Inference for Events with Dependent Risks in Multiple Endpoint Studies

摘要: Abstract Kaplan–Meier and cumulative incidence functions are not sufficient descriptive devices for studies that have multiple time-to-event endpoints. For example, in cancer treatment research the probability of tumor recurrence conditional on having died from treatment-related toxicities prevalence graft-versus-host disease among leukemia-free patients surviving a bone marrow transplant interest. These quantities can be estimated nonparametrically using simple several estimates events with possibly dependent risks. We derive asymptotic distribution theory such by representing Kaplan–Meier, incidence, hazard estimators as sums iid random variables. Variance estimation also follows directly this representation. Two-sample test statistics null presented. Several examples illustrate utility these results.