• 
• 

Three Part Question

In [adult patients attending the emergency department with a suspected rotator cuff injury] is [ultrasound scan a reliable and accurate] tool at [confirming the diagnosis]?

Clinical Scenario

A 55-year-old woman attends the Emergency Department following a fall, during which she sustained an injury to her left shoulder. She complains of pain and is unable to abduct her arm to more than 90 degrees. An x-ray shows no fracture or dislocation and you suspect a rotator cuff injury. You know that MR imaging is a reliable method of confirming this type of injuries but also that it is expensive and time-consuming.

You wonder if ultrasound is as an accurate imaging modality at confirming a rotator cuff injury.

Search Strategy

Google Scholar:
The search run week 2 march 2014 with the terms “ultrasound” AND “rotator cuff” yielded about 14,000 results. For pragmatic reasons only the first 100 papers underwent abstract review for potential articles of high level of evidence. Two systematic reviews and meta-analysis were found. The most recent only was included.
Ovid MEDLINE:
The search was run week 4 March 2014.
Search strategy: {[(exp Shoulder/) OR (shoulder.mp) OR (exp Rotator cuff/) OR (rotator cuff.mp)] AND [(exp ultrasonography/) OR (ultrasonography.mp) OR (ultrasound.mp) OR (ultrason$.mp)]}
The search run was restricted to the past four years only, as the above-mentioned systematic review and meta-analysis was published in 2010. The search was further limited to humans, English language and adult population.
363 papers were retrieved for abstract review. Three were deemed relevant to the clinical question but one was published before the above-mentioned SR.

EMBASE:
Search strategy: [{(exp Rotator cuff/) OR (exp Shoulder/) OR (exp Shoulder girdle/)] AND (exp Ultrasound/) OR (exp Ultrasound scanner/) OR (exp Echography/)]}.
Search run week 4 March 2014 and limited to adults, humans, English language and period 2010 – 2014. 334 papers were found but after abstracts review only three were deemed relevant to the clinical questions (one paper was a duplicate from the Medline search. One was published before the systematic review and was therefore not included for appraisal).
AMED and CINAHL:
Searched with the EMBASE terms week 4 March 2014. No yield.
ClinicalTrials.gov database:
Search run with terms “shoulder” OR “rotator cuff” week 4 March 2014.
586 studies found but only one deemed relevant. It however had an unknown recruitment status and could therefore not be included in our review.
Cochrane Library:
Search using the term “rotator cuff. No reviews were identified.

Search Outcome

4 papers are included in the table

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Sipola P et al, 2010, Finland	77 patients with suspected rotator cuff disorder planned for surgery after 3 months of unsuccessful conservative treatment Cohort underwent USS, MRI and MRA scans	Prospective comparative study	Diagnostic accuracy for USS, MRI and MRA	Sensitivity for undifferentiated tears was 92%, specificity 45%, PPV 91%, NPV 50% Sensitivity for full – vs. partial thickness tear was 83%, specificity 53%, PPV 84%, NPV 50%	Poor presentation of results Convenience sample Two patients did not undergo MRA scan Included degenerative rotator disease USS performed by radiologists (generalisability) Teres minor muscle not analysed
Singisetti K et al, 2011, England	90 patients (96 shoulders) underwent USS for detection of supraspinatus and subscapularis muscles partial or complete tear and this was compared to arthroscopy	Prospective comparative study	Sensitivity, specificity, PPV, NPV for USS and arthroscopy in detecting tears	For supraspinatus tendon, USS sensitivity was 89%, specificity 43%, PPV 76%, NPV 100% For subscapularis tendon tear, USS had a sensitivity of 30%, specificity 100%, PPV 100%, NPV 78%	No confidence interval were provided USS performed by a single radiologist Only supraspinatus and subscapularis muscles were analysed
Smith TO et al, 2011, England	62 diagnostic studies (including 6007 patients and 6066 shoulders) that directly compared the accuracy of USS to arthroscopy/surgical findings Cadaveric/animal studies, patients with tendinosis/tendinopathy excluded No language restrictions	Systematic review and meta-analysis	Pooled sensitivity and specificity for both full - and partial - thickness tear ROC curves Covariate analysis for different frequency transducer heads Effect of different operators background	Partial - thickness tear pooled sensitivity 0.84 (95% CI 0.82-0.86), pooled specificity 0.89 (95% CI 0.87-0.91) Full - thickness tear pooled sensitivity 0.96 (95% CI 0.95-0.97), pooled specificity 0.93 (95% CI 0.91-0.95) Lower accuracy for ultrasonographers and general radiologists compared to musculoskeletal radiologists	Some included studies were not performed by intention-to-treat analysis Potential assessor bias as surgeons not always blinded to USS findings
Youssef MA et al, 2013, Egypt	30 patients with suspected rotator cuff disorder underwent USS and MRI scan The results were compared to arthroscopic and surgical findings	Prospective comparative study	Sensitivity, specificity and accuracy of USS and MRI in detecting rotator cuff disease	USS sensitivity was 97.7%, specificity 98.2%, accuracy 98% (p<0.05)	Looked at cuff disorders (including tendinosis) No confidence intervals provided No subgroup provided for the different muscle tendons

Comment(s)

Rotator cuff disorders accounts for about 10% of all shoulder pain (Meislin) and are a common reason for presentation to Emergency Departments. Clinical examination is frequently considered difficult and unreliable to predict the exact type and extent of injury. Radiological imaging is often required to confirm or refute diagnosis. MRI however is still expensive and time-consuming in comparison to ultrasonography. Furthermore, the portability of USS means that the integrity of the rotator cuff can be assessed by a non-radiologist during a clinic session to allow efficient planning of treatment.

The systematic review by Smith seems to suggest that USS has a diagnostic accuracy high enough to detect full-thickness rotator cuff tears. It is however less accurate in detecting partial-thickness ruptures.

The further collated and appraised studies (Sipola,Youssef, Singisetti) showed different levels of accuracy and this is probably due to several factors like operator’s background and experience, frequency of the transducer head used and sampling.

The meta-analysis however contained sufficient data to provide an indication for the effect of operator profession and unsurprisingly the accuracy was the greatest under the direction of a musculoskeletal radiologist.

It was interesting to note that orthopaedic surgeons performed better than general radiologists (sensitivity 0.95 vs. 0.89 and specificity 0.91 vs. 0.87). This is probably due to their familiarity and knowledge of shoulder anatomy. With the increasing use of USS by non-radiologist in one-stop clinics, this observation should prompt further research on the training type and length required to perform and interpret ultrasound images in this cohort of patients.

Editor Comment

Abbreviations: USS ultrasound scan, MRI magnetic resonance imaging, MRA MRI arthrogram, PPV positive predictive value, NPV negative predictive value

Clinical Bottom Line

Ultrasonography is accurate enough to detect selected ruptures of the rotator cuff muscles. Further studies are however needed to precisely determine the level of training required for its accuracy in the hands of non-radiologists.

References

1. Meislin RJ, Sperling JW, Stitik. TP. Persistent shoulder pain: epidemiology, pathophysiology and diagnosis. Amj J Orthop 2005; 34:5-9.
2. Sipola P, Niemitukia L, Kroger H et al. Detection and quantification of rotator cuff tears with ultrasonography and magnetic resonance imaging - a prospective study in 77 consecutive patients with a surgical reference. Ultrasound in Med. & Biol 2010:36; 1981-1989.
3. Singisetti K, Hinsche A. Shoulder ultranography versus arthroscopy for the detection of rotator cuff tears: analysis of errors. Journal of Orthopaedic Surgery 2011: 19: 76-9.
4. Smith TO, Back T, Toms AP et al. Diagnostic accuracy of ultrasound for rotator cuff tears in adults: a systematic review and meta-analysis. Clinical Radiology 2011: 66; 1036-1048.
5. Youssef MA, Teima AH, Abduo YE et al. Ultrasonographic and MR diagnosis of rotator cuff disorders & shoulder joint instability. The Egyptian Journal of Radiology and nuclear medicine 2013: 44; 835-844.