Sentiment,
Sorry, but you make a mistake. I hope to be able to explain it with the two groups separately.
Take the group of 182 patients. These are the patients with a surgery date equal or more than 36 months. (on the date of the analysis)
You know whether each patient is dead or alive.
If the patient is alive, he is censored. Censored patients are annotated by a small vertical line on the OS curve.
The OS curve goes from 0 to 36 months. You will see no vertical lines from that group of 182 because all these patients had surgery before March 15, 2014 (3 years before date of the analysis March 2017).
44 of the first 182 patients living at least 36 months.
The scientific paper tells us that 30% from these 182 patients are living at least 30 months.
182 x 30%= 55 patients.
That means that 11 patients (55-44) died between 30 and 36 months and you will see 11 step downs on the OS curve between 30 and 36 months for that group.
Now take the second group of 41 patients (223-182). You may assume that these patients behave the same way.
That means that of these 41 patients 30% = 12 will live at least 30 months.
This group of patients had their surgery between March 15, 2014 and September 15, 2014.
Every patient who is still alive will be marked on the OS curve with a vertical line between 30 and 36 months.
If you look at the OS curve, you see 7 censors.
Conclusion: of the 12 patients who have lived for at least 30 months, 7 are still alive and 5 have died between 30 and 36 months.( 5 step downs on the curve)
Now add the 2 groups together: 182+ 41 = 223 patients.
You have 55 + 12 = 67.patients who have lived at least 30 months and more.
You have 11 + 5 = 16 patients who died between 30 and 36 months.
On the OS curve you see 16 step downs in that timeline.
You see 7 patients who are still alive between 30 and 36 months.( 7 vertical lines)
16 died + 7 still alive = 23 patients.
67-23 = 44 patients after the 36 months mark.
Conclusion: If the 7 patients of the second group (41) were all living at least 36 months then you would have 51 patients at 223 and that is 22.8%.(55/223)
That is less than the 24.2% for the first group of 182 patients (44/182 = 24.2%)
