Application And Evaluation Of The New Version Of WHO Fracture Risk Assessment Tool (Frax)

Objective：To analyze the effects of different risk factorson fracture risk by calculating the10-year fracture probabilitiesof Chinese mainland population using the new version of FRAX (Fracture Risk Assessment Tool); and to evaluate the clinicalvalue of FRAX by calculating the10-year fracture probabilitiesin a high-risk population and comparing the results with the NOF(National Osteoporosis Foundation) thresholds to observe thepositive rate just before the actual occurance of fragilityfractures and by comparing the positive results with those of BMD（Body Mineral Density） measurement.Methods：We established four model values:when without femoral neck BMD, the FRAX BMI were10-year hip fracture probability(BMI,Body Mass Index) and10-year major fracture probability(BMI)；when with femoral neck BMD，the FRAX BMD were10-year hip fractureprobability BMD and10-year major fracture probability(BMD)；inorder to study the effects of sex、BMI、BMD and different riskfactors on fracture probabilities using control variable method,meanwhile, we defined patients who had suffered a fragilityfracture as a high-risk population needing anti-osteoporosistreatment. Establishing a time node just before the actualfractures, from December2010to August2013, we questioned198patients treated in our hospital, calculated their10-yearfracture probabilities and compared the results with the NOFthresholds. Patients with a FRAX value lower than the NOFthresholds were considered not detected at high risk by FRAX theday before an osteoporotic fracture (OF). We used paired fourfoldtable chi-square test to compare the positive results of FRAX withthose of BMD measurement.Results：Size of FRAX values were positively correlated withhigh age, low BMI, low BMD and clinical risk factors, and underthe same circumstances, FRAX values of women were generally larger than those of males；the femoral neck BMD played an action ofadjustment and correction action in the operation of the FRAX values. The mean age of the population (n=198) was69.9±10.1yearsold, and78.8%were female. The main fracture sites included:hip(41.9%)、proximal humerus(22.2%)、distal radius(27.8%)、spine(6.6%) and pelvis(1.5%). When not combined with femoral neckBMD, there were only2cases(1.0%) with a10-year major fractureprobability(BMI)≥20%and84cases(42.4%) with a10-year hipfracture probability(BMI)≥3%, in other words, less than half ofthe patients were detected at high risk by FRAX the day beforean OF. For age groups55-59、60-69、70-79and80-90years old,respectively, the positive rates were3.0%、20.5%、66.7%、82.9%(P<0.05). Dual-energy X-ray absorptiometry (DXA) wasavailable for58patients, and there were24cases(41.4%) with aT-score≤-2.5SD(Standard Deviation) and26cases(44.8%) with a-2.5SD<T-score≤-1.0SD(10of them were with ‘previous fracture’and/or ‘secondary osteoporosis’). There were3cases(T-score≤-2.5SD) with a FRAX BMD below the NOF thresholds and6cases(-2.5SD<T≤-1.0SD) with a FRAX BMD equal to or greater than the NOFthresholds. Totally, there were27cases(46.6%) with a FRAX BMI and27cases(46.6%) with a FRAX BMD equal to or greater than theNOF thresholds. Pair-wise comparison of the positive results ofthe three methods were tested by paired fourfold table chi-squarestatistics, and there was no significant difference (P>0.05).Conclusion：FRAX is an economical、convenient and effectiveclinical fracture risk assessment tool. Under the existing NOFthresholds, the Chinese mainland model could only detect only lessthan half of the high-risk population and the positive results ofFRAX had no significant difference with those of BMD. Prospectivepopulation-based studies and pharmacoeconomic researches areneeded to establish ideal Chinese thresholds.