Reliability testing of the Larsen and Sharp classifications for rheumatoid arthritis of the elbow

详细信息    查看全文

• 作者：Nicholas B. Jew ; MD^a ; Anthony M. Hollins ; MD^b ; Benjamin M. Mauck ; MD^a ; Richard A. Smith ; PhD^a ; Frederick M. Azar ; MD^a ; Robert H. Miller III ; MD^a ; Thomas W. Throckmorton ; MD^a ; ^{tthrockmorton@campbellclinic.com}
• 关键词：Basic Science Study ; Validation of Classification System
• 刊名：Journal of Shoulder and Elbow Surgery
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：26
• 期：1
• 页码：140-143
• 全文大小：163 K
• 卷排序：26

文摘

Two popular systems for classifying rheumatoid arthritis affecting the elbow are the Larsen and Sharp schemes. To our knowledge, no study has investigated the reliability of these 2 systems. We compared the intraobserver and interobserver agreement of the 2 systems to determine whether one is more reliable than the other.MethodsThe radiographs of 45 patients diagnosed with rheumatoid arthritis affecting the elbow were evaluated. Anteroposterior and lateral radiographs were deidentified and distributed to 6 evaluators (4 fellowship-trained upper extremity surgeons and 2 orthopedic trainees). Each evaluator graded all 45 radiographs according to the Larsen and Sharp scoring methods on 2 occasions, at least 2 weeks apart.ResultsOverall intraobserver reliability was 0.93 (95% confidence interval [CI], 0.90-0.95) for the Larsen system and 0.92 (95% CI, 0.86-0.96) for the Sharp classification, both indicating substantial agreement. Overall interobserver reliability was 0.70 (95% CI, 0.60-0.80) for the Larsen classification and 0.68 (95% CI, 0.54-0.81) for the Sharp system, both indicating good agreement. There were no significant differences in the intraobserver or interobserver reliability of the systems overall and no significant differences in reliability between attending surgeons and trainees for either classification system.ConclusionThe Larsen and Sharp systems both show substantial intraobserver reliability and good interobserver agreement for the radiographic classification of rheumatoid arthritis affecting the elbow. Differences in training level did not result in substantial variances in reliability for either system. We conclude that both systems can be reliably used to evaluate rheumatoid arthritis of the elbow by observers of varying training levels.