Estimating the extent of tracking in interval-censored chain-of-events data

This paper describes a method for determining whether the times between a c hain of successive events (which all individuals experience in the same ord er) are correlated, for data in which the exact event times are not observe d. Such data arise when individuals are only observed occasionally to deter mine which events have occurred. In such data, the (unknown) event times ar e interval censored. In addition, some individuals may have experienced som e of the events before their first observation and may be lost to follow-up before experiencing the last event. Using a frailty model proposed by Aale n (1988, Mathematical Scientist 13, 90-103) but which has never been used t o analyze real data, we examine whether individuals who develop early marke rs of HIV infection can also be expected to develop antibody and other indi cators of HIV infection more rapidly.