Diagnostic accuracy of FRAX in predicting the 10-year risk of osteoporotic fractures using the USA treatment thresholds: A systematic review and meta-analysis.
OBJECTIVES: The aim of this study was to conduct a systematic review and meta-analysis on the performance of the WHO's Fracture Risk Assessment (FRAX) instrument in predicting 10-year risk of Major Osteoporotic Fractures (MOF) and Hip Fractures (HF), using the USA treatment thresholds, in populations other than their derivation cohorts.
DESIGN: EMBASE and MEDLINE database were searched with search engine PubMed and OVID as well as Google Scholar for the English-language literature from July 2008 to July 2016. Limiting our search to articles that analyzed only MOF and/or HF as an outcome, 7 longitudinal cohorts from 5 countries (USA, Poland, France, Canada, New Zealand) were identified and included in the meta-analysis. SAS NLMIXED procedure (SAS v 9.3) was applied to fit the Hierarchical Summary Receiver Operating Characteristics (HSROC) model for meta-analysis. Forest plot and HSROC plot was generated by Review Manager (RevMan v 5.3).
RESULTS: Seven studies (n=57,027) were analyzed to assess diagnostic accuracy of FRAX in predicting MOF, using 20% as the 10-year fracture risk threshold for intervention, the mean sensitivity, specificity, and diagnostic odds ratio (DOR) along with their 95% confidence intervals (CI) were 10.25% (3.76%-25.06%), 97.02% (91.17%-99.03%) and 3.71 (2.73-5.05), respectively. For HF prediction, using 3% as the 10-year fracture risk threshold, six studies (n=50,944) were analyzed. The mean sensitivity, specificity, and DOR along with their 95% confidence intervals (CI) were 45.70% (24.88%-68.13%), 84.70% (76.41%-90.44%) and 4.66 (2.39-9.08), respectively.
CONCLUSIONS: Overall, using the 10year intervention thresholds of 20% for MOF and 3% for HF, FRAX performed better in identifying patients who will not have a MOF or HF within 10years, than those who will. A substantial number of patients who developed fractures, especially MOF within 10years of follow up, were missed by the baseline FRAX assessment.