Abstract
Introduction
In patients with shoulder pain one of the important initial determinations is to assess the integrity of the rotator cuff. Clinical examination is often inconclusive. Compared with MRI and arthrography, ultrasound allows dynamic evaluation, is non-invasive, less expensive, less time-consuming and more acceptable to patients. The aim of the present study was to evaluate the accuracy of high resolution shoulder ultrasonography compared with arthroscopy in a series of consecutive patients with clinically suspected rotator cuff disease.
Materials and methods
100 shoulders in 99 consecutive patients with shoulder pain who had undergone standardised pre-operative ultrasonography and subsequent arthroscopy between May 2004 and March 2006 were included in the study. There were 53 males and 46 females with a mean age of 59 years. The mean time interval between the ultrasonographic and the arthroscopic examinations was 227 days. For full thickness tears ultrasonography showed a sensitivity of 100%, specificity 83%, positive predictive value 80%, negative predictive value 100% and accuracy 90%. Ultrasonography showed a sensitivity of 83% in detecting partial thickness tears, specificity 94%, positive predictive value 86%, negative predictive value 93% and accuracy 91%.
Discussion
The use of high resolution shoulder ultrasonography to diagnose rotator cuff tears has achieved only limited acceptance by orthopaedic surgeons. Refinements in scanning technique, modern equipment with improved resolution capabilities and clarification of criteria for diagnosing rotator cuff tears have improved the accuracy of shoulder ultrasonography.
Conclusion
In our experience ultrasonography is found to be a highly accurate technique for detecting rotator cuff tears. However the success of an ultrasonographic examination is heavily dependent on the experience of the operator.
