Skip to main content

03-11-2011 | Surgery | Article

‘Real-life’ diagnostic accuracy questioned for ultrasound rotator cuff injury


Free abstract

MedWire News: Real-life ultrasound use for the detection of rotator cuff injuries may not achieve the high levels of diagnostic accuracy reported in clinical trials, suggest study findings from a UK district general hospital.

The analysis, published in the journal Shoulder & Elbow, demonstrated "disappointingly low" sensitivity and specificity for rotator cuff injury when ultrasound was used to examine patients undergoing shoulder arthroscopy compared with reports showing the technique can be as effective as magnetic resonance (MR) imaging.

"The poor results seen reflect the specialist nature of the imaging modality and the learning curve required to perform the scans," suggest Tim McBride (West Midlands Deanery, UK) and Keshav Mathur (Alexandra Hospital, Redditch).

To find out how accurate pre-operative ultrasound is for diagnosing rotator cuff tears in the setting of a district hospital, the team examined data for all 85 patients who underwent ultrasound before shoulder arthroscopy over a 1-year period at a midsized district hospital.

Ultrasound was performed by a single radiologist in two-thirds of the patients, and patients waited an average of 6.2 months for surgery after imaging. In all, 69% of patients had rotator cuff tears detected by the gold standard of arthroscopic investigation. The majority (71%) of tears were full thickness.

Analysis showed that ultrasound was 66% sensitive and 54% specific for the diagnosis of rotator cuff tears, giving an overall accuracy of 62%. The positive and negative predictive values were 76% and 41%, respectively.

Sensitivity fell to just 48% and 35% for detection of full thickness and partial thickness tears, respectively, although specificity increased to 79% and 76%, respectively.

The researchers say their findings have "implications not only for the diagnostic process, but also in terms of theatre time logistics."

McBride and Mathur therefore conclude: "We would recommend a regular combined radiology/shoulder surgery multidisciplinary team meeting in the department not only to feedback the intra-operative findings, but also as a useful learning tool for both specialities, and as a forum for the discussion of more complex cases.

"Should the results of ultrasound scanning continue to be below par, then MR imaging could be considered as a useful diagnostic alternative."

By Lynda Williams

Related topics