Ultrasound imaging has become an essential adjunct to clinical examination when assessing a patient with suspected rotator cuff pathology. With the new high-resolution portable machines it has become feasible for the shoulder surgeon to perform the scans himself in the clinic and save a great deal of time. This study was conducted to examine the accuracy of the ultrasound scans performed by a single surgeon over a period of four years (2001-2004).

The ultrasound findings were uniformly documented and collected prospectively. Out of a total of 364 scanned patients we selected 143 who ultimately received an operation and we compared the surgical findings with the ultrasound reports. The intra-operative findings included 77 full thickness supraspinatus tears, 24 partial thickness tears and 42 normal cuffs. Three full thickness tears were missed on ultrasound and reported as normal/ partially torn. Four normal/ partially torn cuffs were thought to have a full thickness tear. This presents 96.3% sensitivity and 94.3% specificity for full thickness tears. Three partial thickness tears were reported normal on ultrasound and eight normal cuffs were thought to have partial thickness tears. This presents 89% sensitivity and 93.7% specificity for partial thickness tears. The size estimation of full thickness tears was more accurate for large/massive tears (96%) than moderate (82%) and small/pinhole tears (75%). The tear sizes were more often underestimated which may partly reflect disease progression during the unavoidable time lag between scan and surgery.

We conclude that shoulder ultrasound performed by a sufficiently trained orthopaedic surgeon is a safe and reliable practice to identify rotator cuff tears.