High Reliability of a Scoring System for Implant Position in Undisplaced Femoral Neck Fractures.

Abstract

The aim of this study was to evaluate the intrarater and interrater reliability of a scoring system for internal fixation (SIFA) in undisplaced femoral neck fractures (FNF). A reproducibility study with a historical consecutive cohort. The patients were all treated at a level one trauma hospital. A cohort of patients with FNF, treated with 2 parallel implants, was retrieved using a local county register, and the cohort was assessed for undisplaced FNF until the desired sample size was reached. One hundred two patients had their x-ray images reviewed 2 times with a minimum 14 days interval, by 3 raters, all blinded for each other's results, and an independent person analyzed the data with unweighted kappa statistics. Implant positions on the first postoperative anterior-posterior and axial radiographs were first visually assessed, then objectively measured according to SIFA, which consists of 5 individual measurements; screw tip distance, screw femoral shaft angle, inferior screw calcar distance, and screw position in the femoral head in the anterior-posterior and axial radiographs. Visually assessed implant position gave intrarater kappa results for the individual components of SIFA at 0.40-0.75 (rater 1), 0.75-0.98 (rater 2), and 0.69-0.83 (rater 3), with interrater kappa results at 0.18-0.80 (combined range). Objectively measured implant position for the individual components of SIFA gave intrarater kappa results at 0.56-1.0 (rater 1), 0.74-0.89 (rater 2), and 0.81-0.94 (rater 3), with interrater kappa results at 0.56-0.85 (combined range). The overall SIFA score has similar higher intrarater and interrater results for the measured (0.51-0.78) compared with the visual assessment (0.13-0.78). The assessment in the axial view of the screw position in femoral head had the largest difference of acceptable positioning, when comparing visual assessment with the measured. The SIFA demonstrates substantial intrarater and interrater reliability (kappa 0.51-0.78), when raters make measurements of radiographs, but lower reliability (kappa 0.13-0.78) when raters visually assess the radiographs without measuring. The SIFA should be assessed on a large cohort, for correlation with clinical evaluation.