Re: multi-event survival analysis?

Posted by Dale Glaser on Sep 27, 2011; 3:48am
URL: http://spssx-discussion.165.s1.nabble.com/multi-event-survival-analysis-tp4827134p4843964.html

Bruce, I once looked into when one has repeated events in the context of a time-to-event analysis/design, and such recurrent events analysis can be analyzed (which I haven't used) via a R package called "Frailtypack", which has the following description from the R website:   "Frailtypack now fits several classes of frailty models using Penalized Likelihood on the hazard function. 1) A shared gamma frailty model and Cox proportional hazards model. Left truncated, censored data and strata (max=2) are allowed. Clustered and recurrent survival times can be studied (the Andersen-Gill(1982) approach has been implemented for recurrent events). An automatic choice of the smoothing parameter is possible using an approximated cross-validation procedure. 2) Additive frailty models for proportional hazards models with two correlated random effects (intercept random effect with random slope). 3) Nested frailty models for hierarchically clustered data (with 2 levels of clustering) by including two iid gamma random effects. 4) Joint frailty models in the context of joint modelling of recurrent events with terminal event."   .......but I haven't used the GENLINMIXED option in SPSS, so maybe it is equipped to do recurrent events analysis.......I think (but I would have to double-check) one may be able to do such analysis in Stata, and I would defer to SAS users on this listserve to see if the same can be done for that software..but as I said, there is definitely a R Package     Dale Glaser, Ph.D. Principal--Glaser Consulting Lecturer/Adjunct Faculty--SDSU/USD/Alliant 3115 4th Avenue San Diego, CA 92103 phone: [phone] fax: [phone] email: [hidden email] website: www.glaserconsult.com --- On Mon, 9/26/11, Bruce Weaver <[hidden email]> wrote: From: Bruce Weaver <[hidden email]> Subject: Re: multi-event survival analysis? To: [hidden email] Date: Monday, September 26, 2011, 12:53 PM Please clarify what you mean by multiple-event.  I.e., do you mean the same event can occur multiple times per person (e.g., time to drink for pts in rehab); or do you mean that there is only one event per person, but an event is defined as the occurrence of any one of X, Y or Z?  I *think* Rich U was thinking about the second case when he said Cox regression could do this. For repeated occurrences of the same event, I *think* you'd need to use GENLIN or GENLINMIXED, or something like that.  But there may be others who have actually done it who can speak with more authority. J P-6 wrote: > > Hello, > Â > Does anyone know if SPSS can handle a multi-event survival analysis? It > appears not, but thought I should consult the wisdom of the list. > Â > Thanks, > John > ----- -- Bruce Weaver bweaver@... http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." NOTE: My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. -- View this message in context: http://spssx-discussion.1045642.n5.nabble.com/multi-event-survival-analysis-tp4827134p4842687.html Sent from the SPSSX Discussion mailing list archive at Nabble.com. ===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@... (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD