Question: Distant metastases-free survival Time and Events/Censored data
1
gravatar for Masood
4.9 years ago by
Masood20
United Kingdom
Masood20 wrote:

I am trying to perform Kaplan-Meier analysis using histopathological information on its own and using the expression values. I am taking "mets-dd-after-diagnosis" as Time and "Alive/Death" as Events. 

 

Could you please help in suggesting that I am taking the Event variable rightly? or please suggest what should I consider for the Events?

My data table looks like this: Original GEO data set http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE40267

 

geo_accession path. review % of tumour cells ihc_pr ihc_er ihc_her2 node positive grade traditional histological type diag_tumour_size  mets_dd_after_diagnosis  death_dd_after_diagnosis  mets_count  cause_of_death  age_at_diagnosis
GSM989445 70 0 0 0  TRUE 3  NST -1  NA  NA 0  NA 46
GSM989446 95 0 0 0  TRUE 3  NST 2.5  NA  NA 0  NA 51
GSM989447 75 0 0 0  FALSE 3  NST 2 341 407 1  Breast Cancer 66
GSM989448 95 0 0 0  TRUE 3  NST 5 212 788 3  Breast Cancer 26
GSM989449 85 1 0  NA  TRUE 3  MIXED NST AND METAPLASTIC 2 1020 1208 2  Breast Cancer 50
GSM989450 70 0 0 0  FALSE 3  NST 3  NA  NA 0  NA 70
GSM989451 80 0 0 0  FALSE 3  NST 2.5  NA  NA 0  NA 33
GSM989452 90 0 0 0  FALSE 3  NST 4.5 323 542 3  Breast Cancer 67
GSM989453 80 1 0 0  NA 3  NST 8.5 194 200 4  Breast Cancer 57
GSM989454 75 0 0 0  TRUE 3  NST 5 1463 1650 3  Breast Cancer 50
GSM989455 75 0 0 0  FALSE 3  NST 3  NA 6506 0  NA 59
GSM989456 80 0 0 0  TRUE 3  NST 1.5 2163 2541 4  Breast Cancer 47
GSM989457  NA 0 0 0  NA 3  NST 2 253 271 2  other_cause 48
GSM989458 80 0 0 0  FALSE 3  NST 4.5  NA  NA 0  NA 41
GSM989459 90 1 0 0  TRUE 3  NST 3.5 384 528 3  Breast Cancer 67
GSM989460 80 0 0 0  FALSE 3  METAPLASTIC MATRIX PRODUCING 2.5  NA  NA 0  NA 39
GSM989461 75 1 0 0  FALSE 3  NST 4  NA  NA 0  NA 41
GSM989462 75 0 0 0  NA 3  NST 2 307 368 2  Breast Cancer 78
GSM989463 85 0 0 0  FALSE 3  NST 4  NA  NA 0  NA 51
GSM989464 70 0 0 0  TRUE 3  NST 2  NA  NA 0  NA 55
GSM989465 85 0 0 0  FALSE 3  NST 4.2  NA  NA 0  NA 61
GSM989466 70 0 0 0  TRUE 3  NST 5.5  NA  NA 0  NA 52
GSM989467 90 0 0 0  FALSE 3  NST 5 783 1009 3  Breast Cancer 71
GSM989468 60 0 0 0  FALSE 3  NST 2  NA  NA 0  NA 55
GSM989469 70 1  NA 0  TRUE 2  NST 5.5  NA 940 0  other_cause 74
GSM989470 50 0 0 1  TRUE 3  NST 4.5  NA  NA 0  NA 55
GSM989471 30 0 0 0  FALSE 3  NST 4.5  NA  NA 0  NA 41
GSM989472 85 0 0 0  FALSE 3  NST 5.5  NA 6763 0  other_cause 70
GSM989473 80 0 0 0  TRUE 3  NST 4.5 1579 1652 4  Breast Cancer 47
GSM989474 30 0 0 0  FALSE 3  NST 4  NA  NA 0  NA 50
GSM989475 75 0 0 0  FALSE 3  NST 2 341 407 1  Breast Cancer 66
GSM989476 70 0 0 0  NA 3  NST 5.5  NA  NA 0  NA 73
GSM989477 90 0 0 0  TRUE 2  PLEOMORPHIC LOBULAR 5 1736 1894 2  Breast Cancer 72
GSM989478 50 0 0 0  TRUE 3  NST 5 816 1378 2  Breast Cancer 49
ADD COMMENTlink modified 4.9 years ago by Sean Davis26k • written 4.9 years ago by Masood20

are you irani massod???

ADD REPLYlink modified 4.9 years ago • written 4.9 years ago by A3.7k
1

NO! I am not.

ADD REPLYlink written 4.9 years ago by Masood20

sorry then maybe Turkish!! 

ADD REPLYlink written 4.9 years ago by A3.7k
1
gravatar for Sean Davis
4.9 years ago by
Sean Davis26k
National Institutes of Health, Bethesda, MD
Sean Davis26k wrote:

You are right to be confused about what to use for survival analysis for this study.  Survival analysis needs two pieces of information to apply survival analysis for an event of interest:

  1. Time to event OR last followup
  2. Did the event occur

There are two possibilities for "event" in this study, metastasis and death.  Interestingly, this study appears (I may be misunderstanding something) to give you the time to event only when the event occurred and does not tell you the amount of time the patients that did NOT experience an event were observed.  This makes traditional survival analysis impossible.  I would consider contacting the authors about getting complete information.  Short of getting that additional information, you are going to be stuck with comparing "dead vs. alive" or "met vs. no met", neither of which is really ideal.

ADD COMMENTlink written 4.9 years ago by Sean Davis26k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1850 users visited in the last hour