Abstract

Objectives: The aim of this study was to determine intra- and interobserver reliability of AO/ASIF and Jensen classification systems, and to compare reliability when applied by unexperienced and experienced orthopaedic surgeons.

Patients and Methods: The anteroposterior and lateral radiographs of 60 intertrochanteric hip fractures were reviewed and classified by two groups (G1, G2) of orthopaedic surgeons using the AO and Jensen classification systems on two separate occasions three months apart. Each group consisted of five orthopaedic surgeons. Group 1 had less than five years, and group 2 had more than 10 years of experience. Kappa statistical analysis was used for determination of intra- and interobserver variation.

Results: For the AO classification system without subgroups, the mean kappa value was 0.67 (range 0.47-0.90) for intraobserver variation and 0.42 (range 0.10-0.73) for interobserver variation. For the Jensen classification, the mean kappa value was 0.57 (range 0.35-0.80) for intraobserver variation and 0.30 (range 0.10-0.60) for interobserver variation. For the AO classification system with subgroups, the mean kappa value was 0.49 (range 0.21-0.81) for intraobserver variation and 0.23 (range 0.09-0.51) for interobserver variation.

Conclusion: Although these classification systems have disadvantages, this study suggests that AO system without subgroups is more useful than Jensen and AO system with subgroups to classify intertrochanteric fractures of the proximal femur.