摘要
对111例Graves病初发患者用他巴唑治疗6月,停药后随访1年。根据病人一般临床及生化特征,用logistic回归分析法,建立了一个具有4个指标的数学预测模型,其预测停药后1年的预后,具有较高的敏感性(81.5％)和特异性(84.8％),符合率达82.9％(92/111),其结果优于本资料中任何有预测意义的单因素分析。
One hundred and twelvenew cases of Graves′disease were treatedby tapazole for 6 monsths and followed upfor another 12 months. The initial dosewas 30 mg/d. Clinical and biochemical eu-thyroidism was achieved within 1 to 3months, then a maintenance was givenuntil cessation of drugs at 6 mouths.Onehundred and eleven cases completedthe study. Remission and relapse weredefined at the end of follow-up for 1 yearaccording to the presence or absence ofclinical manifestation of hyperthyroidismand the levels of T_3 and T_4. The results of the 12-month follow-upShowed that 46 of 111 cases were in remis-sion, and the remaining 65 cases sufferedrelapse. A logistic regression model with4 variables was established, which includ-ed thyroid suppression rate and goitre sizeby palpation at the end of drug treat-ment, the level of T_3 before therapy,and the patients′age. The model had 81.5%sensitivity, 84.8% specificity and 82. 9%(92/111) accuracy in predicting the out-come of Graves′disease after withdrawalof drug for 1 year. The results were muchbetter than any other univariate analysisin this study.
出处
《华西医科大学学报》
CAS
CSCD
1991年第3期314-317,共4页
Journal of West China University of Medical Sciences
关键词
甲状腺肿
预后
LOGISTIC回归
Graves' disease
Drug treatment
logistic regression model
Estimation of prognosis
