TY - JOUR
T1 - The accuracy of high-resolution ultrasound for evaluating focal lesions of the hand and wrist
AU - Teefey, Sharlene A.
AU - Middleton, William D.
AU - Patel, Vikram
AU - Hildebolt, Charles F.
AU - Boyer, Martin I.
PY - 2004/5
Y1 - 2004/5
N2 - Purpose: Although several articles have described the sonographic features of solid and cystic lesions of the hand and wrist, few have investigated its accuracy for diagnosing such lesions. The purposes of this retrospective study were to determine the accuracy of sonography for diagnosing focal lesions of the hand and wrist against the standard of histologic examination of the resected specimen and to compare the accuracy of sonography against the recorded initial clinical impression. Methods: A medical record search for all sonographic studies of the hand and wrist at our institution was performed. Eighty-four patients had correlative surgery and made up the study group. All sonogram reports were reviewed for the absence or presence of focal lesions and specific diagnoses and all medical records were reviewed for the initial clinical impression. Both were compared with histologic findings and tested for agreement with the κ statistic and for significant differences with the McNemar test. Results: Seventy-six of the 84 patients had focal lesions of the hand and wrist and 8 had normal sonograms. Overall, ultrasound correctly diagnosed 83% of the lesions and the initial clinical impression was correct in 54%. Of the 76 lesions ultrasound correctly diagnosed 87% of cystic lesions, 73% of solid lesions, 75% of tenosynovitis cases and the single-vessel thrombosis. The initial clinical impression was correct in 67% of cystic lesions and 75% of tenosynovitis cases. The solid or cystic nature of the 15 solid lesions and the single-vessel thrombosis could not be determined by history or physical examination alone. Of the 35 cases in which the recorded clinical impression was incorrect or not recorded ultrasound provided the correct diagnoses in 71% of these cases. Agreement with the K statistical analysis was only fair and the recorded initial clinical impression was less accurate than ultrasound. Conclusions: Ultrasound was statistically more accurate than the initial clinical impression for distinguishing solid from cystic lesions of the hand and wrist. Ultrasound was very accurate for specifically diagnosing ganglions and slightly less for solid lesions and tenosynovitis.
AB - Purpose: Although several articles have described the sonographic features of solid and cystic lesions of the hand and wrist, few have investigated its accuracy for diagnosing such lesions. The purposes of this retrospective study were to determine the accuracy of sonography for diagnosing focal lesions of the hand and wrist against the standard of histologic examination of the resected specimen and to compare the accuracy of sonography against the recorded initial clinical impression. Methods: A medical record search for all sonographic studies of the hand and wrist at our institution was performed. Eighty-four patients had correlative surgery and made up the study group. All sonogram reports were reviewed for the absence or presence of focal lesions and specific diagnoses and all medical records were reviewed for the initial clinical impression. Both were compared with histologic findings and tested for agreement with the κ statistic and for significant differences with the McNemar test. Results: Seventy-six of the 84 patients had focal lesions of the hand and wrist and 8 had normal sonograms. Overall, ultrasound correctly diagnosed 83% of the lesions and the initial clinical impression was correct in 54%. Of the 76 lesions ultrasound correctly diagnosed 87% of cystic lesions, 73% of solid lesions, 75% of tenosynovitis cases and the single-vessel thrombosis. The initial clinical impression was correct in 67% of cystic lesions and 75% of tenosynovitis cases. The solid or cystic nature of the 15 solid lesions and the single-vessel thrombosis could not be determined by history or physical examination alone. Of the 35 cases in which the recorded clinical impression was incorrect or not recorded ultrasound provided the correct diagnoses in 71% of these cases. Agreement with the K statistical analysis was only fair and the recorded initial clinical impression was less accurate than ultrasound. Conclusions: Ultrasound was statistically more accurate than the initial clinical impression for distinguishing solid from cystic lesions of the hand and wrist. Ultrasound was very accurate for specifically diagnosing ganglions and slightly less for solid lesions and tenosynovitis.
KW - Accuracy
KW - Ganglions
KW - Hand and wrist ultrasound
UR - http://www.scopus.com/inward/record.url?scp=2442638979&partnerID=8YFLogxK
U2 - 10.1016/j.jhsa.2004.02.001
DO - 10.1016/j.jhsa.2004.02.001
M3 - Article
C2 - 15140479
AN - SCOPUS:2442638979
SN - 0363-5023
VL - 29
SP - 393
EP - 399
JO - Journal of Hand Surgery
JF - Journal of Hand Surgery
IS - 3
ER -