Clinical Accuracy of the Lateral-Anterior Drawer Test for Diagnosing Posterior Cruciate Ligament Rupture

Abstract

Abstract Background Commonly used clinical posterior cruciate ligament (PCL) tests present with diagnostic weaknesses requiring alternative clinical tests. The Lateral-Anterior Drawer test (LAD-test) is a suggested alternative that previously demonstrated concurrent validity in situ. Further in vivo LAD-test clinical accuracy examination is required prior to any recommendation for clinical adoption. Thus, this case–control study aims to (1) investigate the LAD-test’s in vivo interrater and intra-rater reliability; (2) establish LAD-test concurrent validity against MRI as the reference standard; and (3) examine the correspondence between examiners’ professional working experience and LAD-test diagnostic accuracy. Methods Three examiners with different professional experience levels, blindfolded during testing, and blinded from subjects’ identity, medical history, and reference test outcome performed all LAD-testing twice per subject. Reliability analyses included percent agreement, Fleiss’ kappa and Cohen’s kappa coefficients with 95% Confidence Intervals (CIs) and prevalence-adjusted bias-adjusted kappa (PABAK) calculations. Validation parameters included sensitivity, specificity, likelihood ratios (LR + ; LR-), and predictive values (PPV; NPV) each accompanied by 95%CIs; each tester’s percent agreement with the MRI; and their Youden Index. Results The study sample was comprised of 31 subjects of which 14 had a history of unilateral full-thickness PCL-rupture. Their 14 contralateral knees and both knees of 17 healthy subjects served as controls. In vivo LAD-test performance did not produce any negative ramifications for the tested subjects. Interrater reliability was moderate (test-1: Fleiss’κ = 0.41; 95% CI 0.40;0.41; test-2:Fleiss’κ = 0.51; 95% CI 0.50;0.51). Pairwise examiner’s LAD-test outcome agreement ranged from 74 to 89%. Pairwise interrater reliability was fair-to-substantial (κ = 0.27 to κ = 0.65) with moderate-to-substantial PABAK (0.48–0.77). Intra-rater reliability was substantial-to-almost perfect (PABAK 0.65–0.97). Sensitivity and specificity ranged from 57 to 86% and 83 to 98%, respectively. The advanced and novice clinicians’ Youden Indexes were acceptable. The same examiners’ positive likelihood ratios revealed important and relative important effects, respectively. Positive predictive values were considerable for the advanced and novice clinicians, while negative predictive values were high for all examiners. Conclusion Overall, the study results suggested LAD-test practicability. In vivo LAD-test performance did not produce any negative ramifications for the tested subjects. In subjects presenting with a chronic PCL-deficiency (i.e., > 3 months since initial injury), the LAD-test’s clinical accuracy was comparable-to-superior to other commonly used clinical PCL-tests. Future studies to establish the LAD-test’s usefulness in isolation as well as in combination with other clinical tests for acute PCL-rupture diagnostics are warranted. Trial registration number DRKS00013268 (09. November 2017).