欧洲生命质量量表模型的拓展
Expansion of the EQ-5D model

目的:构建以假设状态为基础的欧洲生命质量量表(EQ-5D)拓展模型。方法:采用分层整群随机抽样的方法,抽取200名在校医学生作为应答者。利用L18(2×37)正交设计表选取18个假设健康状态; 健康状态的评分采用视觉标尺法; 在五因素EQ-5D模型(ModelA#)的基础上加入性别、年龄、经济状况3个因素,构建EQ-5D拓展模型(ModelA); 采用广义最小二乘回归法进行数据拟合,构建模型。通过模型逆回归构建以实际体验为基础的五因素EQ-5D模型(ModelSun#)和拓展模型(ModelSun)。分别用这4个模型对22个典型健康状态进行评分。结果:ModelA的模型系数均为负值,与ModelSun系数方向一致,大小略有不同。ModelA#与ModelSun评分的相关系数为0.870 3,ModelA与ModelSun评分的相关系数为0.968 8,较前者提高。结论:ModelA模型合理,预测精度较ModelA#提高,应用范围扩大。
Aim: To build up the EQ-5D developed model based on hypothetical health states.Methods: Using a stratified cluster random sampling method, 200 medical students were selected to be the respondents.A total of 18 hypothetical health states were designed by L18(2×37),and the health states were scored by visual analogue scale method.The three factors including gender, age, and economic status were incorporated into EQ-5D model(ModelA#)to construct EQ-5D developed model(ModelA)fitted by GLS method. The EQ-5D models based on experience health states without(Model Sun#)or including gender, age, and economic status(ModelSun)were constructed by reverse regression. The 4 models were used to score 22 typical hypothetical health states.Results: The direction of model coefficients of ModelA were consistent with ModelSun, and the coefficients were both minus but slightly different. The correlation coefficient between ModelA# and ModelSun was 0.870 3, and that between ModelA and ModelSun was 0.968 8,higher than the former.Conclusion: The EQ-5D developed model(ModelA)has higher prediction accuracy compared with ModelA#, its application value is elevated