“I think I need an MRI for my shoulder”

A female in her mid-twenties recently visited my office, presenting with right shoulder pain.  She works as a painter and fabricator, a pretty physical job, and she is right hand dominant.  She began having shoulder pain 2 months prior after working a double shit. Since then, she has tried stretching and relative rest, naproxen, and has had no improvement in her pain, except notable improvement every time she gets 2 or more days off work in a row.  After working, she states that she experiences extreme pain in her shoulder, as if “it is going to fall off.” Due to the pain, she has not been getting an adequate amount of sleep. 

The patients primary care physician thought she might have a rotator cuff tear and referred her for a second opinion. She stated that she wanted MRI to “see what is going on,” which is not a rare request.

On physical examination, the patient had:

  • a positive Watsons
  • a negative drop-arm with pain
  • negative dynamic labral shear test
  • irritation of her rotator cuff
  • no significant alteration in her scapulothoracic rhythm

After her exam, my initial impression somewhat aligned with her desire for further diagnostic imaging. Instead of going through the hassle of scheduling an MRI, we were able to achieve clarity through imaging in my office, that day, with ultrasound imaging. Faced with the ability to avoid an MRI, the patient agreed to undergo a diagnostic ultrasound of her shoulder.

The images were as follows:

Figure 1: Axial view of biceps -normal

Figure 2: Long axis view of long head of biceps, normal

Figure 3: Coronal view of subscapularis, normal

Figure 4: Supraspinatus coronal view.  This is normal. The hypoechoic signal on the right is artifact.  Note how collagen fibrils can be seen as intact and linear as they course over the humeral head.  Multiple views are obtained to see the whole tendon.

Figure 5: Infraspinatus, normal.  The hypoechogenicity on the left is the red-white junction.

Figure 6: Teres minor coming off humerus to left, normal.

Figure 7: Normal glenohumeral joint and posterior labrum


This is an irritating but benign overuse injury.  There is no macroscopic, anatomic damage to the rotator cuff.  Importantly, she does not need surgery. Since she does not need surgery, she does not need MRI since MRI is a pre-operative study and she did not have labral findings.

Plan of Care

In this case, I advised the patient that the most effective plan would be to rest the shoulder and do physical therapy. Since she will still have to work, she was advised to do more with her left hand to avoid over-using her right shoulder. 

I did not recommend for her to have corticosteroid injections, as they can weaken the rotator cuff and slow certain biochemical healing pathways.  I also advised avoiding NSAIDs (non-steroidal anti-inflammatories such as naproxen and ibuprofen).

If several weeks pass and she has not seen any improvement, I would suspect that she is developing tendinopathy. In order to diagnose, we would use the ultrasound. If that is in fact the case, I would recommend the patient considers PRP (platelet-rich plasma, a way to introduce high concentrations of 7 key growth factors to the injured structure).


Diagnostic ultrasound can provide immediate answers and save the patient a lot of money and time.  Physical Therapy and relative rest and avoiding steroids and NSAIDs are the treatments of choice. If this plan of care did not work and she developed worsening tendinopathy down the line, the Physical Therapist could consider sending her to get consider platelet-rich plasma (PRP) injection for a boost to the tendon healing, which is offered at Integrative Sports Medicine.