Abstract:
|
Interval censoring means that an event time is only known to lie in an interval (L,R], with L the last examination time before the event, and R the first after. In the univariate case, parametric models are easily fitted, and for non-parametric models, the mass is placed on some intervals, derived from the L and R points. Asymptotic results are simple for the former and complicated for the latter. This paper reviews extension to multi-state models, in practice exemplified by two cases, the illness-death model (where the time of illness occurrence is interval censored and the death time is known precisely) and the case known as double censoring (referring to analysis of the time between two events, which are both subject to censoring, like an incubation time, where the time of becoming infected is interval censored). There are severe identifiability issues in case we would use models like those typically applied for data subject to right censoring. Therefore, it may be preferable to consider more restrictive models. Potential models will be discussed.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.