Ahh, the mystical frozen shoulder.  We have all had our share of patients, with frustration typically steaming out of both parties’ ears.  It is a diagnosis usually made of exclusion, but a proper examination can help rule-in the condition.  The onset is usually spontaneous, the cause of the condition is still unknown and not one conservative intervention has been exclusively shown to be superior.

As physical therapists, we strive for all patients to be treated conservatively.  I do this for every condition and I will say 90%+ of patients should NOT need any other type of medical intervention, including medication & especially surgery.  One that continues to stump me though is the challenging treatment of adhesive capsulosis (frozen shoulder).  One of the main reasons is that if the shoulder is truly ‘frozen’, the average length of symptoms could last up ~30 months (average 1 year to 3 years).  Three stages have been described in the literature, and typically are presented by the following graph:

Frozen Shoulder Time Frame

I like treating orthopaedic cases that have a significant within-visit response and can get instantaneous gratification, even though full resolution will take some time. Besides a few interventions in lower realm of evidence (thoracic manipulation case report and dry needling case report), most of the time the treatment is slow. 

This treatment process can be labeled a “difficult condition and even difficult patients”.  With the deadlock of a condition as frozen shoulder can be, clinicians and patients could consider pulling the trigger SOONER than later to proceed with more risky interventions.  One of the most common interventions is manipulation under anesthesia (MUA).  With that said, the question remains: should MUA be considered earlier in the care for frozen shoulder patients?  This could be a possible proposal to improve ROM, decrease pain and decrease disability in a more speedier fashion, THEN proceed with physical therapy care.

Frozen Shoulder Risk Analysis

Let’s delve into this question further.

Anytime a procedure is performed, a risk:benefit analysis must be weighed.  For the most part, any physical therapy procedure is extremely safe with very little adverse events.  A low risk and over 90% success rate, the first choice for treating frozen shoulder should be physical therapy.  However, the success can depend on what you perceive as a short period of time.  Meaning, do you want to push through ~4 months of rehabilitation or will a MUA get you results sooner and get out of the frozen stage quicker?

For the sake of this post, I’ll skip steroid injections (as this is usually combined with ~12 weeks of therapy) and proceed to manipulation under anesthesia.

The procedure of MUA itself definitely has more risk, especially of possible fracture, rotator cuff tears, and labral tears.  Also, some surgeons will avoid the procedure if the patient is a diabetic (insulin-dependent), which is a risk factor for the condition. This does not include the risks associated with anesthesia, which can be very high, including recently noted to double dementia risk.   The benefit must then be weighed.  I have always heard ~70% success rate with MUA. Let’s look at some more evidence:

In a study of 125 patients by Kivimaki, et al 2007 comparing MUA to HEP had this conclusion,

Manipulation under anesthesia does not add effectiveness to an exercise program carried out by the patient after instruction.

On the other hand, Dodenhoff et al 2000 expressed,

We recommend the use of manipulation under anesthetic in primary frozen shoulder to restore early range of movement and to improve early function in this often protracted and frustrating condition.

Vastamaki & Vastamki 2013 found in a low level study but followed patients a mean of 23 yrs showed,

In this group of patients treatment of idiopathic frozen shoulder by MUA led to improvement in shoulder motion and function at a mean 23 years after the procedure

Farrell et al 2005 looked at long-term (15 yrs) with the following results (however, it was after failure of conservative care),

Treatment of idiopathic frozen shoulder by manipulation under anesthesia leads to sustained improvement in shoulder motion and function at a mean of 15 years after the procedure.

Wang 2007 had this conclusion,

We concluded that manipulation under anesthesia when initial conservative management failed speeds up the recovery of idiopathic, post-trauma and post-surgery frozen shoulders and improves shoulder function and symptoms within a short period of time. 

I would imagine physical therapy would be the most cost-effective intervention, but recent systematic review and cost-effectiveness analysis could not provide a conclusive outcome (no surprise since this is a systematic review!).

So in conclusion, the evidence doesn’t really provide us guidelines of using MUA besides suggesting it after failure of conservative measures.   I could not find any studies that solely looked at performing MUA EARLIER (to then decrease the length of disability).  Would the disability be less and recovery speedier?

My opinion:

I don’t have a true answer here and it all depends (don’t you hate that answer!?).  I will say that I have seen many patients who respond solely to physical therapy services and not need anything more invasive, costly or risky as a MUA.  However, there are ~5% of patients who just do not respond to PT.  These have gone on to have MUA (or lysis) and have come back to PT with a significant improvement in ROM and then get to ~90% of prior function within a month. I am trying clinically to find out if they will respond to my care or will have to have another treatment earlier in the course of care.  Pretty much all of my patients do not want to have another procedure and will rather continue with PT services, even if the course of treatment is slow.  That makes a big difference too in regards to how to approach various treatment options.

What are your clinical judgments in regards to this?  If you had a patient right now with frozen shoulder who has not shown improvement in ROM in the last ~4 weeks, but pain is still mild to moderate (NRPS: 2-4/10), would you continue with physical therapy or recommend MUA?  Or, would you suggest continued care with hopes of  transitioning to the ‘thawing’ stage quicker with further PT services?

At what time-frame do you consider conservative therapy to be a failure?  Do you think physical therapists over-treat this condition and should recommend other measures sooner than later?


    1. Hey Bruce,
      Thanks for responding. Do you ever run across someone where you would recommend MUA?

    2. I am a type 1 diabetic and have had both of my shoulders frozen for 2.4 years now. Pt and injections have done very little for me. I am have manipulation done on the 18th. So far I have wasted over 2 years trying and suffering.

  1. Very good article on this subject. thank you for this. What I’ve found is that by having a method where I can palpate and feel for path-anatomic and path-physiological landmarks or areas I can discern which tissues are problematic. With Integrative Manual Therapy a method of physical therapy developed by Sharon Giammatteo, PT PhD, I’ve learned to not only test for ligaments, joints, bone and myofascial components but also to look at arteries such as the aorta, subclavian, carotids arteries and veins, SVC, IVC, Jugulars, brachiocehphlic as well as look for bone bruises in the rib cage and problems in the organs such as the heart and lungs. In these areas it is common to find infection, inflammation and disruptions of membrane walls. There are methods to check among all the findings and find the actual primary areas that are causing the problem including why a body would freeze up a shoulder. By using tech from IMT physical therapy we can help the body heal the bone, or artery or vein or clean up infections or fix up disruptions in membrane wall integrity. When this happens if we get the right structures which we have seen in our clinic on numerous occasions, we see dramatic increases in ROM and decreases in pain. These improvements continue to improve between sessions since the primary cause was found. When looking at good manual therapy done to a shoulder that is not giving immediate or within session results, it may be due to treating the wrong primary problem or area. Looking at the arteries, veins, bones and looking for infection, inflammation and disruptions in membrane wall integrity can be a missing link. It’s what we’ve been using and finding good results with. you can read more about it from our site as we have links to information about IMT physical therapy at http://www.ralphhavensphysicaltherapy.com feel free to email me with any questions or if you’d like any more information on Sharon Giammatteo and Integrative Manual Therapy. Ralph Havens, PT, OCS, IMTC, certified matrix energetics practitioner, at http://www.ralphhavensphysicaltherapy.com
    Ralph Havens Physical Therapy, Bellingham, WA

    1. Hey Ralph,
      Thanks for sharing your knowledge on IMT. I have not taken any of those courses but could be an avenue in the future.


  2. I have had MUA on both shoulders after 3-4 months of therapy getting nowhere. The lates one was this past week. It shows significant inprovement for me. Yea, therapy is still uncomfortable, especially during the first week of going daily, but it has definitely been worth it for me.

    1. Thanks for commenting Kim and glad you have seen an improvement already. It shows that if PT initially does not work, then you should have a better response to MUA. Good luck with the rehab and stick with it! Slow and steady as they say!


  3. I recently had a patient who was a prime example of this. I have had frozen shoulder in the past respond extremely well to PT. With this latest patient, after expiration of the first script she was still lacking a majority of her ROM. Following failed conservative treatment she underwent bilateral MUA. Her motion is markedly improved and her pain has been dramatically reduced. I would say MUA is definitely a viable option for those who have failed conservative treatment.

  4. Hi Harrison,

    Thank you for this informative post. A nice presentation of the topic and use of time frame and risk analysis graph has made the correlation more clear.

    Many of my patients with adhesive capsulitis have responded well to conservative treatment but they were on stage 1-2. I wanted to share a case. I am treating a patient for frozen shoulder with type 2 diabetes. Steroids injection is not an option for me because of her co-morbidity.This patient showed no improvement from last 3 months but now there is a sign of it. What do you suggest at this point as MUA would be my last resort !

    I look forward to reading your new post!


    1. Anu,
      Sounds like your patient is heading in the right direction. This type of condition is not quick responding, but slow and steady. I think we can “speed” up the healing IMO and while improving function and decreasing pain.

      If showing improvement, I would not go to MUA unless your patient would like quicker results. She/he has to weight the risk/benefit ratio.


  5. MUA helped me significantly with my frozen shoulder. I’m now getting that last bit of ROM via PT albeit it isn’t without significant pain from the stretching. Thanks for the good article.

  6. Hi, I am an insulin dependant diabetic for 30 years, Juvenile Diabetes. I am 55 y/o female, I slipped on a oily deck 10 months ago, broken shoulder, torn ligaments and now frozen shoulder, I have been having PT for this period, condition is not improving in the past 3 months. Pain is driving me crazy at night and Its been 10 months since having a decent sleep. Am I too old for MUA, is it not recommended in IDDM, I have not had any cortisone injections and am trying to rely only on paracetamol and it is no help what so ever. What is the risk to diabetics in having MUA
    Thanking you in advance

    1. Cherie,
      Sorry to hear that PT isn’t helping yet. In regards to your questions, being a diabetic raises risks for any procedure. For MUA, the risk of fracture could possibly be higher, but depends on other factors as well. You would need to speak to a surgeon about the risk / benefit.

  7. I had shoulder surgery in January 4 months ago for a torn labrum, unstable shoulder ( two anchors) and two bone spurs removed.
    Sling for three weeks, then PT.
    For the last 10 weeks..no significant improvement in ROM or strength.
    PT is thinking I may need MUA. I see doc next week.
    Is frozen shoulder the same thing as developing scar tissue after surgery?
    How do I know which one is happening?
    Since I already had a tear and two anchors placed…are the risked higher to “undo” what was already done if I have MUA?

    1. Lynn,
      Sorry to hear about your troubles. You may be leading to MUA if no improvement. I suggest doing conservative care for 8-12 weeks so you are almost there…

      So frozen shoulder is a general term and not necessarily telling you exactly what is causing it, but yes, it most likely is scar tissue.

      I would recommend asking your surgeon about risk of other complications since you already had anchors placed. His/her answer may depend on your overall health, the length of time after that procedure and/or any other risk factors.

      Good luck!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: