Reliability of the Tönnis Classification and Its Correlation With Magnetic Resonance Imaging and Intraoperative Chondral Damage.

Abstract

To evaluate the reliability of the Tönnis classification in the setting of femoroacetabular impingement (FAI) hips without dysplasia. Forty-nine patients with FAI underwent preoperative radiography and magnetic resonance imaging (MRI). Radiographs were evaluated in 2 separate settings by 5 observers and graded according to the Tönnis classification. Interobserver and intraobserver reliability was calculated using the κ coefficient. Intraoperative chondral damage was assessed, and chondral damage to the acetabulum (acetabular labrum articular disruption [ALAD] classification) and to the femur (Outerbridge classification) was graded. The Spearman coefficient was computed to quantify the degree of correlation between the Tönnis grade and MRI-detected chondral damage, as well as intraoperative chondral damage. The average intraobserver reliability of the Tönnis classification was moderate (κ = 0.472), and the interobserver reliability was fair (κ = 0.287). Statistically significant positive correlations were found between the Tönnis classification and the ALAD classification (P = .0087) and between the Tönnis classification and femoral chondral damage detected by MRI (P = .0247). A statistically significant correlation was not found between the Tönnis grade and the intraoperative Outerbridge classification of the femur (P = .4969), between the Tönnis grade and acetabular chondral damage on MRI (P = .4969), or between the Tönnis grade and the ability to detect a chondral flap on MRI (P = .2160). No statistically significant correlation was found between the ALAD classification and the presence or absence of a chondral flap on MRI (P = .3538), between the ALAD classification and MRI-detected chondral damage to the acetabulum (P = .103), or between the Outerbridge classification and the degree of chondral damage observed on MRI of the femur (P = .1922). The Tönnis classification and MRI have substantial limitations when evaluating nondysplastic hips with FAI for the degree of chondral damage and arthritis. Level III, retrospective comparative study of prospective data.