How to convert cox regression output into a dataframe?

Question

I ran univariate cox regression model using following code.

res <- coxph(Surv(DFS, DFS.Event) ~ Grade, data = Grade)
summary(res)

It output as follows:

Call:
coxph(formula = Surv(DFS, DFS.Event) ~ Grade, data = Grade)

  n= 172, number of events= 38 
   (12 observations deleted due to missingness)

             coef exp(coef) se(coef)     z Pr(>|z|)
GradeGrade2 1.153     3.167    1.026 1.123    0.261
GradeGrade3 1.006     2.736    1.028 0.979    0.327

            exp(coef) exp(-coef) lower .95 upper .95
GradeGrade2     3.167     0.3158    0.4237     23.67
GradeGrade3     2.736     0.3655    0.3651     20.50

Concordance= 0.539  (se = 0.04 )
Likelihood ratio test= 1.82  on 2 df,   p=0.4
Wald test            = 1.34  on 2 df,   p=0.5
Score (logrank) test = 1.45  on 2 df,   p=0.5

I want to convert the output into a dataframe. My expected output is:

Variable    coef    exp(coef)   se(coef)    z   p   lower 0.95  upper 0.95
GradeGrade2 1.153   3.167   1.026   1.123   0.261   0.4237  23.67
GradeGrade3 1.006   2.736   1.028   0.979   0.327   0.3651  20.5

That is not the print output of coxph. It’s the output of its associated summary method. — IRTFM
– IRTFM, Commented Oct 11, 2023 at 18:11

Friede · Accepted Answer · 2023-10-11 18:22:20Z

As you are looking for a method to extract information from output of summary(coxph(something)), I recommend a self-written function of the form

CoxphToDF <- function(y) {
  stopifnot(class(y) == "summary.coxph")
  
  y <- 
    cbind(y[["coefficients"]], 
          `lower .95` = y[["conf.int"]][, "lower .95"], 
          `upper .95` = y[["conf.int"]][, "upper .95"]
    )
  
  cbind(Variable = rownames(y), as.data.frame(y))  
}

Toy data from ?coxph:

library(survival)

test2 <- list(start=c(1,2,5,2,1,7,3,4,8,8), 
              stop=c(2,3,6,7,8,9,9,9,14,17), 
              event=c(1,1,1,1,1,1,1,0,0,0), 
              x=c(1,0,0,1,0,1,1,1,0,0)) 
bladder1 <- bladder[bladder$enum < 5, ] 

testobject <- summary(coxph(Surv(stop, event) ~ (rx + size + number) * strata(enum),
                            cluster = id, bladder1))

The result looks odd as the row names are identical to the Variable column, but this mimics the expectations.

CoxphToDF(testobject) |> head()
#>                                    Variable        coef exp(coef)   se(coef)
#> rx                                       rx -0.52598436 0.5909733 0.31582587
#> size                                   size  0.06961334 1.0720936 0.10155903
#> number                               number  0.23817961 1.2689371 0.07588471
#> rx:strata(enum)enum=2 rx:strata(enum)enum=2 -0.10632654 0.8991310 0.50423983
#> rx:strata(enum)enum=3 rx:strata(enum)enum=3 -0.17250719 0.8415522 0.55780009
#> rx:strata(enum)enum=4 rx:strata(enum)enum=4 -0.10945460 0.8963229 0.65730033
#>                        robust se          z    Pr(>|z|) lower .95 upper .95
#> rx                    0.31523946 -1.6685232 0.095211902 0.3185925  1.096226
#> size                  0.08863123  0.7854268 0.432203369 0.9011346  1.275486
#> number                0.07458532  3.1933847 0.001406155 1.0963594  1.468680
#> rx:strata(enum)enum=2 0.33395783 -0.3183831 0.750194332 0.4672589  1.730168
#> rx:strata(enum)enum=3 0.39867743 -0.4326987 0.665233705 0.3852354  1.838383
#> rx:strata(enum)enum=4 0.50636317 -0.2161583 0.828864362 0.3322365  2.418141

^{Created on 2023-10-11 with reprex v2.0.2}

lower 0.95 and upper 0.95 on both the lines are same. I am supposed to get 0.4237 and 23.67 in line one and 0.3651 20.50 in line two. But I am getting same values at both the lines

Collectives™ on Stack Overflow

How to convert cox regression output into a dataframe?

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related