孕产妇VTE预防知信行量表的构建及信效度检验

目的 编制孕产妇静脉血栓栓塞症(venous thromboembolism,VTE)预防知信行量表,并进行信效度检验。方法 采取文献检索和专家焦点访谈方法构建初步的条目池,通过两轮专家咨询,进一步修改条目池,形成初步的量表,并进行预调查。采用便利抽样的方法,选取于2021年9—10月于长沙市某三甲医院产科门诊就诊孕产妇为研究对象,采取临界比值法、相关系数法、Cronbach's α系数法、因子分析法四种方法进行条目筛选,结合专业最终决定条目的去留,并对形成的量表进行信效度检验。结果 共回收有效问卷 308份。形成53个条目的孕产妇VTE预防知信行量表,其中知识维度包含28个条目,态度维度16条目,行为维度9条目。知识、信念、行为三个维度及总量表的Cronbach's α 系数分别是0.969、0.961、0.870、0.964,分半信度分别是0.912、0.901、0.701、0.694,重测信度分别是0.816、0.679、0.715、0.799。内容效度(content validity index,CVI):条目水平内容效度指数(item-level CVI,I-CVI)介于0.83~1间,均大于0.780。量表水平的内容效度指数(scale-level CVI,S-CVI)中平均 S-CVI(average scale-level CVI,S-CVI/Ave)为0.96,全体一致S-CVI(universal agreement scale-level CVI,S-CVI/UA)为0.83。结构效度:探索性因子分析结果显示提取8因子模型,累计贡献率 73.532%;验证性因子分析结果显示误差平方根近似值(root mean square error of approximation,RMSEA)=0.086,增量拟合指数(increnental fit index,IFI)=0.850,塔克-刘易斯指数(Tacker-Lewis index,TLI)=0.841,比较拟合指数(comparative fit index,CFI)=0.849。聚敛效度:各条目因子载荷大于均 0.3,各个潜变量的平均方差抽取量(average variance extracted,AVE)大于0.5,组合信度(construct reliability,CR) 大于0.8。区分效度:知识、态度、行为维度之间显著相关,相关系数为正数且均小于0.5,并小于AVE平方根。结论 本问卷具有良好的信度和效度,可用于孕产妇VTE预防知信行的调查和评估。

Construction of a knowledge-attitude-practice scale of venous thromboembolism prevention in pregnant and lying-in women and its reliability and validity test

Objective To develop a knowledge-attitude-practice scale of venous thromboembolism (VET) prevention in pregnant and lying-in women, and to test its reliability and validity. Methods A preliminary item pool was constructed by literature retrieval and expert focus interview. After two rounds of expert consultation, the item pool was further modified to form a preliminary scale, and a preliminary survey was conducted. The convenience sampling method was adopted to select pregnant and lying-in women who visited department of obstetrics of a third-grade class-A hospital in Changsha from September to October 2021 as the research subjects. Four methods including critical ratio method, correlation coefficient method, Cronbach's α coefficient method and factor analysis method were employed to pick out the items, and the final decision was made based on the specialty. The reliability and validity of the scale were tested. Results A total of 308 valid questionnaires were retrieved. The scale with 53 items was formed, including 28 items concerning knowledge dimension, 16 items concerning attitude dimension and 9 items concerning behavior dimension. Cronbach's α coefficients of knowledge, attitude, practice and total amount scale were 0.969, 0.961, 0.870 and 0.964, respectively. The split reliability was 0.912, 0.901, 0.701 and 0.694, respectively. The retest reliability was 0.816, 0.679, 0.715 and 0.799, respectively. As for content validity index (CVI), item-level CVI (I-CVI) ranged from 0.83 to 1, which were all greater than 0.780. Among scale-level CVI (S-CVI), average scale-level CVI (S-CVI/Ave) was 0.96, and universal agreement scale-level CVI(S-CVI/UA) was0.83. As for structural validity, exploratory factor analysis results showed that the cumulative contribution rate of the 8-factor model was 73.532%. The results of confirmatory factor analysis displayed that root mean square error of approximation (RMSEA), increnental fit index (IFI), Tacker-Lewis index (TLI) and comparative fit index (CFI) were 0.086, 0.850, 0.841 and 0.849, respectively. As for convergence validity, the factor load of each item was greater than 0.3, average variance extracted (AVE) of each latent variable was greater than 0.5, and construct reliability (CR) was greater than 0.8. As for discriminant validity, the dimensions of knowledge, attitude and behavior were significantly correlated, and the correlation coefficient was positive and less than 0.5 as well as AVE square root. Conclusion The scale has good reliability and validity, and can be used for the investigation and evaluation of VTE prevention in pregnant and lying-in women.