Detection of rotator cuff tears by surgeon-performed ultrasound scan

A. Iossifidis, E Ibrahim, P. Mitra, C. Petrou

Abstract


INTRODUCTION: The aim of this study was to determine whether the greatest diagnostic accuracy for the detection of a surgically relevant rotator cuff tear is provided by a scan in the radiology department, either ultrasound or Magnetic resonance, or by an ultrasound scan performed in the clinic by a shoulder surgeon.

MATERIALS AND METHODS: Consecutive patients due to undergo an arthroscopy for rotator cuff disease were scanned by the operating surgeon. The presence or absence of a surgically relevant cuff tear on scan was compared with arthroscopic findings as gold standard. The surgeon’s log was then compared with the radiologist results of both ultrasound and MRI performed for the same patients in our institution. Subgroup analysis was performed to compare the results of scans reported by specialist and non-specialist radiologists.

RESULTS: Surgeon-performed Ultrasound scan on 88 shoulders was compared to 57 departmental Magnetic resonance scans and 41 departmental Ultrasound scans. Compared to the predictive values obtained for all radiology reported scans combined, Surgeon-US had significantly greater specificity (98.2 v 81.8, p=0.008), Positive predictive value (97.6 v 79.6, p=0.010), and overall accuracy (95.0 v 84.0, p=0.019). No significant difference was found on comparing sensitivity (91.1 v 86.7) or Negative predictive value (93.2 v 88.2).

CONCLUSION: A surgeon performed Ultrasound scan for the detection of surgically relevant rotator cuff tears can provide equal accuracy to departmental Magnetic resonance or specialist radiologist-performed US and better accuracy than US performed by a non-specialist radiologist. These findings support the development of ‘one stop’ shoulder clinics. To our knowledge, this is the first study to directly compare both radiology-reported US and MRI with surgeon-performed US in the same group of patients.


Full Text: PDF HTML

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.