[R] Calculating survival for set time intervals

[Thomas Lumley]

On Sat, 10 Jun 2006, Gregory Pierce wrote:

> Hello friends and fellow R users,
>
> I have successfully tabulated and entered my survival data into R and
> have generated survival curves. But I would like to be able to determine
> what the survival rates are now at one month, three months, six months
> and one year.
>
> I have a data set, via.wall, which I have entered into R, and which
> generates the following Surv object:
>
> Surv(Days,Status==1)
>  [1] 648+   3  109  241   99    7  849+ 105    3+ 539+ 121   42  490
> 21  870+
> [16] 175   20  434  289  826+ 831+ 664  698+   5   24  655+  18    7+
> 85+  65+
> [31] 547+   8    1   55+  69  499+ 448+   0  158+  31  246+ 230+  19
> 118+  54
> [46]  48+  45+  21+ 670+ 585  558+ 544+ 494  481+ 474+ 472+ 461  447
> 446+ 443+
> [61] 429+ 423+ 401  395+ 390  390+ 389+ 383+ 383+ 373+ 362+ 354  344+
> 342  336+
> [76] 335+ 326+ 306  300+ 292  284+ 280+ 271  246+ 237+ 234  233+ 233
> 230+ 230+
> [91] 226+ 225+ 218+ 215  211+ 199+ 191+ 191  190+ 184+ 169+ 163+ 161+
> 153  150
> [106] 129+ 110+ 107+ 100+  84+  77+  69+  52+  38+  11+
>> names(wall.via)
> [1] "Description" "Patient"     "Physician"   "MRN"         "Age"
> [6] "Status"      "Days"        "Cr"          "INR"         "BR"
> [11] "MELD"        "type"
>
> I can guess pretty accurately by looking at the graph what the survival
> rates are at each interval, but I would like to understand how to
> instruct R to calculate it. Hope I have made this clear. I am just a
> beginner, so forgive me if this is trivial. It just isn't clear to me.

You will have generated survival curves with survfit().  Give them to 
summary() and specify the times you want. For example, using one of the 
built-in data sets:

> fit <- survfit(Surv(time, status) ~ x, data = aml)
> summary(fit,times=c(10,20,30,40))
Call: survfit(formula = Surv(time, status) ~ x, data = aml)

                 x=Maintained
  time n.risk n.event survival std.err lower 95% CI upper 95% CI
    10     10       1    0.909  0.0867        0.754        1.000
    20      7       2    0.716  0.1397        0.488        1.000
    30      5       1    0.614  0.1526        0.377        0.999
    40      3       2    0.368  0.1627        0.155        0.875

                 x=Nonmaintained
  time n.risk n.event survival std.err lower 95% CI upper 95% CI
    10      8       4    0.667   0.136       0.4468        0.995
    20      6       1    0.583   0.142       0.3616        0.941
    30      4       3    0.292   0.139       0.1148        0.741
    40      2       1    0.194   0.122       0.0569        0.664


 	-thomas


Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle