摘要目的:应用经典测量理论（CTT）和项目反应理论（IRT）评价电子健康素养Web 3.0量表在糖尿病患者中的条目特性和适用性，为评估糖尿病患者的电子健康素养水平提供新的有效工具。方法:采用横断面研究设计，便利抽样方法选取2022年11月至2023年5月青岛大学附属烟台毓璜顶医院内分泌科住院部的230例糖尿病患者作为调查对象，采用自制的一般资料调查表和电子健康素养Web 3.0量表进行调查。采用CTT中的相关系数法、Cronbach α系数、Guttman折半信度系数、结构效度、区分效度、收敛效度及IRT中的Samejima等级反应模型对量表条目进行评价。 结果:CTT结果显示，电子健康素养Web 3.0量表的Cronbach α系数为0.973，Guttman折半信度系数为0.894，各条目得分与量表总分间的相关系数为0.708~0.845（均 P<0.01）。探索性因子分析提取3个公因子，累计方差贡献率为76.75%，验证性因子分析模型的拟合度良好。各维度组合信度均>0.60，平均方差萃取量>0.50。各维度平均方差萃取量平方根（0.817、0.875、0.845）大于此维度与其他维度的相关系数（0.742、0.763、0.715）。IRT结果显示，各条目区分度为2.174~4.054，难度系数随等级增加而递增（- 2.471~2.187），各条目平均信息量为1.217~2.609。 结论:电子健康素养Web 3.0量表经CTT和IRT检验在糖尿病患者中具有较好的特性，可作为评估我国糖尿病患者电子健康素养水平的有效工具。

abstractsObjective:To evaluate the item properties and applicability of the eHealth Literacy Scale in Web 3.0 context (eHLS-Web 3.0) among diabetic patients by applying classical test theory (CTT) and item response theory (IRT) to provide a new effective tool for assessing the level of eHealth literacy among diabetic patients.Methods:A cross-sectional study design was adopted and 230 diabetic patients from the inpatient endocrinology department of the Affiliated Yantai Yuhuangding Hospital of Qingdao University were selected by convenience sampling from November 2022 to May 2023 and were surveyed by using the self-general information questionnaire and eHLS-Web 3.0. The correlation coefficient method, Cronbach α coefficient, Guttman split-half reliability coefficient, and structural validity in CCT were used, and discriminant validity, convergent validity, and Samejima Graded Response model in IRT were used to evaluate the scale items. Results:The CTT results showed that Cronbach α coefficient for the eHLS-Web 3.0 was 0.973, the Guttman split-half reliability coefficient was 0.894, and the correlation coefficient between the items and the total score of the scale ranged from 0.708 to 0.845 (all P<0.01). Exploratory factor analysis extracted three common factors with a cumulative variance contribution rate of 76.75%, and the confirmatory factor analysis model had a good fit. The reliability of each dimension combination was >0.60, and the average variance extracted was >0.50. The square root of the average variance extracted from each dimension (0.817, 0.875, 0.845) was greater than the correlation coefficient between this dimension and other dimensions (0.742, 0.763, 0.715). The IRT results showed that the discrimination parameter of all the items ranged from 2.174 to 4.054, the difficulty coefficients increased with the increase of the level (- 2.471 - 2.187), and the average information content of all the items ranged from 1.217 to 2.609. Conclusions:The eHLS-Web 3.0 has good characteristics among diabetic patients as tested by CTT and IRT, and can be used as an effective tool to assess the eHealth literacy level of diabetic patients in China.