Assessment of the GNRB Arthrometer: High Inter‐Rater Reliability Alone Does Not Support Replacement of the KT1000 in Anterior Cruciate Ligament Reconstruction Patients

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Purpose: To evaluate the inter‐rater reliability and sensitivity of the GNRB arthrometer compared with the KT1000 and clinical examination with the Lachman test for patients with an anterior cruciate ligament (ACL) injury. Methods: Patients with ACL injury were examined preoperatively and postoperatively by 2 examiners. One examiner also performed a clinical examination and an examination with the KT1000. Side‐to‐side difference of ≥3 mm was classified as an ACL injury for both the GNRB and the KT1000. The interclass correlation was used to evaluate inter‐reliability, and a paired t‐test was used to identify differences between examiners. The sensitivity was calculated preoperatively. Results: A total of 20 patients underwent preoperative assessment with the GNRB, whereas 17 were available postoperatively. The GNRB diagnosed 12/20 and 13/20 for the 2 examiners, respectively, as having ACL injuries preoperatively, whereas the KT1000 diagnosed 19/20. Postoperatively, 5/17 and 4/17, respectively, had a side‐to‐side difference of ≥3 mm with the GNRB versus 7/17 for the KT1000, hence indicating an ACL injury or insufficient reconstruction. Clinical examination concluded that 15/17 had a manual Lachman of ≤+1. The interclass correlations between examiners were 0.80 for the involved side preoperatively and 0.84 postoperatively, considered moderate to good. Sensitivity for the GNRB was 60% and 65% for examiners 1 and 2, respectively. For the KT1000, it was 95%. Conclusions: In this study, the GNRB had a high inter‐rater reliability, even with limited experience with the device. However, it was inferior to the KT1000 in correctly identifying ACL deficiency preoperatively in patients with ACL injury. Level of Evidence: Level III, prospective comparative study.

This publication has 20 references indexed in Scilit: