Collections > Electronic Theses and Dissertations > Analysis of Complex Time-to-Event Data

The number needed to treat is a tool often used in clinical settings to illustrate the effect of a treatment. It has been widely adopted in the communication of risks to both clinicians and non-clinicians. We introduced a definition of the number needed to treat for time to event data with competing risks using the cumulative incidence function and suggest non-parametric and semi-parametric inferential methods for right censored time to event data in presence of competing risks. In HIV-1 clinical trials the interest is often to compare how well treatments suppress the HIV-1 RNA viral load. We propose an endpoint based on the probability of the viral load being suppressed, and suggest that treatment differences be summarized using the mean restricted time a patient spends in the state of viral suppression. In the standard analysis of competing risks data, proportional hazards models for cause-specific hazards are fit using the same time scale for all causes of failure. We propose estimating cumulative incidence function by fitting regression models for the cause-specific hazard functions using different time scales for each cause. We establish consistency and asymptotic normality of the proposed estimator and assess its performance in simulations. The method is illustrated with stage III colon cancer data obtained from the Surveillance, Epidemiology, and End Results (SEER) program of National Cancer Institute. In competing risks setup, the data for each subject consists of an observed event time, censoring indicator, and the cause of failure, given that the failure is observed. When it is not possible to obtain information about the cause of failure, we suggest a non-parametric method in which the probabilities of failure from each specific cause are first estimated using local polynomial regression, and then these estimates are used to estimate the cause-specific cumulative hazards and cumulative incidence functions. The method is illustrated using the data on infections in patients from the United States Cystic Fibrosis Foundation Patient Registry.