Garvan Nomogram法评估城区中老年人骨折风险效果分析  

Effect of Garvan Nomogram on prediction of fracture risk among urban middle-aged and elderly people

作　　者：王俊[1,2] 王雪君 方贞 

机构地区：[1]宁波市江东区明楼街道社区卫生服务中心,浙江宁波315040 [2]宁波市江东区疾病预防控制中心 [3]宁波市江东区东柳街道社区卫生服务中心

出　　处：《中国公共卫生管理》2017年第4期476-479,共4页Chinese Journal of Public Health Management

基　　金：宁波市科技计划项目(2014C50061)

摘　　要：目的探索应用澳大利亚的骨折风险评估方法 Garvan Nomogram评估法在国内中老年居民中的适用性。方法对50~89岁422名居民进行调查及骨密度(BMD)检测,用Garvan Nomogram方法进行评估。应用配对t检验比较根据体重和股骨颈BMD的预测风险情况,以ROC曲线下的面积(AUC)对不同Garvan Nomogram预测的准确性进行评估。结果根据体重预测5年和10年髋部骨折风险分别为0.65%和1.26%,男性显著高于女性;5年和10年任意部位骨折风险分别为7.06%和14.00%,男性显著低于女性。根据股骨颈BMD预测5年和10年髋部骨折风险分别为2.61%和4.97%,5年和10年任意部位骨折风险分别为8.48%和16.52%。根据BMD预测骨折风险显著高于根据体重预测风险。骨量正常、骨量减少和骨质疏松对象的预测骨折风险呈递增状态。根据体重预测的骨折风险对骨质疏松的预测准确性较低,AUC均<0.70;根据BMD预测的骨折风险对骨质疏松的预测有一定准确性,AUC为0.80左右。研究对象1年随访发生骨折9例(2.13%)。该方法对一年骨折风险的预测准确性较低,AUC均<0.70。结论 Garvan Nomogram方法可通过临床危险因素来预测骨折风险,根据股骨颈BMD预测风险较根据体重预测风险要更准确。Objective To study the applicability of the fracture risk assessment method of Garvan Nomogram from Australian to predict fracture among middle-aged and elderly population in China. Methods A questionnaire was adopted and bone mineral density（ BMD） were tested among 422 residents aged 50-89. The prediction was conducted by using of Garvan Nomogram method. Paired t test was used to compare the Garvan-based probability with weight and that with femur neck（ FN）BMD. Area under curve（ AUC） of ROC analysis was used to assess the accuracy of Garvan-based probability. Results The Garvan-based hip fracture probability of 5-years and 10-years with weight were 0. 65% and 1. 26%,respectively,the probability among males was higher than that of females; the any fracture probability were 7. 06% and 14. 00%,respectively,the probability among males was lower than females. The Garvan-based hip fracture probability of 5-years and 10-years with femur neck BMD were 2. 61% and 4. 97%,respectively,the any fracture probability of 5-years and 10-years were 8. 48%and 16. 52%. The probability with BMD was higher than that with weight. The probability among them with normal BMD,osteopenia and osteoporosis was increasing in turn. The accuracy of the Garvan-based probability with weight to predict osteoporosis was low（ AUC〈0. 70）,while the accuracy of the Garvan-based probability with BMD was better（ AUC about0. 80）. The any fracture rate was 2. 13%（ 9 subjects with fracture） after 1 year follow-up. The accuracy of the Garvan-based probability to predict 1-year fracture was low（ AUC〈0. 70）. Conclusion Garvan nomogram model can predict the osteoporosis fracture risk with clinical risk factors,the accuracy of probability with fermord neck BMD is better than that with weight.

关 键 词：Garvan nomogram评估方法 骨折 骨质疏松症 骨密度 

分 类 号：R161.7[医药卫生—公共卫生与预防医学]