Evaluation of the reproducibility of the Tronzo classification for intertrochanteric fractures of the femur.

Objective: To evaluate the inter-observer reproducibility of the Tronzo classification for transtrochanteric fractures using the kappa concordance coefficient (κ).

Methods: Twenty radiographic images in anteroposterior view on hips with transtrochanteric fractures of the femur were used. These were classified by 12 observers using the Tronzo method. The images were presented in sequence and a questionnaire containing all the options of the Tronzo classification was filled out, along with a simplified classification using Tronzo divisions into two groups (stable and unstable). The data were analyzed by means of the kappa concordance test.

Results: The following kappa indices were found: for images with stable fractures (Tronzo 1 and 2), 0.11; for images with unstable fractures (Tronzo 3, 3 variant, 4 and 5), 0.52; and for the complete classification, 0.44 (moderate concordance). In turn, the simplified classification did not increase the concordance rates.

Conclusion: The Tronzo classification is unsuitable for clinical practice. We suggest that another system should be used or created for this type of fracture.