To Have Or Not To Have Surgery: A Multi-Part Series

Part 2: The Shoulder


As a continuation of last month’s newsletter on knee surgery, in this installment of The Incision Decision we will discuss the shoulder. The basic principles of decision-making are essentially the same for the shoulder as they are for the knee; radiographs, in the absence of sinister pathology and in isolation, should not be the determining factor in deciding whether or not to have shoulder surgery.

I see many patients in my office with radiologic findings of rotator cuff tear, yet they remain functional. In a study that I have referenced on many occasions and in the video above (Extremity Pain of Spinal Source) by Rosedale et al (2019), the authors found that greater than 40% of extremity pain (such as in the shoulder) has its origin in the spine. Unless the spine is first ruled out as a source of the symptoms, we will miss the true cause of the symptoms almost 50% of the time (a flip of a coin) and, by extension, treat the wrong thing!

The story of a patient I saw for his shoulder more than twenty years ago will illustrate this point. The patient was previously seen for low back pain, had a good result, and wanted to return for a trial of physical therapy before undergoing a recommended rotator cuff repair. The patient was told that he had a complete tear on his MRI and that physical therapy would not help, that he required surgery if he ever hoped to regain any function in his arm. Despite the recommendation, the patient chose a trial of conservative treatment (physical therapy) prior to having surgery. We first ruled out the neck (that is, the spine) as a source of the patient’s symptoms and he did, indeed, have a shoulder injury.

The only problem with the “MRI diagnosis” of complete tear was that the patient could lift his arm over his head. This would not be physically possible with a complete tear. It would be comparable to trying to pull a sled with a rope that had been cut. The patient was seen for six visits and discharged with a fully functional arm. No surgery was required.

Guidelines for when to consider shoulder surgery (in consultation with your surgeon):

  • The shoulder joint is unstable and/or so painful that there is an inability to perform normal activities of daily living, and it has not responded to a course of conservative treatment (physical therapy).
  • There is a persistent hemarthrosis (the joint is hot and red for an extended period of time and not responding to conservative care). This may indicate a fracture or infection, a tear in the muscles that stabilize and move the shoulder (rotator cuff), the lining of the joint capsule (synovial membrane) or the cartilage lining the joint (labrum). It may also indicate that other structures within the shoulder have been damaged.

Outside of an emergency situation (e.g. fracture/trauma or serious pathology) an invasive procedure should not be the first choice. Remember, the axiom; ‘Do as little as possible, but as much as necessary’ applies equally to the shoulder as it does to the knee when deciding whether or not to have a surgical procedure.

It is important to note that, in the case of the knee, instability can become a safety issue that may leads to falls. The difference with the shoulder is that one might not fall as a result of a shoulder injury, but might not be able to prevent falls due to loss of shoulder function (an indirect result as opposed to a direct result seen in the knee).

A brief history on the evolution of shoulder arthroscopy (Dessai, 2020):

The invention of the incandescent light bulb by Edison in 1879, led to the introduction of the laparo-thoracoscope in 1910. Attempts were made to use this device in the knee joint as well. Development of the arthroscope really took off after the introduction of “cold-light” and rod lens optical system by Hopkins in 1960. Kenji Takagi and later Masaki Watanabe get the credit for developing the modern form of arthroscopy.

The spillover of knee arthroscopy into the shoulder was inevitable and began in the 1980s. Shoulder arthroscopy started with instability repair, followed by subacromial decompression (relieving the pressure between the bone on top of the shoulder joint and the top part of the upper arm bone). Through the 1980s and 1990s, with developments in biotechnology, more sophisticated tools and anchors became available, leading to refinement of instability repair procedures.

Remember, arthroscopic surgery is not a panacea and certainly no substitute for a good clinical evaluation.

Total Shoulder Replacement surgery:

The first Total Shoulder Replacement surgery was performed in 1893 by the French surgeon Jules Emil Pean, as treatment for a 37-year-old baker dying of a tuberculosis infection in his (R) shoulder. Pean used a platinum and rubber implant in the patient that lasted 2 years and allowed the young man to return to his work.

The next generation of prostheses came in the 1950s, with the advent of the plastic and acrylic humeral head (the ball part of the joint). The implants were typically fixed with plates and screws farther down the humerus (upper arm bone) and failed due to the screws being pulled out of the bone. Several iterations of implant designs and materials appeared through the early 1980s, again, with limited success. By the 1990s, new designs resulted in marked functional improvements and consistently more positive outcomes. Currently, up to 90% of shoulder replacements last for 10 years or more.

As with the knee, what I frequently hear from patients is; ‘My X-ray is bone on bone,’ or, ‘I was told that ‘this is the worst I’ve ever seen,’ or ‘according to your MRI, surgery is the only way you will have the use of your arm.’

Again, the image should not make the diagnosis. It is just one piece of a multi-factorial presentation, including, but not limited to current function and mobility, the patient’s lifestyle and needs, risk/benefit analysis, recovery time, patient age and health status, each to be carefully considered. The surgery-decision-making criteria remain the same for the shoulder as in the knee: you are no longer able to perform your daily activities and/or your shoulder is so painful that you are unable to function.

Understanding Frozen Shoulder:

Any discussion of shoulder dysfunction and its treatment must include frozen shoulder, a musculoskeletal condition unique to the shoulder. According to Griggs et al (2000), the largest single group of patients have no detectable underlying cause for their symptoms (primary idiopathic frozen shoulder). A substantial group of patients with diabetes mellitus (Type II Diabetes) are considered as a separate group, Diabetic Frozen Shoulder, since the course is usually more severe (with regard to pain and limitation and tend to be more protracted (lasting much longer), presumably due to compromised circulation common in diabetes.

Patients with frozen shoulder will report that the symptoms come on spontaneously, without any apparent cause. Radiologic findings are frequently negative and the symptoms are rarely associated with pathologic findings. The diagnosis is made with the history and clinical presentation:

The three phases of frozen shoulder:

  • Initial phase: characterized by significant should pain with movement and/or at rest
  • Freezing phase: characterized by a decrease in pain with progressive limitation in shoulder elevation and rotational motions
  • Thawing phase: characterized by progressive increase in range of motion and function

Surgery is rarely an option with frozen shoulder. Occasionally, manipulation under anesthesia is performed, where the patient is put to sleep and the shoulder is forcibly moved. This procedure typically results in a worsening of the symptoms and a more protracted course. Physical therapy is typically the treatment of choice. With no intervention, the typical course is 18 months to resolution. With appropriate physical therapy, the length of the clinical course can be significantly shortened (6-9 months).

I will recommend a book by T.R. Reid, entitled ‘The Healing of America,’ that chronicles the author’s search for a solution to his very painful shoulder that was fractured decades earlier, repaired successfully, but started to give him severe pain with functional limitation forty years after the initial injury. The search takes Mr. Reid to the major industrialized countries as he explores how each of those country’s health care systems would manage the injury.

As with the total knee replacement, the stronger, more fit and flexible you are going into the surgery, the faster and more complete the recovery. If you have any questions about conservative treatment for shoulder pain or dysfunction, do not hesitate to contact me.

For the right reason, at the right time, in the hands of a skilled surgeon, surgical outcomes are typically excellent, with a result that could not be achieved in any other way.

Yours in health,



Desai SS. History and evolution of shoulder arthroscopy. J Arthrosc Surg Sport Med 2020;1(1):11-5.

Flatow EL, Harrison AK. A history of reverse total shoulder arthroplasty. Clin Orthop Relat Res. 2011 Sep;469(9):2432-9

Griggs SM , AhnA, GreenA. Idiopathic adhesive capsulitis: a prospective functional outcome study of nonoperative treatment. J Bone Joint Surg [Am]2000;82-A:1398–1407. 

Robinson, CM, Seah KTM, Chee, YHP, Hindle, Murray, IR. Frozen Shoulder. J Bone Joint Surg [Am]2012;94-B:1-9.

Rosedale R, Rastogi R, Kidd J, Lynch G, Supp G, Robbins SM. A study exploring the prevalence of Extremity Pain of Spinal Source (EXPOSS). J Man Manip Ther. 2020 Sep;28(4):222-230.