Reliability analysis of CT torsion assessment after closed cephalomedullary nailing of trochanteric femoral fractures. A comparison study of six established methods.

Background: Posttraumatic maltorsion and implant failure after closed reduction of proximal femoral fractures remain a cause of concern. Although the reproducibility of torsion measuring techniques on CT for femoral shaft fractures has been thoroughly analyzed, little is known about the trochanteric fractures. Apart from the well-known CT limitations, posttraumatic alteration of bony landmarks makes torsional assessment even more challenging. Main goal of this study was to examine the reliability of different CT techniques on trochanteric femoral fractures after closed nail fixation. Secondary goal was to see whether the measurements within the examined population were influenced by the fracture type and patient age or BMI.

Methods: 20 cases (AO.31-A1 or -A2) were retrospectively examined. Six established CT techniques for torsional assessment were performed from three different investigators twice at different time points. The intraclass correlation coefficient (ICC for 95% CI) was used to analyze the interobserver and intraobserver reliability.

Results: The Hernandez method (0.986) followed by the Jend method (0.982) by a mean difference of <1° showed the highest reliability. Although increasing fracture complexity from A1 to A2 led to an overall worsening of the measurement precision, the Hernandez and Jend techniques revealed a very good consistency. Within the examined population, age and BMI had no impact on the precision of the measurements.

Conclusions: The Hernandez and Jend methods represent reliable alternatives for torsional assessment of trochanteric femur fractures treated with closed nail fixation when compared to the other measurement techniques here involved. Documentation of the torsion measuring method used in each case constitutes an essential element of the radiological reports.