Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: RE: Relative survival - strel


From   Tim Evans <Tim.Evans@wmciu.nhs.uk>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   RE: st: RE: Relative survival - strel
Date   Fri, 7 Oct 2011 13:26:03 +0100

Sam, Steve

Yes I suppose you could read a survival analysis text to get a better understanding but I still think that in order to apply what you have read, you would need to look at the calculation in the ado file in order to know exactly what it is doing. 

The method that -strel- and other similar programmes use for calculating crude or observed survival is the actuarial (life table) method. You don't necessarily need to read a specific Stata book as this is implemented in other stats packages and the methods behind it are well know. You should be able to get the same crude survival using -ltable- in Stata, with your code it would be this:

ltable _t death_5y, interval(5)

A little back ground to it can be easily found by google search:

'In the life-table method, times-to-event (failures and censorings or withdrawals without failure) are grouped into convenient intervals. The ratio of cases that failed to the number at risk at the beginning of the interval is computed for each interval, and, from this, the cumulative proportion of survivals and failures.'

Specifically:

'An important assumption is that cases censored within an interval are at risk of failing for half the interval or, as implemented, that half the censored cases are at risk in that interval.'

If you want a good introductory book, then Cancer Registration Principles and Methods (1991) produced by IARC is a very good starting place. With reference to survival analysis, this pdf chaper (http://www.iarc.fr/en/publications/pdfs-online/epi/sp95/sp95-chap12.pdf) specifically deals with it. Page 164 onwards deals with the lifetable method.

Hope this helps.

Tim


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Steven Samuels
Sent: 06 October 2011 22:40
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: RE: Relative survival - strel


Sam, 
I suggest that you read a survival analysis text or look at one of the many tutorials to be found on the net, e.g. Stephen Jenkins's "Survival Analysis Using Stata:" http://www.iser.essex.ac.uk/survival-analysis

Steve

On Oct 6, 2011, at 6:49 AM, Tim Evans wrote:

Hi Sam,

strel is a programme for calculating relative survival, updated and maintained by the London School of Hygiene and Tropical Medicine (for Nicks information!).

I think you have to look at how strel is calculating crude survival by looking in the ado file. 

I think you could say there are two types of crude survival  - one where you simply divide number of events by the number of cases (which takes no account of survival time) - so 'Crude' crude survival, or you can have a sophisticated crude survival which takes account of the survival time or person years in the analysis and gives a weighted crude survival. This is what strel does - if you look at the code in the ado file, you will see this:

gen Crude=exp(-sum(deaths/p_years*_width))*100

This is why you don't get an exact division in the manner you suggest. An exact division would be if the survival time of all of the death cases was 2.5 years.

I think its upto you to decide which is more useful for you, but the above is common practice and if you used strs (an alternative programme in Stata) it also calculates crude survival like this.

As an aside, I don't know what data you are calculating relative survival on, but your break points seem too wide to take into account variations in underlying mortality. By having only one break point of 5 years, you are assuming that there is no change in risk of mortality between 0-5 years. Without knowing your data, I doubt this is a correct assumption.

Hope this helps

Tim


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Sam Leary
Sent: 06 October 2011 11:07
To: statalist@hsphsun2.harvard.edu
Subject: st: Relative survival - strel

Dear all,  I am a bit of a novice at survival analysis, but am trying to 
calculate relative survival using strel.  My variables are

death_5y = 1 if died by 5 years after diagnosis, 0 if still alive at 5 years
age_diag = age at diagnosis (years)
age_death = age at death or 5 years, whichever came first (years)

I have downloaded a lifetable which has the following variables:
country calendar_year sex age rate
and I have only kept the data for the calendar year relating to the start 
date of my audit.  I have also sorted this file by age then sex.

So I have used the following commands to calculate the crude and relative 
survival at 5 years after diagnosis:

stset age_death, failure(death_5y) origin(age_diag)
strel 0 5 using "ltab_s1",mergeby(sex)

My question is, as there are 257 deaths out of a total of 468, why is the 
crude survival percentage 41.86 using strel, rather than (468-257)/468 = 
45.1?

Many thanks,

Sam.

----------------------
Dr Sam Leary
Senior Lecturer in Statistics

Working hours:
Monday, Tuesday and Thursday
7.30am - 2.30pm

Lifecourse Epidemiology and Population Oral Health Research Group
Department of Oral and Dental Sciences
Bristol Dental School
Lower Maudlin Street
Bristol
BS1 2LY
Tel 0117 342 3264
Fax 0117 342 4443
E-Mail: S.D.Leary@bristol.ac.uk
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

_DISCLAIMER:
This email and any attachments hereto contains proprietary information, some or all of which may be confidential or legally privileged. It is for the exclusive use of the intended recipient(s) only. If an addressing or transmission error has misdirected this e-mail and you are not the intended recipient(s), please notify the author by replying to this e-mail. If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail or any attachments, as this may be unlawful.


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

_DISCLAIMER:
This email and any attachments hereto contains proprietary information, some or all of which may be confidential or legally privileged. It is for the exclusive use of the intended recipient(s) only. If an addressing or transmission error has misdirected this e-mail and you are not the intended recipient(s), please notify the author by replying to this e-mail. If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail or any attachments, as this may be unlawful.


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index