Improving the reporting of death and investigating the role of probiotics in mucositis and infections in children with cancer

This thesis examines the classification of death, and the use of probiotics to reduce and prevent infection-related mortality in children diagnosed with cancers. It comprises of three main parts. The first part describes a study that validated a consensus-based definition of treatment-related mortality (TRM) and cause-of-death attribution system. This took place in a single institution in Leeds, the UK outside the centre it was initially developed (Toronto, Canada). Two consultants and two clinical research associates independently classified deaths as TRM or “not treatment-related” according to an algorithm. When TRM occurred, reviewers applied the cause-of-death attribution system, and inter-relater reliability was then assessed. This study demonstrated that the classification and cause of death attribution systems can be implemented in different health care settings, but that further research is required for patients receiving palliative care. The second part of this thesis describes a systematic review and meta-analysis that investigated the efficacy and safety of probiotics in people with cancer. Probiotics appear safe to deliver and may reduce the incidence of diarrhoea and duration of fever, but, heterogeneity, unclear bias, and a lack of paediatric participants demonstrated uncertainty in these findings. Findings from this systematic review were used to develop a randomised-controlled feasibility study. The third part reports the first study undertaken to evaluate the feasibility of undertaking an randomised-controlled trial (RCT) to investigate the use of probiotics (Symprove) to prevent or reduce mucositis and infection in children with cancers in the UK. Evaluation suggested that a RCT is feasible, but further considerations are needed to address significant barriers to recruitment and adherence to the capture of data that were identified. Findings from this study have been used to develop a parallel biological sub-study that can be undertaken in a future RCT.