School of Public Health and Health Systems, University of Waterloo, Waterloo, Ontario, Canada.
Departments of Paediatrics and Epidemiology and Biostatistics, Western University, London, Ontario, Canada.
Epilepsia. 2018 Mar;59(3):668-678. doi: 10.1111/epi.14008. Epub 2018 Jan 28.
The aim of this study was to develop and validate a brief version of the Quality of Life in Childhood Epilepsy Questionnaire (QOLCE). A secondary aim was to compare the results described in previously published studies using the QOLCE-55 with those obtained using the new brief version.
Data come from 373 children involved in the Health-related Quality of Life in Children with Epilepsy Study, a multicenter prospective cohort study. Item response theory (IRT) methods were used to assess dimensionality and item properties and to guide the selection of items. Replication of results using the brief measure was conducted with multiple regression, multinomial regression, and latent mixture modeling techniques.
IRT methods identified a bi-factor graded response model that best fits the data. Thirty-nine items were removed, resulting in a 16-item QOLCE (QOLCE-16) with an equal number of items in all 4 domains of functioning (Cognitive, Emotional, Social, and Physical). Model fit was excellent: Comparative Fit Index = 0.99; Tucker-Lewis Index = 0.99; root mean square error of approximation = 0.052 (90% confidence interval [CI] 0.041-0.064); weighted root mean square = 0.76. Results that were reported previously using the QOLCE-55 and QOLCE-76 were comparable to those generated using the QOLCE-16.
The QOLCE-16 is a multidimensional measure of health-related quality of life (HRQoL) with good psychometric properties and a short-estimated completion time. It is notable that the items were calibrated using multidimensional IRT methods to create a measure that conforms to conventional definitions of HRQoL. The QOLCE-16 is an appropriate measure for both clinicians and researchers wanting to record HRQoL information in children with epilepsy.
本研究旨在开发和验证儿童癫痫生活质量问卷（QOLCE）的简短版本，并比较使用新简短版本与之前使用 QOLCE-55 发表的研究结果。
数据来自参与多中心前瞻性队列研究——儿童癫痫生活质量研究的 373 名儿童。使用项目反应理论（IRT）方法评估维度和项目特性，并指导项目选择。使用多元回归、多项回归和潜在混合建模技术对简短测量的结果进行复制。
IRT 方法确定了一个双因素等级反应模型，该模型最适合数据。删除了 39 个项目，形成了一个包含 16 个项目的 QOLCE（QOLCE-16），在所有 4 个功能领域（认知、情感、社会和身体）中都有相同数量的项目。模型拟合度非常好：比较拟合指数为 0.99；塔克-刘易斯指数为 0.99；均方根误差近似值为 0.052（90%置信区间[CI] 0.041-0.064）；加权均方根为 0.76。之前使用 QOLCE-55 和 QOLCE-76 报告的结果与使用 QOLCE-16 生成的结果相当。
QOLCE-16 是一种多维健康相关生活质量（HRQoL）测量方法，具有良好的心理测量学特性和较短的估计完成时间。值得注意的是，项目使用多维 IRT 方法进行校准，以创建符合 HRQoL 常规定义的测量方法。QOLCE-16 是临床医生和研究人员在记录癫痫儿童 HRQoL 信息时的一种合适的测量方法。
