基于概化理论与项目反应理论的肾病综合征生命质量测定量表评价

Evaluation of quality of life instrument for nephrotic syndrome basedon generalization theory and item response theory

目的本研究旨在评价慢性病之肾病综合征测定[QLICD-NS(V2.0)]量表评估性能。方法选取2021年3-11月广东医科大学附属医院肾内科确诊的203例肾病综合征(NS)患者进行QLICD-NS(V2.0)量表的测评,评价方法为概化理论(GT)与项目反应理论(IRT)。利用Multilog 7.0软件进行IRT的等级反应模型(GRM)分析得出每个条目的难度、区分度系数和信息量。结果GT结果显示QLICD-NS(V2.0)量表4个领域的全域总分贡献比分布均匀,4个领域的概化系数均>0.50。除社会功能外,其他3个领域的被试方差分量均大于条目方差分量,各领域的可靠性指数均>0.50。IRT结果显示QLICD-NS(V2.0)量表的区分度为0.82;除了条目TNS7和TNS8,其余条目的难度系数均-3~3且呈单调递增。结论QLICD-NS(V2.0)量表在生理功能、心理功能领域的信度较好,在社会领域和特异功能领域尚可,说明本研究研制的QLICD-NS(V2.0)量表性能良好。

Objective To evaluate the performance of the Quality of Life Instrument for Chronic Disease Nephrotic Syndrome[QLICD-NS(V2.0)]in patients with nephrotic syndrome.Methods A total of 203 patients with nephrotic syndrome(NS)diagnosed in the Department of Nephrology,Affiliated Hospital of Guangdong Medical University from March 2021 to November 2021 were selected for QLICD-NS(V2.0)evaluation,and the evaluation methods were generalization theory(GT)and item response theory(IRT).The difficulty,discrimination coefficient and information amount of each item were obtained by using Multilog 7.0 software to analyze the grade response model(GRM)of IRT.Results The results of GT showed that the contribution ratio of the global total score in the four domains of QLICD-NS(V2.0)scale was evenly distributed,and the generalization coefficient of the four domains was greater than 0.50.Except for social function,the variance component of the participants in the other three domains were greater than the item variance component,and the reliability index of each domain was greater than 0.50.The results of IRT showed that the discrimination degree of QLICD-NS(V2.0)scale was 0.82.Except for items TNS7 and TNS8,the difficulty coefficients of the other items ranged from-3 to 3 and increase monotonically.Conclusion The QLICD-NS(V2.0)scale has good reliability in physiological function and psychological function,and is acceptable in social field and special function.The QLICD-NS(V2.0)scale developed in this study has good performance.