The purpose of this study was to evaluate the accuracy of community based MRI of the wrist compared to arthroscopic exam. Literature reports of imaging accuracy are largely based upon University center studies and presence of complete tears. In this study, there were multiple center MRI facilities due to regional insurance restrictions. Arthroscopic exam found higher than reported incidence of incomplete interosseous ligament tears.

20 consecutive patients undergoing clinical, radiographic, MRI and arthroscopic evaluation were included in the study. No patient had any prior wrist surgery. A single Hand surgeon examined all patients, reviewed all plain radiographs, MRI’s and performed all arthroscopic surgeries. The Radiologist’s interpretation of the MRI was then compared to the arthroscopic findings. Patients with both contrast and noncontrast MRI were included. All surgical reports included description of the Triangular Fibrocartlage Complex, Scapholunate and Lunotriquetral structures.

The subject group was comprised of 9 males and 11 females. Average age of males was 36 and for females 34. There were 11 workman’s compensation and 9 private insurance patients. Twenty-five percent, 5 of 20, MRI’s had a false negative reading. There were no false positives found. 14 of 20 (70%) had additional finds at arthroscopy that were not reported on the MRI. These included 65% Scapholunte tears, 15% Lunotriquetral tears, and 40% TFCC tears. Of the SL tears not reported, 4 were complete tears and 6 were incomplete. There were no complete Lunotriquetral tears and 3 incomplete Lunotriquetral tears. The overall sensitivity of the MRI was 70% (14 of 20). Only fifteen percent (3 of 20) of MRI’s were 100% accurate.

In conclusion, in this metropolitan area, community based MRI’s are not as accurate as the literature reports the accuracy of MRI’s when compared with arthroscopic findings. Partial tears of the SL and LT ligaments were frequently missed although suspected by clinical exam and confirmed by arthroscopy.