Determining of migraine prognosis using latent growth mixture models  

作　　者：Bahar Tasdelen Aynur Ozge Hakan Kaleagasi Semra Erdogan Tufan Mengi 

机构地区：[1]Department of Biostatistics, Mersin University Faculty of Medicine, Mersin, Turkey [2]Department of Neurology , Mersin University Faculty of Medicine, Mersin, Turkey

出　　处：《Chinese Medical Journal》2011年第7期1044-1049,共6页中华医学杂志（英文版）

摘　　要：Background This paper presents a retrospective study to classify patients into subtypes of the treatment according to baseline and longitudinally observed values considering heterogenity in migraine prognosis. In the classical prospective clinical studies, participants are classified with respect to baseline status and followed within a certain time period. However, latent growth mixture model is the most suitable method, which considers the population heterogenity and is not affected drop-outs if they are missing at random. Hence, we planned this comprehensive study to identify prognostic factors in migraine.Background This paper presents a retrospective study to classify patients into subtypes of the treatment according to baseline and longitudinally observed values considering heterogenity in migraine prognosis. In the classical prospective clinical studies, participants are classified with respect to baseline status and followed within a certain time period. However, latent growth mixture model is the most suitable method, which considers the population heterogenity and is not affected drop-outs if they are missing at random. Hence, we planned this comprehensive study to identify prognostic factors in migraine.

关 键 词：classifcation longitudinal studies ＂ migraine growth mixture models 

分 类 号：F[经济管理]