Should we repair rotator cuffs?

Should we repair rotator cuff tears - and if so, how?

Mr Lennard Funk MSc, FRCS(Tr&Orth)

Prepared for the MSc Orthopaedic Engineering, University of Wales, 2005

In this presentation we shall explore the literature on rotator cuff tears and the evidence for repairing cuff tears.   There is great variation amongst surgeons as to the management of rotator cuff tears biased by experience and their understanding of the literature, skills levels and regional variations.   I therefore undertook to try and obtain as balanced a view as possible from the reliable and relevant studies in the current literature.  


Are cuff tears normal?


There is adage amongst Orthopaedic Surgeons that "grey hair equals cuff tear" and the general teaching has been that cuff tears are a normal part of the aging process.   Sher et al in 1995 looked at asymptomatic patients using Magnetic Resonance Imaging (MRI).   They found that 54% of asymptomatic shoulders over of the age of 60 years had evidence of a rotator cuff tear, 28% of those were full thickness tears, and 26% were partial thickness tears.   However other studies comparing MR scans with arthroscopy have demonstrated that MR scanning is only 75-95% accurate in diagnosing rotator cuff tears.   Therefore the true incidence of degenerate tears of the rotator cuff of asymptomatic shoulder is not exactly known.


Is Sub Acromial Decompression Alone Adequate for Treatment of Rotator Cuff Tears?


Hoe-Hanson et al in 1999 looked at 39 patients undergoing sub acromial decompression alone.   13 had small cuff tears less than 2 cms in size, 13 partial tears and 13 patients with intact rotator cuffs.   On reviewing the patients at six years they found that there was no significant difference between the groups with regards to range of motion, strengths and pain relief with no difference in the Constant scores.   They concluded that the functional outcome after Arthroscopic Sub Acromial Decompression was not obviously related to the preoperative degree of cuff pathology where there were small rotator cuff tears.  


A similar study performed by Massoud et al (JBJS 2002) looked at sub acromial decompression alone for chronic small and medium sized cuff tears in 118 shoulders between two and five years after sub acromial decompression.   They found a satisfactory outcome in 59% of patients under the age of 60 and 87.5% of those over the age of 60.   An unsatisfactory outcome was related to manual work, duration of symptoms of more than twelve months and younger age.   Further surgery in the form of a rotator cuff repair was required in 20% of the total number of patients in the study.   They found that half of those tears had increased in size.  


Ellman et al in Arthroscopy 1993 prospective study selected patients into three groups based on the size of tear measured during surgery, physiological aids, handedness and activity level.   They found that small tears (less than 2cms) in older individuals not involved in strenuous activities did well with sub acromial decompression alone.   Patients with larger tears (2-4cms) did poorly (50% satisfactory).   With massive irreparable tears the pain relief was satisfactory on a limited goals basis but there was no improvement in strength or range of motion.   Their study emphasised the importance of patient selection based on the above criteria in selecting the treatment for rotator cuff tears.  


In other studies looking at sub acromial decompression alone and comparing it with rotator cuff repairs (Wiley 1991, Zyliac 1194, Godsman 1997, Savoie 1997) all found that with large rotator cuff tears there was deterioration over time and the results of the sub acromial decompression alone was inferior to rotator cuff repair.  


In summary of the data thus far, it would appear that Arthroscopic Sub Acromial Decompression alone is good for pain relief in the presence of rotator cuff tears in elderly patients (>60 years) on a limited goals basis, patients with low functional demand and massive irreparable tears in elderly patients with low functional demand.   Only one study (O'Hanson et al) demonstrated good results for sub acromial decompression alone for full thickness rotator cuff tears of a small size.  


What is the natural history of Rotator Cuff Tears?


Yamanaka (Clin Orthop, 1994) followed 40 partial articular surface cuff tears over 14 months and found that 10% had healed, 10% had decreased in size and 80% enlarged to become full thickness tears.

Yamaguchi (JSES, 2001) longitudinally followed asymptomatic rotator cuff tears over a 5-year period to assess the risk for development of symptoms and tear progression using ultrasound scans and clinical assessments. 50% of the group developed symptomatic cuff tears during the 5-year period. 50% of the entire group's tears increased in size. No patient had a decrease in size of tear. Only 22% of the remaining asymptomatic patients had progression of their tears. Thus, even without tear progression 78% developed symptomatic tears.


Therefore, it would appear that articular surface partial tears generally progress to full thickness tears. The full thickness tears are at risk of progressing and full thickness tears are likely to become symptomatic.

What are the results of rotator cuff repair?

The published results of rotator cuff repairs have been variable. This may be related to patient selection and older techniques.

  Summary of few large studies:

Author, year



Tear size


Murray, Snyder 2002


Av 39 m

Medium and Large


Jones, FH Savoie 2003


1 - 5 yr

Large and Massive  


Wolf   20 04


4 - 10 yr

Full thickness


La, Burkhart 2004


Av 23 m

Large and Massive



Author, year


Mean Age (years)

Average Follow-up (months)

Excellent & good clinical results

Cuff integrity (imaging method)

Wolf et al,   2004 (20)





Not done

Bouileau et al, 2005 (16)





71% (CTA)

Charousset et al, 2008 (19)





75% (CTA)

  *LaFosse et al, 2007 (18)





89% (CTA)

*Huijsmans et al, 2007 (21)






  [* indicates studies with double row rotator cuff repair technique only]

Most studies have only reported the size of the cuff tear and patient age in relation to the results. However, our understanding of the importance of the rotator cuff muscle quality (atrophy and fatty infiltration) in relation to repair results is now clearer.

We also appreciate the relevance of the broad insertion area of the rotator cuff on the humerus, known as the footprint area. We now attempt to recreate this with our repair techniques.

Progressive degenerative cuff tears are have poorer tendon quality than traumatic tears in good tendon tissue and are more likely to fail. Repair of these tears is likely to fail and we need to look at alternative methods to manage these patients, other than rotator cuff repair.

In summary, based on our understanding of the published literature and experience - Yes!
But… this does depend on a number of factors, not all of which are easy to objectively quantify. A traumatic rotator cuff tear in a young, active patient would be an indication for early repair. An atraumatic, degenerate tear in an elderly, low-demand patient would be not be a good candidate for repair. However, the grey zone between those two extremes is less clear and it is best to apply principles of shared decision making. In fact, one of the key advances of modern medical practice, in the current era of internet based information explosion, is such shared decision making between the surgeon and the patient. Patients must be involved in decisions regarding their health , especially when is comes to surgical treatment. Hence, decisions to proceed with rotator cuff repair are best supported with evidence based discussions regarding management options, complications, recovery times and anticipated outcomes.



This website is certified by Health On the Net Foundation. Click to verify. This site complies with the HONcode standard for trustworthy health information:
verify here. satisfies the INTUTE criteria for quality and has been awarded 'editor's choice'.

The material on this website is designed to support, not replace, the relationship that exists between ourselves and our patients. Full Disclaimer