第十四章生存分析的SAS实现

第十四章 生存分析的 SAS 实现例 14-2 McKelvey et al(1976)收集了 3 期的某型淋巴瘤患者的生存时间（天）。分别是6，19，32，42，42，43+，94，126+，169+，207，211+，227+，253，255+，270+，310+，316+，335+，346+。SAS 分析程序data ch14_2; 定义数据集名input t c@@; 输入生存时间变量t和截尾情况变量ccards; 开始输入数据6 1 19 1 32 1 42 1 42 1 43 0 94 1 126 0 169 0 207 1 211 0 227 0 253 1 255 0 270 0 310 0 316 0 335 0 346 0;proc lifetest plots=(s); 利用lifetest过程进行生存分析并作生存函数图time t*c(0); 指定时间变量和截尾变量并指出数据截尾时截尾变量的取值run;SAS 软件输出结果Product-Limit Survival Estimates Survival Standard Number Number t Survival Failure Error Failed Left 0.000 1.0000 0 0 0 19 6.000 0.9474 0.0526 0.0512 1 18 19.000 0.8947 0.1053 0.0704 2 17 32.000 0.8421 0.1579 0.0837 3 16 42.000 . . . 4 15 42.000 0.7368 0.2632 0.1010 5 14 43.000* . . . 5 13 94.000 0.6802 0.3198 0.1080 6 12 126.000* . . . 6 11 169.000* . . . 6 10 207.000 0.6121 0.3879 0.1167 7 9 211.000* . . . 7 8 227.000* . . . 7 7 253.000 0.5247 0.4753 0.1287 8 6 255.000* . . . 8 5 270.000* . . . 8 4 310.000* . . . 8 3 316.000* . . . 8 2 335.000* . . . 8 1 346.000* . . . 8 0 NOTE: The marked survival times are censored observations. Summary Statistics for Time Variable t Quartile Estimates Point 95% Confidence Interval Percent Estimate [Lower Upper) 75 . . . 50 . 94.000 . 25 42.000 32.000 . Mean Standard Error 181.701 24.497 NOTE: The mean survival time and its standard error were underestimated because the largest observation was censored and the estimation was restricted to the largest event timeSummary of the Number of Censored and Uncensored Values Percent Total Failed Censored Censored 19 8 11 57.890.000.250.500.751.00t05010015020025...