计算

先构造两个多变量 Cox 分析模型

install.packages("ezcox")#先安装包
packageVersion("ezcox")#0.4.0版本
library(survival)
library(ezcox)
 lung$ph.ecog <- factor(lung$ph.ecog)
zz <- ezcox(lung, covariates = c("sex", "ph.ecog"), controls = "age", return_models = TRUE)
#=> Processing variable sex
#==> Building Surv object...
#==> Building Cox model...
#==> Done.
#=> Processing variable ph.ecog
#==> Building Surv object...
#==> Building Cox model...
#==> Done.
 zz
#$res
# A tibble: 6 x 12
#  Variable is_control contrast_level ref_level n_contrast n_ref    beta    HR lower_95 upper_95 #p.value global.pval
#  <chr>    <lgl>      <chr>          <chr>          <dbl> <dbl>   <dbl> <dbl>    <dbl>    <dbl>   <dbl>       #<dbl>
#1 sex      FALSE      sex            sex              228   228 -0.513  0.599    0.431    0.831 0.00218    #0.000857
#2 sex      TRUE       age            age              228   228  0.017  1.02     0.999    1.04  0.0646     #0.000857
#3 ph.ecog  FALSE      1              0                113    63  0.359  1.43     0.969    2.11  0.0712     #0.000551
#4 ph.ecog  FALSE      2              0                 50    63  0.857  2.36     1.5      3.7   0.0002     0.000551
#5 ph.ecog  FALSE      3              0                  1    63  2.11   8.23     1.09    61.8   0.0406     0.000551
#6 ph.ecog  TRUE       age            age              228   228  0.0108 1.01     0.992    1.03  0.251      #0.000551

#$models
# A tibble: 2 x 5
# Variable control model_file                                                                    model   status
  <chr>    <chr>   <chr>                                                                         <list>  <lgl> 
#1 sex      age     "C:\\Users\\ASUS\\AppData\\Local\\Temp\\Rtmp0GDxAP/ezcox\\ezcox_1c30cfa3a80"  <coxph> TRUE  
#2 ph.ecog  age     "C:\\Users\\ASUS\\AppData\\Local\\Temp\\Rtmp0GDxAP/ezcox\\ezcox_1c3060a35bfa" <coxph> TRUE  

#attr(,"class")
#[1] "ezcox" "list" 
#attr(,"controls")
#[1] "age"
#####获取模型
得到计算结果后，我们下一步提取模型。

mds <- get_models(zz)

查看模型结构，其实是列表而已

str(mds, max.level = 1)
#> List of 2
#>  $ Surv ~ sex + age    :List of 19
#>   ..- attr(*, "class")= chr "coxph"
#>   ..- attr(*, "Variable")= chr "sex"
#>  $ Surv ~ ph.ecog + age:List of 22
#>   ..- attr(*, "class")= chr "coxph"
#>   ..- attr(*, "Variable")= chr "ph.ecog"
#>  - attr(*, "class")= chr [1:2] "ezcox_models" "list"
#>  - attr(*, "has_control")= logi TRUE

Show time 使用 show_models()。

#先安装forestmodel包
library(remotes)
remotes::install_github("ShixiangWang/forestmodel")
library("forestmodel")
show_models(mds)
```![png](https://img.haomeiwen.com/i21455872/1018a5d57e2bc8e5.png?imageMogr2/auto-orient/strip%7CimageView2/2/w/1240)

# Set model names
show_models(mds, model_names = paste0("Model ", 1:2))

.png

对比发现改变了model名字，内容是不变的.

png

让图形更精简一些，可以将模型合并并去掉控制变量。

# Merge all models and drop control variables
show_models(mds, merge_models = TRUE, drop_controls = TRUE)

png

一步生成森林图

show_forest(lung, covariates = c("sex", "ph.ecog"), controls = "age")

png
cox森林图也可用ggforest()画1，两者对比，看你自己喜欢哪个。

另外，非常感谢王诗翔老师的邮件回复，帮助我解决问题，让我能跟着文章做下去！
参考：王诗翔「r<-包」使用 ezcox 展示森林图