Re: st: Recurrent Events and Terminal Event

[Steven Samuels <sjsamuels@gmail.com>]

Your question is unclear.  Do you wish to study the distribution and predictors of the recurrent events (the correct technical term), predictors of the terminal event (including possibly the recurrent events), or both?

For the former, see Section 3.2 http://www.stata.com/support/faqs/stat/stmfail.html#order,  especially 3.2.3 and 3.2.4. You can add information on the nature of prior events in the form of possibly time-dependent covariates. See also the section on recurrent events in  Kalbfleisch and  Prentice, Analysis of Failure Time Data, 2nd Edition, Wiley, 2002. 

I don't think that the possibility of a terminal event should much affect the analysis of recurrent events, but I could be wrong. If you have a reference with an analysis you'd like to emulate in Stata, please share it.  For the specific topic of heart attacks: if  death "closely" follows  a heart attack, I suggest that you end observation with last attack and not restart the clock.

If your interest is to "adjust" for the recurrent events in studying  causes of the terminal event, you have a serious conceptual problem: the recurrent events are on the causal pathway. 

Steve



Steve

On Aug 21, 2011, at 4:58 PM, Iulia Cioroianu wrote:

Hello,

Does anyone know of any attempts to deal with repeated events in the
presence of a related terminal event in Stata? For example, recurring
heart attacks that are associated with a higher risk of death.
Any ideas or examples are appreciated.
Thank you,

Iulia
*
*   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/