Purpose:​ The usefulness of FRAX in predicting major bone fractures in patients with end-stage kidney disease on maintenance hemodialysis treatment has been confirmed in previous studies. For meaningful clinical use, the prognostic and intervention FRAX thresholds need to be established.

Methods:​ The primary aim of our study was to calculate the optimal cut-off point of FRAX for the best prediction of an increased bone fracture risk in dialysis patients and additionally, to propose its intervention threshold, indicating the need for antifracture pharmacological treatment. The study included 718 hemodialysis patients, who were followed up for two years. Thirty low-energy major bone fractures were diagnosed during the study period. We used the Polish version of FRAX (without the DXA examination) and some particular variables of the FRAX calculator. The optimal cut-off point for prediction of an increased major bone fracture risk was based on the analysis of the sensitivity and specificity curves of FRAX.

Results:​ The analysis revealed FRAX >5% (sensitivity of 70.0%, specificity of 69.8%) as the prognostic threshold for major bone fractures. Its sensitivity for bone fracture prediction was significantly higher, but specificity lower than those of FRAX ≥10%, used in general Polish population. The reason for this can be an underestimation of bone fracture risk with FRAX in dialysis patients.

Conclusions:​ We conclude that the FRAX prognostic threshold for identification of an increased risk of major bone fractures in hemodialysis patients is >5%. We propose to use this specific value of FRAX as an intervention threshold for pharmacological antifracture treatment in hemodialysis patients.