Poster 111: Poor Inter-observer Reliability of TT-TG Measurement Among Orthopaedic Sports Medicine Surgeons Regardless of Measurement Technique

OBJECTIVES: Tibial tubercle-trochlear groove distance (TT-TG) is an important radiologic parameter in the diagnosis and treatment of patellar instability. However, there is substantial heterogeneity in the literature regarding the most appropriate measurement protocol for assessing TT-TG distance. The objective of this study was to evaluate the inter- and intra-observer reliability of TT-TG distance measurement in healthy knees performed by fellowship-trained orthopaedic sports medicine surgeons. METHODS: A convenience sample of 29 magnetic resonance imaging scans (MRIs) of the knee was assembled after screening to ensure there was no ligamentous rupture, meniscal pathology, cartilage injury, subchondral edema, or abnormal masses in the bone or muscle. An a priori power analysis determined that this number of MRIs would provide greater than 90% power to detect a statistically significant difference in intraclass correlation of 0.95 (a < 0.05). Three fellowship-trained orthopaedic sports medicine surgeons read the MRIs and recorded the TT-TG distance using their method of choice. After a minimum of 5 months, the MRIs were read again, this time using a standardized evidence-based method to measure the TT-TG distance. Kappa statistics were calculated to evaluate inter- and intra-observer reliability. RESULTS: Among the initial TT-TG distance measurements, kappa was equal to -0.006 (P = 0.690) when assessing inter-observer reliability. Introduction of the standardized measurement protocol was associated with no significant change in inter-observer reliability (kappa = 0.124, P = 0.070). The individual TT slice that was selected for TT-TG distance measurement demonstrated a kappa statistic of 0.192 (P < 0.001) for the surgeons’ method of choice and a kappa statistic of 0.037 (P = 0.455) for the standardized method. Similarly, TG slice selection using the surgeons’ method of choice yielded a kappa statistic of -0.005 (P = 0.949) and a kappa statistic of -0.102 (P = 0.140) for the standardized method. CONCLUSIONS: Our findings demonstrate poor inter-observer reliability of TT-TG distance measurement in healthy knees that was not improved with the application of a standardized evidence-based measurement protocol. TT and TG slices selected by surgeons varied widely. Future research should focus on comparing TT-TG distance measurements across a wider variety of knees (ranging from healthy to those with patellar instability), as this may improve the inter-observer reliability of TT-TG measurement.