A Marginal Semi-parametric Method to Analyse Recurrent Event Data in the Presence of Terminal Events
by Dr James Cui
Abstract: Recurrent event data are frequently encountered in clinical and epidemiological studies. In some cases, these recurrent events can be precluded by the occurrence of a terminal event, such as death. The terminal event is usually not independent of the recurrent event and thus, appropriate analytical methods taking account of this dependent relationship could improve data analyses. The aim of this talk is to describe the recent development in the analysis of recurrent event data when a terminal event is present, and introduce a marginal semi-parametric model that can be implemented in Stata software. We applied this method to data from the Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) study and found that the protective effect of pravastatin was significant in prevention of both a first myocardial infarction (MI) event and subsequent MI events, death from coronary heart disease (CHD) and non-CHD death after adjusting for age, gender and smoking status. The proposed method combines the analytic methods for ordered and unordered multiple failure time data into one framework and utilizes more information relating to the biological process of a disease.
For More Information: Dr Owen Jones: O.D.Jones@ms.unimelb.edu.au