精确logistic回归及其SAS应用程序

刘启军; 曾庆; 周燕荣

Abstract

刘启军,曾庆,周燕荣.精确logistic回归及其SAS应用程序[J].Chinese journal of Epidemiology,2003,24(8):725-728

精确logistic回归及其SAS应用程序

Exact logistic regression and its perf ormance to SAS system

Received:September 09, 2002

DOI：

KeyWord: logistic回归精确 SAS 小样本

English Key Word: Logistic regression, exact SAS Small sample

FundProject:

Author Name	Affiliation
LIU Qi-jun	Department of Health Statistics, Chongqing University of Medical Sciences, Chongqing 400016, China
ZENG Qing	Department of Health Statistics, Chongqing University of Medical Sciences, Chongqing 400016, China
ZHOU Yan-rong	Department of Health Statistics, Chongqing University of Medical Sciences, Chongqing 400016, China

Hits: 4851

Download times: 2492

Abstract:

目的　精确logistic回归在最大似然法估计结果不可靠或者不存在情况下,作为传统logistic回归的一种补充方法可适用于小样本资料、资料结构不平衡及高度分层资料。方法　对37例乳腺癌患者术后预后资料采用SAS进行精确logistic回归分析。结果　发现组织学分级、淋巴结转移两因素对乳腺癌的术后预后具有统计学意义。结论　在样本量很小,最大似然法估计无效的情况下,精确logistic回归是一种极好的分析手段。

English Abstract:

Objective To explore the feasibility of exact logistic regression,used as a complemental method for the maximum liklihood estimation,and to analyse with data small sample,unbalanced structure and highly stratal nature under the situations of questionable results or inexistence of the maximum likelihood estimation.Methods Data from 37 postoperative breast cancaer cases were analyzed in 1997 by exact logistic regression under SAS system.Results Data caculated by SAS software showed that Quasi-complete separation of data points was detected but the results of maximum likelihood estimation did not exist,SAS outputs conflicted the results of the last max imum likelihood iteration (likelihood Chi-square and score Chi-square have similar P,less than 0.05, but the Wald chi-square had a larger P,more than 0.05).Under conditional exact parameter estimation,it appeared that:(1)the joint effect of conditional score statistics was 21.12 with P =0.0003;(2)for individual parameters,the effect conditional score statistics of histological classification(grades)was 5.80 with P=0.0208;axillary node metastasis (diversion)was5.74 with P=0.0195;tumor size(size)was 0.79,with P=0.6476.The effects of tumor histolog ical classification and axillary node metastasis were statistically significant on breast cancaer tumour.Conclusion Exact logistic regression seemed to be a very useful method in analy zing data from small sample when the maximum likelihood estimation was either with no effect or did not exist

View Fulltext Html FullText View/Add Comment Download reader