For patients with blood diseases, long-term remission is often only possible after haematopoietic stem cell transplantation (HSC). Although transplantation can be life-saving, a number of post-transplant events can increase mortality. This thesis provides the first insight into patient outcomes after HSC transplantation using cord blood donated to the UK National Health Service Cord Blood Bank (NHS-CBB). In the NHS-CBB dataset, event times were incompletely observed for some event types. Missing event times were assumed to have occurred in a known, finite, time-period. Hence, the missing event times were considered interval-censored. Methods for handling interval-censored event times can be categorised as (i) applying multiple imputation (MI) strategies or (ii) taking a full maximum likelihood (FML) approach. I focused on MI strategies, rather than FML methods, because of their flexibility. Using simulation studies, I evaluated MI strategies in competing risks and multi-state model (MSM) analyses, examining the extent to which interval boundaries, the data distribution, and analysis model should be accounted for when data were missing at random (MAR) and missing not at random (MNAR). I found that MI by predictive mean matching (PMM), in which sampling is from a set of observed times without reliance on a specific parametric distribution, resulted in least biased estimates when event times were MAR (conditional on observed data). Furthermore, in MSM analysis, I found that applying PMM separately for each sub-group of patients with a different pathway through the MSM tended to reduce bias and standard error. Finally, I applied the best MI methodology from my simulation studies in an analysis of the NHS-CBB dataset. My results suggest that application of MI methods to the NHS-CBB dataset reduced bias and improved precision in estimates, compared with complete case analysis. My approach ensures that accurate information is available to inform decisions for both clinicians and patients.