骨质疏松高风险人群中医辨识量表的测试优化研究

A study of testing and optimization of the traditional Chinese medicine identification scale for high-risk population of osteoporosis

目的:对骨质疏松高风险人群中医辨识量表进行测试优化。方法:以2020年9月至2021年1月在北京中医药大学第三附属医院、中国中医科学院望京医院及北京市西城区广外医院门诊就诊的患者为调查对象,要求调查对象完成骨质疏松高风险人群中医辨识量表、国际骨质疏松基金会(International Osteoporosis Foundation,IOF)骨质疏松风险一分钟测试题、亚洲人骨质疏松症自我筛查工具(osteoporosis self-assessment tool for Asians,OSTA)量表、疼痛视觉模拟量表(visual analogue scale,VAS)、简明健康状况调查表(short form 36 health survey questionnaire,SF-36)及骨密度检测(包括腰椎和髋部)。评价骨质疏松高风险人群中医辨识量表的可行性、信度、效度及判别效果。结果:①可行性评价结果。共发放骨质疏松高风险人群中医辨识量表250份,回收236份,有效量表228份,回收率94.4%,有效率91.2%。调查对象完成骨质疏松高风险人群中医辨识量表的时间均≤5 min。所有调查对象均同时完成骨密度检测及其余4个量表,调查对象完成所有量表的时间均≤20 min。②信度评价结果。骨质疏松高风险人群中医辨识量表的克龙巴赫α系数为0.882、Spearman-Brown系数为0.891,其中针对前10个条目的克龙巴赫α系数为0.770。③结构效度评价结果。经探索性因子分析,骨质疏松高风险人群中医辨识量表的32个条目中,手足烦热和健忘2个条目的负荷值在所有因子上均<0.4,予以剔除。对剩余30个条目重新进行探索性因子分析,共提取出9个公因子,累积方差贡献率为61.215%。9个公因子分别代表下肢症状、头目症状、疼痛症状、睡眠情绪、头面体征、寒热症状、躯体体征、躯体症状、头面症状9个维度,对应的中医证候体现了骨质疏松症患者肾精亏虚、肾阳虚、肝肾阴虚、肝郁血虚、脾胃虚弱等病机特点,也与原发性骨质疏松症中医证型基本相符。对量表中前10个条目进行探索性因子分析,共提取出3个公因子,累积方差贡献率为61.200%。3个公因子分别代表躯体症状、疼痛及躯体体征3个维度。应用SPSS AMOS24软件对上述3因子分析模型进行验证性因子分析,结果显示模型拟合较好。④效标关联效度评价结果。骨质疏松高风险人群中医辨识量表前10个条目评分和量表总分(30个条目)与骨密度分类、腰椎骨密度、髋部骨密度及SF-36中8个方面得分均呈负相关(骨密度分类:r=-0.190,P=0.004;r=-0.231,P=0.000;腰椎骨密度:r=-0.144,P=0.030;r=-0.198,P=0.003;髋部骨密度:r=-0.218,P=0.001;r=-0.272,P=0.000;SF-36生理功能:r=-0.280,P=0.000;r=-0.330,P=0.000;SF-36生理职能:r=-0.324,P=0.000;r=-0.346,P=0.000;SF-36身体疼痛:r=-0.430,P=0.000;r=-0.434,P=0.000;SF-36总体健康:r=-0.460,P=0.000;r=-0.495,P=0.000;SF-36活力:r=-0.220,P=0.001;r=-0.292,P=0.000;SF-36社会功能:r=-0.282,P=0.000;r=-0.326,P=0.000;SF-36情感职能:r=-0.174,P=0.009;r=-0.259,P=0.000;SF-36精神健康:r=-0.251,P=0.000;r=-0.320,P=0.000),与OSTA指数、疼痛VAS评分均呈正相关(OSTA指数:r=0.153,P=0.021;r=0.140,P=0.035;疼痛VAS评分:r=0.453,P=0.000;r=0.430,P=0.000);IOF骨质疏松风险一分钟测试结果与前10个条目评分不存在相关性(r=0.127,P=0.055),与量表总分(30个条目)呈正相关(r=0.168,P=0.011)。⑤判别效果。以骨密度诊断结果为金标准,基于骨质疏松高风险人群中医辨识量表前10个条目评分和量表总分(30个条目)判别骨量异常的受试者操作特征曲线下面积分别为0.636[95%CI(0.564,0.709)]和0.650[95%CI(0.579,0.721)],最佳截断值分别为2分和5分,即符合骨质疏松高风险人群中医辨识量表前10个条目中的2个及以上,或符合全部30个条目中的5个及以上,就可判定为骨量异常,为骨质疏松高风险人群。结论:骨质疏松高风险人群中医辨识量表具有较好的可行性和信度;优化后的量表具有良好的效度,具备一定的判别骨量异常的能力,可用于骨质疏松高风险人群的筛查评估。

Objective:To test and optimize the traditional Chinese medicine(TCM)identification scale for high-risk population of osteoporosis(OP).Methods:Patients who were treated in Beijing University of Chinese Medicine Third Affiliated Hospital,Wangjing Hospital of CACMS and Beijing Xicheng Guangwai Hospital from September 2020 to January 2021 were selected as the subjects.The participants were asked to complete the TCM identification scale for high-risk population of OP,the International Osteoporosis Foundation(IOF)one-minute OP risk check,the osteoporosis self-assessment tool for Asians(OSTA)scale,the pain visual analogue scale(VAS),the short form 36 health survey questionnaire(SF-36)and bone mineral density(BMD)measurement(including lumbar spine and hip).The feasibility,reliability,validity and discriminative effects of the TCM identification scale for high-risk population of OP were evaluated.Results:①Results of feasibility evaluation.Two hundred and fifty TCM identification scale questionnaires for high-risk population of OP were handed out,and 236 ones returned,among which 228 ones were valid,with the response rate of 94.4%and the effective rate of 91.2%,and the time for participants to finish the scale did not exceed 5 minutes.All participants finished the BMD measurement and the other 4 scales synchronously,and the time for them to finish all scales was less than 20 minutes.②Results of reliability evaluation.The Cronbach’sαcoefficient of the TCM identification scale for high-risk population of OP was 0.882,and the Spearman-Brown coefficient was 0.891,with a Cronbach’sαcoefficient of 0.770 for the first 10 items.③Results of structural validity evaluation.As indicated by exploratory factor analysis,among the 32 items in the TCM identification scale for high-risk population of OP,2 items,i.e.,“vexing heat in the extremities”and“forgetfulness”,had factor loadings of<0.4 across all factors and were therefore excluded.An exploratory factor analysis was re-performed on the remaining 30 items,9 common factors were extracted,with a cumulative variance contribution rate of 61.215%.The 9 common factors represented 9 dimensions,including lower limb symptoms,head and eye symptoms,pain symptoms,sleep emotions,head-face signs,cold and heat symptoms,physical signs,physical symptoms,and head-face symptoms.These dimensions corresponded to the TCM syndromes of OP patients,including kidney-essence deficiency,kidney-yang deficiency,liver-kidney yin deficiency,liver-qi stagnation and blood deficiency,and spleen-stomach weakness,which were generally consistent with the TCM syndromes of primary OP.Another exploratory factor analysis was performed on the top 10 items of the scale,3 common factors were extracted with a cumulative variance contribution rate of 61.200%.These 3 common factors represented 3 dimensions,i.e.,physical symptoms,pain signs and physical signs.The models generated from the above 3 common factors analyses were further validated through confirmatory factor analysis using SPSS AMOS24 software,and the results indicated a good model fit.④Results of criterion-related validity evaluation.The scores of the top 10 items of the TCM identification scale for high-risk population of OP and the total score(30 items)were negatively correlated with BMD classification,lumbar spine BMD,hip BMD and scores of 8 aspects in SF-36(BMD classification:r=-0.190,P=0.004;r=-0.231,P=0.000;lumbar spine BMD:r=-0.144,P=0.030;r=-0.198,P=0.003;hip BMD:r=-0.218,P=0.001;r=-0.272,P=0.000;SF-36 physical function:r=-0.280,P=0.000;r=-0.330,P=0.000;SF-36 role-physical:r=-0.324,P=0.000;r=-0.346,P=0.000;SF-36 bodily pain:r=-0.430,P=0.000;r=-0.434,P=0.000;SF-36 general health:r=-0.460,P=0.000;r=-0.495,P=0.000;SF-36 vitality:r=-0.220,P=0.001;r=-0.292,P=0.000;SF-36 social function:r=-0.282,P=0.000;r=-0.326,P=0.000;SF-36 role-emotional:r=-0.174,P=0.009;r=-0.259,P=0.000;SF-36 mental health:r=-0.251,P=0.000;r=-0.320,P=0.000).The scores of the top 10 items of the TCM identification scale for high-risk population of OP and the total score(30 items)were positively correlated with the OSTA index and pain VAS score(OSTA index:r=0.153,P=0.021;r=0.140,P=0.035;pain VAS score:r=0.453,P=0.000;r=0.430,P=0.000).The results of IOF one-minute OP risk check were not correlated with the scores of the top 10 items(r=0.127,P=0.055),but were positively correlated with the total score(30 items)of the scale(r=0.168,P=0.011).⑤Discriminative effects.With BMD diagnosis results as the gold standard,based on the scores of the top 10 items and the total score(30 items)of the TCM identification scale for high-risk population of OP,the discriminative effects of the TCM identification scale for high-risk population of OP was evaluated by using the receiver operating characteristic(ROC)curve,and the areas under the curve(AUC)of participants with abnormal bone mass were 0.636(95%CI(0.564,0.709))and 0.650(95%CI(0.579,0.721))respectively,with the optimal cutoff values of 2 points and 5 points respectively,which suggested that participants who scored 2 points or more on the top 10 items or 5 points or more on all 30 items of the TCM identification scale could be identified as abnormal bone mass and considered as individuals at high risk of OP.Conclusion:TCM identification scale for high-risk population of OP has good feasibility and reliability.The optimized scale has good validity and a certain ability to discriminate abnormal bone mass,and can be used for screening and evaluation of high-risk population of OP.