Classification and Treatment of Pediatric Tibial Spine Fractures: Assessing Reliability Among a Tibial Spine Research Interest Group.

Background: Treatment decisions for patients with tibial spine fractures depend heavily on radiographic measurements. The purpose of this study was to determine whether existing classification systems and radiographic measurements are reliable among a multicenter tibial spine research interest group. A secondary purpose was to evaluate agreement in treatment of tibial spine fractures.

Methods: Using a deidentified radiographic imaging series and identical imaging software, we examined the interobserver and intraobserver reliability of the Meyers and McKeever classification, as well as a cohort of measurements of tibial spine fractures and treatment recommendations. Forty patients were included based on previous reliability studies. Interobserver and intraobserver data were analyzed using kappa and intraclass correlation coefficient reliability measures for categorical and continuous variables, respectively.

Results: Good interobserver reliability was seen with superior displacement measurements of the anterior portion of the tibial spine fracture (0.73, 0.78) and excellent intraobserver reliability with an intraclass correlation coefficient of 0.81. Several measurements demonstrated moderate interobserver and intraobserver reliability including posterior-proximal displacement, and length and height of the tibial spine fracture. Moderate intraobserver reliability was seen with a majority of measurements and classification schemata (0.42 to 0.60) except for a poor agreement in posterior-sagittal displacement (0.27). Classifying tibial spine fractures according to the original Meyers and McKeever classification demonstrated fair agreement [κ=0.35, 0.33 (inter); 0.47 (intra)]. When combining Type III and IV, agreement increased for both reviews [κ=0.42, 0.44 (inter); 0.52 (intra)]. A total of 24 (60%) fractures were classified as 3 different types. There was fair agreement in both reviews regarding open reduction (either open or arthroscopic) versus closed reduction for initial treatment [κ=0.33, 0.38 (inter); 0.51 (intra)].

Conclusions: Measurement of superior displacement of the anterior portion of tibial spine fractures on the lateral images is the only radiographic assessment with good interobserver and intraobserver reliability. Reliability of radiographic measurements and a modified classification for tibial spine fractures remains fair, and perhaps unacceptable, even among a group of pediatric sports medicine specialty-trained surgeons.

Level Of Evidence: Level III-diagnostic reliability study of nonconsecutive patients.