The Survival Analysis Of Veterans Administration Lung Cancer Dataset

AIMThis study will describe the notion about survival analysis and the variability between diverse methods and test statistics used in analyzing survival data. METHODSThis study will be using the Veterans Administrative Lung Cancer dataset(R Dataset) of 137 patients having advanced inoperable Lung cancer, carried out by the US Veterans Administration. Out of 137 male patients, 69 patients received the Standard test while 68 patients received Chemotherapy test and 64 patients died in each treatment group while only 9 patients were alive (censored) at the end of the study. This dataset will be analyzed using the SPSS version to determine whether there is a statistically remarkable difference between the survival distributions of the cell types of Lung cancer using Life tables, and also between the Standard test and Chemotherapy test treatment groups using Kaplan Meier method. Various covariates were also analyzed to find out if they have any effect on the death of the male patients with advanced inoperable lung cancer using Cox-Regression. RESULTSThe Wilcoxon (Gehan) statistic from Life table method shows a p-value of 0.000 less than the 95% confidence level indicating that there is a statistical significance between the Lung cancer cell types. The three statistical tests of Kaplan-Meier method which include Log Rank (Mantel-Cox) test, Breslow (Generalized Wilcoxon) and Tarone Ware tests showed that the survival distribution between the two Lung cancer treatment groups are equal using an alpha level of 0.05.The Standard test and the Chemotherapy test drug treatment show the same effect on the survival time of the patients, no treatment is more beneficial than the other. Cox Regression result showed that Squamous cell type2(4), Adeno-carcinoma cell type (1), Small cell type (3) and karnofsky score are statistically significant while Large cell type2 (2), prior therapy, age, diagnosis time and the treatment group were found not significant. CONCLUSIONThe treatment difference does not show any remarkable effect on the survival time of male patients with Lung Cancer.