risktable stsgraph

[laure rouch]

Dear stata users,

I am writing to you about the calculation of numbers at risk post sts graph. I am studying the time (suivi) until the event "conversion".

My data are as follows. When I run the command sts graph, risktable.

The numbers at risks are :

time 0 : 375 patients
time 2 : 327 patients
time 4 : 287 patients
time 6 : 110 patients.

I do not understand because during the follow-up, 105 patients have the event "conversion". I thought the numbers at risk represented at each time, the number of patients who don't have done the event so far. Normally at 6 years follow up (the max), 105 patients did the event so 270 did not and are still at risk isnt'it?
Could you explain me how this number at risk is calculated and what it means exactly. I don't understand how 110 patients can be "at risk" at 6 years follow-up? It doesn't make sense.

Thank you in advance for your help,

LR

* Example generated by -dataex-. To install: ssc install dataex
clear
input int id float suivi byte conversion
  1   2 1
  2   5 0
  3   5 0
  4   6 0
  5   6 0
  6   5 0
  7   6 0
  8   1 1
  9   1 1
 10   5 0
 11   6 0
 12   1 1
 13   4 1
 14   5 0
 16   5 0
 17   6 1
 18   5 0
 20   2 1
 21   6 0
 22   5 0
 23   3 0
 24   5 0
 26   5 0
 27   5 0
 28   4 1
 29   6 0
 30   1 1
 31   2 1
 32   5 0
 33   6 0
 34   5 0
 35   2 0
 36   1 1
 38   5 0
 39   5 0
 40   5 0
 41   5 0
 42   5 0
 43   5 0
 44   5 0
 45   5 0
 46   5 0
 47   6 0
 48   3 1
 49   6 0
 51   1 1
 52   1 1
 53   5 0
 54   6 0
 55   6 0
 56   2 1
 57   5 0
 58   6 0
 59   5 0
 60   5 0
 61   5 0
 62   1 1
 63   5 1
 64   2 1
 65 1.5 1
 66   4 1
 67   5 0
 69   6 0
 70   1 1
 71   5 0
 72   3 0
 73   2 0
 74   1 1
 75   5 0
 76   1 1
 77   6 0
 78   5 0
 79   3 1
 80   5 0
 81   6 0
 82   2 1
 83   6 0
 84   6 0
 85   3 1
 86   5 0
 87   1 1
 88   6 0
 89 1.5 1
 90   5 0
 91   6 0
 92   5 0
 93   6 0
 94   1 1
 95   5 0
 96   5 0
 97   5 0
 98   1 1
 99   3 0
100   6 0
101   5 0
102   5 0
103   4 1
104   5 0
105   3 1
106   6 0
end

[Bjarte Aagnes]

Not all of your observations have the event of interest (conversion == 1), Those having conversion == 0 are censored and contribute to the total number at risk up to the time that they ceased to be followed (suivi):

sts list

shows the number lost in the fourth column (Net Lost).

[laure rouch]

Thanks Bjarte Aagnes. Obviously...Beginner's question...sorry.

Laure