Variable length time to event studies are based on observing predetermined total numbers of events. Many of these trials involve following patients over a long span of time. Research teams need to determine the conditions under which studies can be finalized in order to put together resources necessary to the dissemination of study results. We show prediction strategies for determining the timing of the final events in blinded settings. Our strategies estimate subject accrual from staggered entry and the evolution of hazard rates. We model hazard rates based on the time in study vs. time since the beginning of the trial. Future accruals and times in study are estimated by fitting mixtures of B splines to Poisson intensity rates.
We propose a Basic and an Advanced strategy for determining the timing of events in blinded settings. The Basic strategy utilizes separate processes for accrual and calendar event times. The Basic strategy, although simple to implement, does not account for the evolution of hazard rates and does not provide model flexibility. The Advanced strategy, addresses these limitations and utilizes separate processes for accrual and time in study.