| Author |
 Topic  |
|
|
IrinaMosh
Germany
Posts |
 Posted - 10/20/2005 : 09:13:42 AM
|
Origin Version (Select Help-->About Origin): 7,5
Operating System:Win
Hi,
I would like to ask someone who ever worked with Survival statistics, how exactly to fill in tables for Kaplan-Meyers analysis. I have three groups of animals and I need three graphs. I have actually only survival time and number of animals , so two colomn tables for each group. Zensor I guess is a number of animals. What I found in tutorial didn#t work in my case.
Would be grateful for your kind advice.
Ina |
|
|
Yerika
USA
Posts |
Posted - 10/20/2005 : 3:26:12 PM
|
It's my understanding that Kaplan-Meyer graphs have time on the x-axis and percentage of the surviving population on the y axis.
If you have a number of animals, figure out the percentage of survivors in a new column by going to the Column menu, and choosing "Set Column Values." In the dialog box, type in the column name where you are keeping number of animals, divided by the maximum number of animals, for example: col(b)/max (assuming your y data is in column b)
Then highlight the percentage column and choose the type of graph you want from the Plot menu. You can repeat this process for each of the sets of data you have, and create separate graphs, or graph each set of data in one graph using the layer tool.
If you need more information or I have misunderstood your question, please let me know.
|
 |
|
|
azhu
China
4 Posts |
Posted - 10/20/2005 : 11:07:58 PM
|
Hi, IrinaMosh,
In Kaphan-Meier, an observation having a censored time value represents some outcome other than failure such as prematurely leaving the study due to chance factors or surviving beyond the completion of the study.
So looks like that your data should be analyzed by Life Table Ananlysis.
Usually life table data are as following. Please note that usually time interval in life table analysis are the same.
time censor_num dead_num
0 10 1
1 15 1
2 6 2
3 8 1
4 12 6
5 7 8
quote:
Origin Version (Select Help-->About Origin): 7,5
Operating System:Win
Hi,
I would like to ask someone who ever worked with Survival statistics, how exactly to fill in tables for Kaplan-Meyers analysis. I have three groups of animals and I need three graphs. I have actually only survival time and number of animals , so two colomn tables for each group. Zensor I guess is a number of animals. What I found in tutorial didn#t work in my case.
Would be grateful for your kind advice.
Ina
Edited by - azhu on 10/20/2005 11:12:26 PM |
|
|
|
IrinaMosh
Germany
Posts |
Posted - 10/21/2005 : 05:47:07 AM
|
[quote]
Hi, Azhu
I guess your advise is up to point.
Even when it sounds stupid for you, what is this magic censor number? I really have no idea. How to calculate it or what does it mean.
time censor_num dead_num
0 10 1
1 15 1
2 6 2
3 8 1
4 12 6
5 7 8
Thank you in advance
Ina
|
|
|
|
IrinaMosh
Germany
Posts |
Posted - 10/21/2005 : 05:51:24 AM
|
Hi Yerika,
Thank you for your advice. The point only is that I would like to skip calulations of percentage. I guess that with Origin is possible to avoid this. But I don't know how. This is my trouble.
Ina
|
|
|
|
azhu
China
4 Posts |
Posted - 10/23/2005 : 10:25:48 PM
|
Hi, Ina,
Would you like to show me some of your data so that I can have more idea about it.
|
|
|
|
IrinaMosh
Germany
Posts |
Posted - 10/25/2005 : 04:16:21 AM
|
Hi Azhu,
E.g. I had two groups of animals:
1 group 8 survived after OP, I applied medicine then:
Hours Number of dead
14 1
17 2
23 1
36 1
60 1
2 group 13 survived and then with drug
Hours Number of dead
14 2
17 1
23 1
28 3
36 1
60 1
I need to compare groups. I guess time intervals I should unify, but what is zensor here?
Grateful in advance
Ina |
|
|
|
azhu
China
4 Posts |
Posted - 10/25/2005 : 06:19:19 AM
|
Hi, Ina,
For example, if we examine a group of living subjects, we know that each one is alive today, but we do not know their future date of death. we call them right censored values.
I assume that you stopped examining after 60 hours. Then in Group1, 2 is still alive when you stopped examining, we can call the 2 values as censored values.
To do Kaphan-Meier analysis, you data should be written as following
Group1
0-censored, 1-dead(failure)
14 1
17 1
17 1
23 1
36 1
60 1
60 0
60 0
Group2
0-censored, 1-dead(failure)
14 1
14 1
17 1
23 1
28 1
28 1
28 1
36 1
60 1
60 0
60 0
60 0
60 0
To do comparison, we need to plot the 2 survival function in a graph
1. Do Kaphan-Meier analysis to plot survival plt respectively.
2. Select Edit: Merge All Graph Windows to merger the 2 survival plot in a graph.(Make sure that in the merger graph dialog we select 1 row and 1 column)
According to your data in survival plot, we can roughly conclude that survival status in group 2 is better.
Edited by - azhu on 10/25/2005 07:13:52 AM
Edited by - azhu on 10/25/2005 07:21:14 AM |
|
|
| |
Topic |
|
|
|
Survival statistic