摘要
生存率是评价慢性疾病尤其恶性肿瘤诊洽效果的常用指标。临床试验方面的小样本生存资料的统计分析一般采用Kaplan-Meter乘积限估计,并常以生存率曲线的形式表达。寿命表法比较适用于大样本,分组生存资料;常用于流行病学随访资料、医院肿瘤登记资料、人群肿瘤登记资料的生存率分析。为了适应越来越多的肿瘤医疗或登记机构在肿瘤生存分析方面的需求,我们采用传统的寿命表方法,在微机上开发了一个用BASIC语言编写的生存分析程序:LIFTAB。
Survival analysis is a statistical method frequently used in the analysis of datacollected from clinical trials epidemiological follow-up studies. hospital-based cancerregistry and population-based cancer registry.The large statistical softwares or packagessuch as BMDP.SAS.EPILOG and so on.are able te perform different survival anslyses,but they usually occupy large spaces of hardware in the microcomputers. and are hardlyto use. In addition, there is limitation on number of case records. and the univariate analyses are usually Kaplan-Meier Product-limit estimate in most packages. LIFTABis a microcomputer program which is written in interpreted BASIC for calculatingthe survivorship function and its standard error by using life table method, andsummary measures of survival times such as 50% survival time. The average age ofpatient group is also provided. The nummber of case records to be analyzed is unlimited.So it is very suitable for statistical analysis for large data set in a microcomputer. Thestatistical tests by using nonparametrie test of logrank for two or more groups, and thestatistical comparison for rates at different time points between two groups, can also beperformed. Again, the program is able to perform statistical test for trend analysis forseveral patient groups by grouping according to category prognostic variables. The rowBASIC code is compiled by interpreted BASIC, so it may run under the MS-DOS orPC-DOS (version 2. 0 or higher) directly. An example analysis is given to illustrate theinput and output features of program LIFTAB. The examples included in the programare female breast cancer survival data during 1980--84, and colon cancer survival dataduring 1972--76 and 1980--84 collected by Shanghai Cancer Registry.
出处
《肿瘤》
CAS
CSCD
北大核心
1991年第3期106-109,共4页
Tumor
