Alcoholic subtypes based on multiple assessment domains. Validation against treatment outcome.
Alcoholic subtypes based on multiple assessment domains. Validation against treatment outcome.
复制标题
基于多个评估领域的酒精亚型。
DOI：
--
复制
发表时间：
1986
期刊：
影响因子：
--
通讯作者：
R. D. Walker
中科院分区：
文献类型：
--
作者：
D. Donovan;D. Kivlahan;R. D. Walker
近年来，在酗酒人群中的亚型发展领域的研究经历了一些变化。首先，从基于先验比较方法的亚型派生转向最近使用的后验相关方法，如聚类分析。第二，增加了对多变量的关注，将来自若干实际和/或理论上相关的评估领域的多种衡量标准纳入分类过程。第三，更加重视派生集群的外部验证。本章简要回顾了这些问题，使用西雅图退伍军人管理局医疗中心的研究实例，通过一系列专注于酒精亚型发展的研究来说明这些转变中的每一个。最后，将给出最近的亚型识别和验证研究的结果，以说明多域集群分析策略的困难和优势。
Research in the area of subtype development within alcoholic populations has undergone a number of changes over the recent years. First, there has been a shift from subtype derivation based on a priori comparative approaches toward a more recent use of a posteriori correlational approaches such as cluster analysis. Second, there has been an increased multivariate focus in which multiple measures from a number of practically and/or theoretically relevant assessment domains are entered into the classificatory process. Third, there has been an increased emphasis on the external validation of derived clusters. The present chapter briefly reviews these issues, using examples of research from the Seattle Veterans Administration Medical Center to illustrate each of these shifts across a series of studies focusing on the development of alcoholic subtypes. Finally, the results of a recent subtype identification and validation study will be presented to illustrate the difficulties in and advantages of a multidomain cluster analytic strategy.