such as suprascapular nerve impingement, can be difficult to diagnose if the muscular atrophy that accompanies the pain is not identified. Atrophy of other muscle groups may be visible, which may provide a solid clue to the diagnosis. The "Popeye sign" characteristic of a ruptured long head of the biceps tendon is easily visible in patients when they flex their elbows. In AC joint separations, there may be a step-off that results in a prominence of the end of the clavicle.

More acute causes of shoulder pain, including dislocation, fracture, and joint inflammation, are often visible. In anterior shoulder dislocation, the patient tends to hold the affected arm close to his or her body. This has been termed the "dead arm" sign. The deltoid will look full and the acromion will seem more prominent. Conversely, with posterior dislocation, there will be posterior fullness, the coracoid and acromion will be prominent, and the patient will tend to hold the arm adducted and internally rotated with the palm on the abdomen. Normally, there is a clear history of trauma in these cases, unless the patient has a history of recurrent dislocations.

Clavicular fractures will often produce a bump or deformity at the fracture site. Glenohumeral joint effusions tend to fill in the normal indentation between the head of the humerus and the anterior chest wall, but this may be hard to detect with inspection alone.

If acute trauma is suspected, then the direction of the examination changes. The clinician needs to check for neurovascular integrity, associated trauma, and internal injury. A detailed discussion of acute shoulder trauma is beyond the scope of this article, but it is well described in the "Suggested Reading" list.
 

KEYS TO PALPATION

Palpation can often be revealing with shoulder pain. Acute fractures and dislocations are very likely to have localized deformity and tenderness. Patients with AC joint arthritis or injury will often have localized tenderness in that location.

A recently described test, the Paxinos test, can be helpful in identifying AC sources of pain. To conduct this test, the clinician places a thumb over the posterior lateral aspect of the acromion and the index and middle finger over the mid-clavicle. Pressure is then applied anteriorly and superiorly with the thumb and inferiorly with the index and middle finger, stressing the AC joint. The test result is positive if the pressure causes or increases pain in the AC joint.

Alternatively, AC joint pain can be elicited with forced adduction of the arm across the chest wall. Palpation of the biceps groove may elicit pain, as can palpation over the supraspinatus and subacromial bursa. Discretely localized pain can also be a good indication that the patient may benefit from local therapy, such as a steroid injection.
 

RANGE-OF-MOTION TESTING

The next step in the examination of patients with a painful shoulder is to check for active and passive range of motion. When passive range of motion is greater than active range of motion, it is much more likely that the patient has a periarticular condition such as rotator cuff tendinitis or tear than glenohumeral arthritis or frozen shoulder.

There are several possible range-of-motion tests that have been described. Asking the patient to make the football touchdown sign is often a good first range-of-motion test. On the affected side, the patient will often not be able to fully abduct.

Another general screen for active range of motion is the Apley's scratch test. The patient is asked first to put the hand on the affected side behind the back and reach up as high as possible, as if to scratch the back. This assesses internal rotation of the shoulder and especially the function of the subscapularis muscle. If there is a tear or tendinitis, the patient will be unable to raise the thumb to the normal level, which is about T8. The patient is then asked to raise the arm on the affected side behind the head and reach down the back as far as possible. This assesses external rotation and infraspinatus/ teres minor function. Normally, a person will be able to reach to the T4 level. Finally, the patient is asked to reach across the chest with that same arm to the acromion of the other shoulder. This tests adduction and will often trigger pain in patients with AC joint problems.

For a guide to the normal range of shoulder motion in adults, see the box below.

Returning to our patient, the examining clinician found no evidence of muscular atrophy or asymmetry on inspection. The patient was unable to do a full touchdown sign. On his left side, he was only able to get the arm up to about 70 degrees of abduction. With Apley's scratch test, the patient was near-normal on external rotation (reaching up the back), but had diminished range of motion on internal rotation (down the back). When the examiner abducted the arm, range of motion was improved by about 30 degrees. The Paxinos test was negative.

At this point, the strongest suspicion is for a rotator cuff or impingement problem with either tendinitis, subacromial bursitis, or a tear. The next step is to test specifically for these conditions.

back to top

IMPINGEMENT AND BICIPITAL TENDINITIS TESTS

The classic test for impingement was described by Neer and is based on compression of the greater tuberosity of the humerus against the anterior edge of the acromion. Standing behind the patient, the examiner stabilizes the scapula with one hand and flexes the arm forward. The patient may report a "catch" or aggravation of shoulder pain between 60 and 120 degrees.

Hawkins described a modified version of this test that can be done immediately after the Neer test. The examiner flexes the arm to 90 degrees, flexes the elbow to 90 degrees, and then internally rotates the elbow, all the while supporting the arm at the elbow. This has been called the "empty can" or "beer can" test because it simulates emptying a can.

Several maneuvers have been developed for identifying bicipital tendinitis. These are generally based on the concept of stressing the long head of the biceps tendon as it passes through the biceps groove at the head of the humerus. In the Yergason test, the clinician flexes the elbow to 90 degrees and then resists supination of the hand while palpating the biceps groove. Since there is no motion of the shoulder during this test, localized pain in the biceps groove is very suggestive of inflammation of the biceps tendon.

Alternatively, Speed's test elicits biceps tendon pain with shoulder flexion with an outstretched arm against resistance.
 

TESTS FOR ROTATOR CUFF TEARS

The rotator cuff muscles and tendons can be remembered by the mnemonic "SITS," beginning with the supraspinatus superiorly, then the infraspinatus, teres minor, and subscapularis posteriorly. The most common site of a rotator cuff tear will be in the supraspinatus muscle and tendon.

Several tests have been shown to be useful in diagnosing rotator cuff tears: the supraspinatus weakness test, the drop arm test, and the weakness of external rotation test. The supraspinatus is responsible for shoulder abduction, especially in the 30- to 90-degree range. To effectively isolate the supraspinatus, the patient should be asked to abduct the arm to 45 degrees in about 30 degrees of forward flexion with the thumb pointed down. He or she should then be asked to resist downward pressure.

The drop arm test is a variant in which the arm is passively abducted more than 90 degrees and the patient is asked to lower the arm with the clinician's support. With a complete rotator cuff tear, the patient will not be able to maintain 90 degrees of active abduction and thus the arm falls or drops. This is the most specific sign for a complete rotator cuff tear, but the sensitivity is less than 20%.

External rotation is mainly accomplished by the infraspinatus and teres minor muscles. With the elbow flexed at 90 degrees, the patient should be asked to externally rotate the arm against resistance. In the same position, the examiner can test internal rotation strength by having the patient push the hand toward the abdomen against resistance.

Less commonly, the subscapularis muscle is affected by trauma or a tear. Weakness of this muscle can be tested by the lift-off test. With the hand behind the back, as in Apley's scratch test, the patient is asked to lift the hand off the back against resistance.
 

TESTS FOR THORACIC OUTLET SYNDROME

Several tests for thoracic outlet syndrome have been described, but care must be taken in interpreting the results since normal patients may test positive in some cases. For all of these tests, provocation of symptoms or loss of the radial pulse, or both, indicates a positive result.

The classic maneuver is Adson's test, in which the patient turns his or her head to the affected side, lifts the chin up, then takes a deep breath and holds it for 30 seconds while the examiner palpates the radial artery.

The elevated arm stress test is the most sensitive test. With the arms elevated to at least 90 degrees and hyperabducted, the patient opens and closes his or her fists for three minutes. The examiner then palpates the radial pulse. Pulling the shoulders down and posteriorly for 30 seconds and applying pressure with the thumbs at the anterior scalene muscle near the first rib can also be useful tests.

How did our patient do with his shoulder pain tests? He had pain at about 70 degrees of abduction and had a positive Hawkins test, but he did not have a positive drop arm test and his external rotation strength was intact. He was, however, somewhat weak on testing of supraspinatus strength, but this maneuver was painful.

At this point, the clinician should be most suspicious for rotator cuff tendinitis or subacromial bursitis in this patient. However, there should still be some concern for a rotator cuff tear due to the supraspinatus weakness.

If a localized problem such as rotator cuff inflammation or bicipital tendinitis is suspected, the lidocaine test can help determine if steroid injection will be helpful and can also help confirm the diagnosis. For this procedure, 3 to 5 ml of lidocaine are injected into the subacromial bursa. If this clearly relieves the patient's pain and range of motion improves, then the test is positive. This would rule out significant glenohumeral arthritis. Patients with a complete rotator cuff tear will not have improved range of motion after lidocaine injection, and patients with frozen shoulder will still have limited range of motion. The same test can be done in the bicipital groove for bicipital tendinitis.

Our patient did receive a lidocaine injection into the subacromial bursa. His range of motion improved, and his supraspinatus strength was intact on repeat testing. Thus, it can be concluded that this patient's primary problem is rotator cuff tendinitis or subacromial bursitis and that he is unlikely to have a full rotator cuff tear.
 

DIAGNOSTIC STRATEGY FOR ROTATOR CUFF TEARS

Two recent studies—one in the United States, the other in Australia—have examined the utility of bedside tests for rotator cuff tears in adults. Both examined a large number of patients who had shoulder pain and had definitive testing with arthrography that identified tears. The studies both found that many of the tests described for rotator cuff tears have low sensitivity and specificity, with resultant false negatives and positives.

Weakness on external rotation was a key finding in both studies. The U.S. study found that if a patient was over age 65, had nocturnal shoulder pain, and had weakness on external rotation, there was a 91% chance of a significant rotator cuff tear. The sensitivity of this strategy, however, was relatively low since those without any of these findings still had a 30% to 40% chance of a tear.

The Australian study combined weakness on external rotation with supraspinatus weakness and a positive impingement sign in its prediction rule. If all three were positive, then the post-test probability for a tear was 98%. With such patients, physical therapy should certainly be considered, and, in the younger patient for whom early surgery would be indicated, possibly an orthopedics referral. If all three were negative, then the chance of a rotator cuff tear was only 0.5%, which effectively rules out a significant tear. Patients with one or two positive findings had intermediate probabilities; most of these patients can be treated initially with conservative therapy, with referral reserved as an option for those who fail to respond.
 

WHEN TO ORDER IMAGING STUDIES

Most patients presenting with shoulder pain in primary care settings do not need an imaging study immediately. The main indications for plain films are a history of recent trauma, evidence of AC separation or glenohumeral arthritis on physical examination, prior surgery on the shoulder, or a rotator cuff tear on examination.

A Boston study examined predictors of therapeutically important X-ray findings in patients presenting to an emergency department with shoulder pain. If a patient has visible swelling, then an X-ray should be ordered because more than half of such patients in the study had a therapeutically important X-ray finding. If there is no history of a fall, then the clinician can feel comfortable not ordering a film since the rate of important findings was only 1%. If the patient did fall and also reports pain at rest, then the next step is to evaluate range of motion. If it is normal, then an X-ray has low yield, the Boston study found. This algorithm needs further validation in other study populations, but it provides a reasonable approach to determining who would benefit from plain film imaging.

Other tests such as arthrograms and magnetic resonance imaging (MRI) are most helpful in acute trauma situations; they do not play a significant role in the initial management of patients with shoulder pain. An MRI can be very helpful for patients who do not respond to conservative therapy or those who have clear indications for early surgical referral, such as patients with evidence of acute rotator cuff tears. Also, it is important to remember that many asymptomatic older patients may have evidence of chronic rotator cuff tears, either full or partial, so routine MRI scanning is not a cost-effective approach with them.

To return to the case presentation, the patient did not have obvious swelling and there was no history of a fall. His pain was mostly on active motion of the shoulder and he had good passive range of motion. Thus, a plain film was not done at the initial evaluation.
 

MOST EFFECTIVE THERAPIES

The most commonly encountered situation in primary care settings with regard to shoulder pain is an older patient with a several-month history of such pain. In these cases, the first step is to initiate conservative therapy for a period of four to six weeks. Many patients will improve during this time period and will not require referral or more intensive interventions. For patients with periarticular shoulder pain, who comprise the vast majority of patients, analgesics and anti-inflammatory agents and range-of-motion and strengthening exercises are the mainstay of treatment. A recent systematic review found that, in general, exercise regimens are effective for shoulder pain, although the methodological quality of these studies is low.

The initial phase of therapy for rotator cuff tendinitis begins with avoidance of activities that involve reaching or lifting. Patients should apply ice to the lateral shoulder region and begin range-of-motion exercises. One of the best initial exercises is pendulum swings (Codman's exercise) with a light weight (less than 10 lbs). Patients can use a filled plastic gallon milk jug, which weighs about 8 lbs, for this exercise. The patient bends slightly forward and holds the weight close to the body. The swings are in a small arc, less than a foot in diameter, and should be performed once or twice a day for five minutes. The goal is to stretch the shoulder and reduce impingement. Of course, this exercise is not recommended in patients with shoulder instability or AC joint separations. Nonsteroidal anti-inflammatory drugs can be a useful adjunct, but they were not as effective as exercises in one of the trials reviewed by the Cochrane Collaboration.

After two to three weeks of such therapy, when the pain has subsided, the patient can begin strengthening exercises and additional range-of-motion exercises. At this point in therapy, many patients find that moist heat is effective in reducing pain and in improving range of motion. The key exercises are "wall walking," in which the patient stands facing a wall and gradually "walks" the fingers on the affected side up the wall as far as possible. Patients can also use a cane or broomstick to passively abduct the affected arm. As the patient's condition improves, the same maneuvers that can be used to diagnose rotator cuff tendinitis, such as the Apley's scratch test or full abduction, can be done to continue to improve the flexibility of the shoulder.

To strengthen the rotator cuff, the therapeutic focus should be on improving internal and external rotation. A simple exercise for this is to have the patient position himself or herself in a doorway and do isometric internal and external rotation against the door jamb. Resistance bands that patients can use for the same purpose are also available. Stretching the posterior capsule can be accomplished by adducting the shoulder with the elbow flexed and using the unaffected arm to stretch the shoulder further.

The patient in our case presentation started icing his shoulder and doing pendulum exercises. There was initial improvement, but he still had pain on lifting his arm and also at night.

Many patients will improve with four to six weeks of ice application, range-of-motion exercises, and mild strengthening. However, those who do not improve may be good candidates for steroid injection; a systematic review found several trials that showed improved results compared with physical therapy. Steroid injections are especially useful for the patient with a positive lidocaine test and clearly localized pain. The two most common areas for injection are the subacromial bursa and the biceps groove. In the latter area, the clinician must be cautious not to inject directly into the tendon because this may weaken it and lead to rupture.

The technique for injection is straightforward. After a sterile prep, a mixture of lidocaine 1% and 20 to 40 mg of methylprednisolone (about 3 ml in total volume) is injected using a 22-gauge, 1-1/2-inch needle. Similar doses of fluorinated steroids can be used but they may produce a higher incidence of local skin atrophy. Some obese patients may require a longer needle. The patient should be instructed to ice the shoulder that day, restrict use for three days, and avoid activities that stress the shoulder (such as overhead lifting) for several weeks.

Our patient was re-examined and found again to have impingement signs, but his supraspinatus strength had improved. His subacromial bursa was injected with lidocaine and methylprednisolone. He rested over the next few days and then resumed his exercise program. When he was seen two months later, he was much improved. He still had occasional twinges of pain with certain movements, but his nocturnal pain had resolved.
 

INDICATIONS FOR REFERRAL

An orthopedic referral should be considered for patients who fail to respond to conservative therapy within a few months and those who have evidence of an acute rotator cuff tear on examination. Patients with more acute problems such as AC joint separation, trauma, or biceps tendon rupture should also be referred after the initial evaluation. That being said, the majority of patients with chronic shoulder pain in primary care settings can be cared for successfully without referral. Physical therapy referral is important for patients who need to resume a sport or occupation, for those who fail initial therapy, and for those who have difficulty carrying out a self-directed program of exercise. Physical therapists can also help with mobilization of the shoulder.

One year after his evaluation, our patient had been pruning trees when he felt his shoulder pop. He had recurrence of the shoulder pain on lifting the arm, as well as recurrence of nocturnal pain, and on examination he had both weakness of supraspinatus and external rotation and a positive drop arm test. This did not improve with lidocaine injection. The diagnosis was thought to be an acute rotator cuff tear.

He had an MRI that showed a complete rupture of the supraspinatus tendon. Since this was an acute tear and the patient was in good health otherwise, he underwent surgical repair. When last seen, he had successfully rehabilitated his shoulder and had good strength and range of motion.
 

COMMON PROBLEM

Shoulder pain is a common problem in primary care that results in important functional limitations for patients. A systematic approach to the history and physical examination can lead to an accurate diagnosis and effective therapy. Most patients do not require either imaging or referral initially and can be managed successfully in the primary care setting.

back to top

Printable version of this article

E-mail this article