Reviewing the methodology surrounding missing data in research and statistical analysis, clarifying why it can contribute to misleading results.
Missing data is present in almost all research. However, it is also a well-recognised problem in the analysis and reporting of clinical research due to its potential to introduce bias into the results. Patient-reported outcomes measures, which are increasingly used in clinical research, can be particularly susceptible to missing data. This presentation will review the methodology surrounding missing data in research and statistical analysis, clarifying why it can contribute to misleading results. Guidance for the handling and reporting of missing data in clinical research will be presented, and compared to current practice, with a focus on randomised controlled trials.
