Abstract
BACKGROUND: Recurrent patellar dislocation (RPD) is a multifactorial disease that affects young and active people. Patellar height measurements are used clinically to screen and diagnose knee conditions. However, there are no known studies that have assessed and compared the performance of patellar height indices for predicting the incidence of RPD, which could be used to recommend surgical treatment after primary patellar dislocation. This case-control study aimed to determine if the patellar height index could be used to predict the incidence of RPD, and to identify the optimal method in terms of its diagnostic ability for RPD. METHOD: Altogether, 133 patients (52 patients with RPD [Group R] and 81 sex- and age-matched controls [Group C]) were enrolled in this study. The Insall-Salvati (IS), Blackburne-Peel (BP), Caton-Deschamps (CD), and modified IS (mIS) methods were used to measure the patellar height index. The intra-observer and inter-observer reliabilities of these four methods were determined using intraclass correlation coefficients. A receiver operating characteristic curve analysis was performed to evaluate the predictive ability of each index and identify the cut-off values that indicated significantly increased risk of RPD. RESULTS: Patient demographics were similar between the two groups. The inter-observer and intra-observer reliabilities were good for all four methods. In patients with RPD, the mean index values for the four methods were significantly higher than those in the matched controls. The area under the curve (AUC) values for IS, BP, CD, and mIS were 0.91 (standard error [SE], 0.03; 95% confidence interval [CI], 0.84-0.96), 0.72 (SE, 0.05; 95% CI, 0.63-0.81), 0.86 (SE, 0.03; 95% CI, 0.79-0.92), and 0.96 (SE, 0.01; 95% CI, 0.94-0.99), respectively. CONCLUSION: Patellar height indices had high predictive performance for the incidence of RPD. The mIS method had the highest AUC.