Abstract：
OBJECTIVE: We have developed a new patient-centered, preference-based generic health-outcome measure, Château-Santé Base (CS-Base), which is based on a novel multiattribute preference response (MAPR) measurement framework. This study aimed to generate a first utility set for the CS-Base, making it suitable for use in health-economic evaluations.
METHODS: CS-Base comprises 12 health attributes: mobility, vision, hearing, cognition, mood, anxiety, pain, fatigue, social functioning, daily activities, self-esteem, and independence, each with 4 levels. Our methodology to generate utilities for the CS-Base was 2-fold. First, we derived coefficients from patient MAPR data to calculate CS-Base values. Subsequently, these were normalized to a 0.0 to 1.0 utility scale, in which 0.0 signifies dead. The dead position was estimated using general population data from a discrete choice experiment (discrete choice experiment + dead), using a division-value strategy, which localize the position of states better or worse than dead.
RESULTS: We analyzed MAPR data from 3222 patients and discrete choice experiment + dead data from 1995 respondents. All MAPR coefficients were negative, logically ordered, and significantly different from the reference level. The dead position was denoted by a division value of -148.385. Utility values spanned from -0.071 to 1.0, and only 53 of 16 777 216 states were deemed worse than dead.
CONCLUSIONS: This study introduced the first CS-Base utility set, underlining a 2-step utility derivation method. This method, blending societal and patient views, surpasses traditional preference-based approaches, yielding firmer results. However, improvement of the normalization procedure is expected. Estimating CS-Base utilities is an ongoing process that gains precision over time.
METHODS: CS-Base comprises 12 health attributes: mobility, vision, hearing, cognition, mood, anxiety, pain, fatigue, social functioning, daily activities, self-esteem, and independence, each with 4 levels. Our methodology to generate utilities for the CS-Base was 2-fold. First, we derived coefficients from patient MAPR data to calculate CS-Base values. Subsequently, these were normalized to a 0.0 to 1.0 utility scale, in which 0.0 signifies dead. The dead position was estimated using general population data from a discrete choice experiment (discrete choice experiment + dead), using a division-value strategy, which localize the position of states better or worse than dead.
RESULTS: We analyzed MAPR data from 3222 patients and discrete choice experiment + dead data from 1995 respondents. All MAPR coefficients were negative, logically ordered, and significantly different from the reference level. The dead position was denoted by a division value of -148.385. Utility values spanned from -0.071 to 1.0, and only 53 of 16 777 216 states were deemed worse than dead.
CONCLUSIONS: This study introduced the first CS-Base utility set, underlining a 2-step utility derivation method. This method, blending societal and patient views, surpasses traditional preference-based approaches, yielding firmer results. However, improvement of the normalization procedure is expected. Estimating CS-Base utilities is an ongoing process that gains precision over time.
摘要：
目的：我们开发了一种新的以患者为中心，基于偏好的通用健康结果衡量标准，CS-Base,它基于一种新的多属性偏好响应(MAPR)测量框架。这项研究旨在为CS-Base生成第一个实用程序集，使其适用于卫生经济评估。
方法：CS-Base包含12个健康属性：移动性，愿景,听力,认知，心情,焦虑,疼痛,疲劳,社会功能,日常活动,自尊,独立,每个都有四个层次。我们为CS-Base生成实用程序的方法是双重的。首先,我们从患者MAPR数据中导出系数以计算CS-Base值.随后，这些被归一化为0.0-1.0效用量表，其中0.0表示“死”。“死亡”位置是使用来自离散选择实验(DCE+死亡)的一般人群数据估计的，使用“除法值”策略，将状态的位置定位为比死亡更好或更糟糕。
结果：我们分析了3,222名患者的MAPR数据和1,995名受访者的DCE+死亡数据。所有MAPR系数均为负，逻辑有序，与参考水平有显著差异。“死”位置由-148.385的除法值表示。效用值从-0.071到1.0，在16,777,216个州中只有53个州被认为比死亡更差。
结论：本研究引入了第一个CS-Base实用程序集，强调两步效用推导法。这种方法,融合了社会和患者的观点，超越传统的基于偏好的方法，产生更坚实的结果。然而,归一化程序的改进是预期的。估计CS-Base实用程序是一个持续的过程，随着时间的推移而获得精度。
方法：CS-Base包含12个健康属性：移动性，愿景,听力,认知，心情,焦虑,疼痛,疲劳,社会功能,日常活动,自尊,独立,每个都有四个层次。我们为CS-Base生成实用程序的方法是双重的。首先,我们从患者MAPR数据中导出系数以计算CS-Base值.随后，这些被归一化为0.0-1.0效用量表，其中0.0表示“死”。“死亡”位置是使用来自离散选择实验(DCE+死亡)的一般人群数据估计的，使用“除法值”策略，将状态的位置定位为比死亡更好或更糟糕。
结果：我们分析了3,222名患者的MAPR数据和1,995名受访者的DCE+死亡数据。所有MAPR系数均为负，逻辑有序，与参考水平有显著差异。“死”位置由-148.385的除法值表示。效用值从-0.071到1.0，在16,777,216个州中只有53个州被认为比死亡更差。
结论：本研究引入了第一个CS-Base实用程序集，强调两步效用推导法。这种方法,融合了社会和患者的观点，超越传统的基于偏好的方法，产生更坚实的结果。然而,归一化程序的改进是预期的。估计CS-Base实用程序是一个持续的过程，随着时间的推移而获得精度。
