Abstract
PURPOSE: The primary objective of this study was to investigate the association between patient-reported history elements and anterior cruciate ligament (ACL) tears. The secondary objectives were to evaluate the predictive validity of history elements and develop clinically interpretable decision rules to aid diagnostic reasoning. METHODS: This was a retrospective analysis of data collected during two prospective studies. Patient history elements including demographics, mechanism of injury, signs/symptoms at the time of injury, and subsequent symptoms since the injury were collected using a pre-existing departmental questionnaire and cross referenced with magnetic resonance imaging results. Association and predictive validity were investigated through penalised and unpenalised logistic regressions, and classification and regression tree analyses. RESULTS: Of the 173 included participants, 87 participants (56 males) had an ACL tear with the remainder reported as having posterior cruciate ligament, meniscal, patellofemoral joint, medial and/or lateral knee injuries. There was a significant negative association between seeing deformity (p = 0.028, odds ratio [OR]: 0.091, 95% confidence interval [CI]: 0.011-0.766) or isolated anterior knee pain (p = 0.038, OR: 0.068, 95% CI: 0.005-0.864) at the time of injury, and an ACL tear. Other patient-reported history elements, used individually or in combination, demonstrated limited clinical utility for differentiating ACL tears from other traumatic knee injuries. CONCLUSION: Patient-reported deformity and isolated anterior or medial knee pain at the time of injury were significantly associated with the absence of an ACL tear. Other history elements were not significantly associated with an ACL tear, highlighting the limited clinical utility of history taking and the importance of physical examination after knee injury. LEVEL OF EVIDENCE: Level I.