CHARTA Foundation, Milan, Italy.
Value Health. 2013 Jul-Aug;16(5):814-22. doi: 10.1016/j.jval.2013.04.008. Epub 2013 Jun 19.
To estimate a value set for the calculation of Italian-specific quality-adjusted life years (QALYs), based on preferences elicited on EuroQol five-dimensional (EQ-5D) questionnaire health states using the time trade-off technique.
The revised standard Measurement and Valuation of Health protocol was followed. Twenty-five health states, divided into three groups and given to 450 subjects, were selected to obtain 300 observations per state. Subjects aged 18 to 75 years were recruited to be representative of the Italian general adult population for age, sex, and geographical distribution. To improve efficiency, face-to-face interviews were conducted by using the Computer Assisted Personal Interviewing approach. Several random effects regression models were tested to predict the full set of EQ-5D questionnaire health states. Model selection was based on logical consistency of the estimates, sign and magnitude of the regression coefficients, goodness of fit, and parsimony.
The model that satisfied the criteria of logical consistency and was more efficient includes 10 main effect dummy variables for the EQ-5D questionnaire domain levels and the D1 interaction term, which accounts for the number of dimensions at levels 2 or 3 beyond the first. This model has an R(2) of 0.389 and a mean absolute error of 0.03, which are comparable to or better than those of models used in other countries. The utility estimates after state 11111 range from 0.92 (21111) to -0.38 (33333). Italian utility estimates are higher than those estimated in the United Kingdom and Spain and used so far to assess QALYs and conduct cost-utility evaluations in Italy.
A specific value set is now available to calculate QALYs for the conduction of health economic studies targeted at the Italian health care system.
基于时间权衡技术对欧洲五维健康量表（EQ-5D）问卷健康状况的偏好评估，估算用于计算意大利特异性质量调整生命年（QALYs）的效用量值。
遵循修订后的健康测量与评估标准协议。选择了 25 种健康状态，分为三组，分配给 450 名受试者，每个状态获得 300 个观察值。招募年龄在 18 至 75 岁的受试者，以代表意大利普通成年人群的年龄、性别和地域分布。为了提高效率，采用计算机辅助个人访谈方法进行面对面访谈。测试了几种随机效应回归模型来预测 EQ-5D 问卷健康状态的全部数据集。模型选择基于估计的逻辑一致性、回归系数的符号和大小、拟合优度和简约性。
满足逻辑一致性标准且效率更高的模型包括 EQ-5D 问卷域水平的 10 个主效应虚拟变量和 D1 交互项，该交互项考虑了超出第一级的第二级或第三级的维度数量。该模型的 R(2)为 0.389，平均绝对误差为 0.03，与其他国家使用的模型相当或更好。状态 11111 之后的效用估计值范围从 0.92（21111）到-0.38（33333）。意大利的效用估计值高于在英国和西班牙进行的效用评估中估计的数值，并且迄今为止，这些数值一直用于评估意大利的 QALYs 和进行成本效用评估。
现在有了一个特定的效用量值集，可以用于对意大利医疗保健系统进行的健康经济研究计算 QALYs。
