This term applies to an uncommon condition wherein the ligamentous capsule surrounding the shoulder tightens up. This restricts motion. This develops as a person fails to move her arm through a full arc of motion allowing the capsular ligaments to tighten very slowly. Progressively, more pain will develop at extremes of motion resulting in a stiff “frozen” shoulder. Neck pain resulting from muscle strain or arthritis and heart pain coming from restricted coronary blood flow may radiate into one’s arm or shoulder and subacromial bursitis may also cause discomfort that initially caused the person not to be moving her arm normally. In fact, one of the most common scenarios in the past was an old person with a broken wrist. She would limit her arm motion while protecting the injury and end up with a stiff, frozen shoulder sometimes associated with stiff fingers. People with diabetes are overwhelmingly the most likely to develop a frozen shoulder, in their case often times without any obvious precipitating event. The incidence of frozen shoulder in the general population is 2-5% but the people with diabetes have an incidence of 10-20%. The physiology leading to this kind of frozen shoulder is probably different from than that in a non-diabetic.
Treatment is based on trying to give a person pain relief and preventing them from losing further motion than they already have by the time that they come in to see the doctor. The natural course of frozen shoulder occurs over about a two-year span after which it essentially burns out leaving a person with some restricted motion. When we see a person with frozen shoulder we try and provide them with medication to minimize pain. There is no absolute proof that it speeds recovery. Medication spans from corticosteroid injections which have some temporary relief in some patients but not in others, non-narcotic analgesics such as Tylenol, tramadol and also probably some use for gabapentin, which is widely used in the setting of pain management and an extremely safe medication. Assisted range of motion with the help of a therapist after applying heat can be beneficial to many people and for those who aren’t showing improvement of their problem both with diminishing pain and possibly improving motion, then if they have lost a lot of motion stretching the shoulder out with the shoulder completely numb, i.e. an interscalene block, or general anesthesia, or both, are indicated. The role of arthroscopy in frozen shoulder is debatable. There are those that think it is less traumatic to a person to reach inside the shoulder and cut the capsule, though this is much more expensive and time-consuming. I am not certain the role has been proven in general. I will reserve it for someone with a severe case who can’t maintain the motion gained after manipulation. This is a very small subset.
An untreated frozen shoulder takes about two years or more to go from start to finish for a person to run the course as far as the pain is concerned. Frequently the shoulder will end up with some restricted motion but pain-free if untreated; asking each person to live in pain, however, for two years without any intervention while she can’t move the arm is generally unacceptable. The person is just hurting too much and wants to take advantage of some sort of help.
The person who gets frozen shoulder on one side has a 34% chance for developing it on the other side as well if it develops spontaneously. Over the years I have seen a number of patients who have come in with one shoulder only to be followed in a year or more with the opposite shoulder.
A frozen shoulder is diagnosed on the basis of symptoms including decreased motion, particularly reduced ability to reach behind your back. An arthrogram or an MRI scan may be required to be certain that the rotator cuff is not also torn. This might demand more early aggressive surgical treatment. Treatment initially should consist of a non-addicting analgesic such as acetaminophen and perhaps a corticosteroid injection to see if the person does get significant relief from same since it is such a non-harmful treatment. If a significant benefit of more than a day or two is not experienced with a cortisone injection, it should probably not be repeated. If one gets lasting benefit, a repeat injection might be warranted a few weeks later.
A trial of NSAID (anti-inflammatory pills) may be of benefit and is certainly low risk. If the patient is not showing dramatic progress then manipulation under anesthesia in the operating room should be considered leaving open the option of possible arthroscopic capsular release. Home exercises should be continued indefinitely.
The most widely used treatment is physical therapy consisting of hot compresses and assisted range of motion wherein the therapist helps the person pull her arm through an arc of motion. During this time the patient is also taking analgesics such as acetaminophen (Tylenol) and hydrocodone in severe cases. There may be some benefit for nonsteroidal anti-inflammatory medication such as meloxicam, ibuprofen, Celebrex, etc. Corticosteroid injection will sometimes give some temporary relief but hasn’t been proven in general to be very effective. The resisting cases are best treated by gentle stretching of the joint while fully relaxing the muscles under general anesthetic in the operating room. This is known as “manipulation” of the frozen shoulder. This is followed by an aggressive daily physical therapy program as an outpatient. Normally the person has been experiencing a frozen shoulder for a few months before they had gotten to the point they would be stiff enough unattended to warrant this type of intervention.
For late presenting frozen shoulder with severe stiffness and loss of motion and/or pain, an occasional patient may be best treated with arthroscopic shoulder capsular release followed by aggressive physical therapy. In this situation, rather than gently stretching the shoulder without an actual operation, the shoulder is manipulated and then a fiberoptic scope is inserted for visualization. A series of small surgical tools are inserted into the joint to release the tightened capsule under direct vision. Because of the very high cost and extra risk such as nerve injury that accompany arthroscopic capsular release, this is not generally considered a first line of treatment by those who are practical. In fact, even when offered manipulation most patients become further motivated to benefit from physical therapy and a home exercise program.
The Risks of Treatment
The risk of gentle manipulation of the shoulder is almost zero. There is a theoretical risk of breaking the humerus, i.e. a fracture, particularly in the case of someone with osteoporotic (weak) bone. There is a theoretic risk to the nerves around the shoulder or brachial plexus going into the arm from the neck. I have never actually heard of a case of this occurring.
Arthroscopic capsular release on the other hand has the remote risk of infection and a significant risk of axillary nerve injury because it lies so close to the underside of the glenohumeral capsule.