Epidemiology of hip fracture and the development of FRAX in Ukraine

• Published in: Archives of osteoporosis Vol. 12; no. 1; p. 53
• Main Authors: Povoroznyuk, VV, Grygorieva, NV, Kanis, JA, EV, McCloskey, Johansson, H, Harvey, NC, Korzh, MO, Strafun, SS, Vaida, VM, Klymovytsky, FV, Vlasenko, RO, Forosenko, VS
• Format: Journal Article
• Language: English
• Published: London Springer London 01-12-2017
• Subjects: Aged; Aged, 80 and over; Algorithms; Endocrinology; Female; Hip Fractures - epidemiology; Hip Fractures - etiology; Humans; Incidence; Male; Medicine; Medicine & Public Health; Middle Aged; Models, Theoretical; Original; Original Article; Orthopedics; Osteoporotic Fractures - epidemiology; Osteoporotic Fractures - etiology; Probability; Risk Assessment - methods; Risk Factors; Ukraine - epidemiology; Hip fractures; FRAX; Austria; Epidemiology; Ukraine
• ISSN: 1862-3522; 1862-3514
• DOI: 10.1007/s11657-017-0343-2

Summary
A country-specific FRAX model has been developed for the Ukraine to replace the Austrian model hitherto used. Comparison of the Austrian and Ukrainian models indicated that the former markedly overestimated fracture probability whilst correctly stratifying risk.
Introduction
FRAX has been used to estimate osteoporotic fracture risk since 2009. Rather than using a surrogate model, the Austrian version of FRAX was adopted for clinical practice. Since then, data have become available on hip fracture incidence in the Ukraine.
Methods
The incidence of hip fracture was computed from three regional estimates and used to construct a country-specific FRAX model for the Ukraine. The model characteristics were compared with those of the Austrian FRAX model, previously used in Ukraine by using all combinations of six risk factors and eight values of BMD (total number of combinations =512).
Results
The relationship between the probabilities of a major fracture derived from the two versions of FRAX indicated a close correlation between the two estimates (
r
 > 0.95). The Ukrainian version, however, gave markedly lower probabilities than the Austrian model at all ages. For a major osteoporotic fracture, the median probability was lower by 25% at age 50 years and the difference increased with age. At the age of 60, 70 and 80 years, the median value was lower by 30, 53 and 65%, respectively. Similar findings were observed for men and for hip fracture.
Conclusion
The Ukrainian FRAX model should enhance accuracy of determining fracture probability among the Ukrainian population and help to guide decisions about treatment. The study also indicates that the use of surrogate FRAX models or models from other countries, whilst correctly stratifying risk, may markedly over or underestimate the absolute fracture probability.