Home | Research| Experience | Activities

Research interests

Survival and longitudinal data analysis
Recurrent events; Gap times; Informative censoring; Heterogeneity
Methodologies
Nonparametric; Semiparametrc; likelihood; Estimating equation; Frailty

Thesis work

Statistical modeling univariate and alternating bivariate gap time data
Recurrent events arise very frequently in medical follow-up studies, such as recurrent opportunistic infections, and alternating hospital stay and community stay. The gap time perspective, which focus on the time between the two successive events is prefered if the events are chronologically ordered or the cyclying pattern is clear. Analyzing gap times of recurrent events helps us to investigate the individual-specific as well as episode-specific covariate effect, and the trend of the durations of gap times. The analyis is challenged by correlation between gap times and the induced dependent censoring (Gelber et al., 1989). We want to develop marginal or conditional regression model for gap times under proper assumption to approach these problems.

Inference and application research

Application to infectious disease, epidemiology and mental health
Current work includes trend analysis of extramedical analgesic use and dependence over the study period, statistical testing for the potential causal pathway between psychiatric disorders and drug use disorders, evaluation of the potential factors associated with adverse effects of adult male circumcision in HIV-infected and uninfected men, and assessing the treatment effect on disease progression for HIV-infected subjects.