如何将参数生存曲线叠加到Kaplan-Meier生存曲线上？

已完成的分析代码 #

Kaplan-Meier分析 #

KMsufit1_2 <- survfit(survobj1_2 ~ 1, data = set_1_2_grp)

参数生存分析 #

Gamma1_2 <- flexsurvreg(survobj1_2 ~ 1, data = set_1_2_grp, dist = 'gamma')
gen_Gamma1_2 <- flexsurvreg(survobj1_2 ~ 1, data = set_1_2_grp, dist = 'gengamma')

实现方案 #

以下提供两种常用的实现方式，可根据需求选择：

方式1：基于ggplot2 + survminer的灵活绘图 #

适合需要自定义图表样式的场景，步骤如下：

1. 加载依赖包

library(survminer)
library(flexsurv)
library(ggplot2)

2. 绘制基础Kaplan-Meier曲线并提取ggplot对象

km_plot <- ggsurvplot(KMsufit1_2, data = set_1_2_grp, 
                      surv.median.line = "hv", # 可选：添加中位生存时间参考线
                      ggtheme = theme_bw())$plot

3. 提取参数模型的生存曲线数据

# Gamma分布模型的生存数据
gamma_surv_data <- summary(Gamma1_2, type = "survival", tidy = TRUE)
# 广义Gamma分布模型的生存数据
gengamma_surv_data <- summary(gen_Gamma1_2, type = "survival", tidy = TRUE)

4. 叠加参数曲线并美化图表

km_plot +
  geom_line(data = gamma_surv_data, aes(x = time, y = est), color = "red", linetype = "dashed") +
  geom_line(data = gengamma_surv_data, aes(x = time, y = est), color = "blue", linetype = "dotted") +
  labs(title = "Kaplan-Meier vs 参数生存曲线",
       x = "时间", y = "生存概率") +
  scale_color_manual(name = "模型类型",
                     values = c("black", "red", "blue"),
                     labels = c("Kaplan-Meier", "Gamma分布", "广义Gamma分布")) +
  theme(legend.position = "bottom")

方式2：基于基础绘图系统的快速叠加 #

代码简洁，适合快速生成对比图：

library(flexsurv)

# 绘制Kaplan-Meier曲线
plot(KMsufit1_2, main = "Kaplan-Meier与参数生存曲线对比",
     xlab = "时间", ylab = "生存概率", col = "black")

# 叠加Gamma分布参数曲线
lines(Gamma1_2, type = "survival", col = "red", lty = 2)

# 叠加广义Gamma分布参数曲线
lines(gen_Gamma1_2, type = "survival", col = "blue", lty = 3)

# 添加图例区分不同曲线
legend("bottomleft", 
       legend = c("Kaplan-Meier", "Gamma分布", "广义Gamma分布"),
       col = c("black", "red", "blue"), 
       lty = c(1, 2, 3))

内容的提问来源于stack exchange，提问作者Hari