PURPOSE: Missing data can cause substantial bias, particularly in health economics and outcomes research (HEOR). However, there is a clear lack of guidance and best practices on the statistical methods to use when dealing with missing data in the HEOR context and how to report on these issues. This has implications within the wider economic modelling used in HEOR, with the aim of guiding decision-making. This workshop aims to review current practices on how missing data are accounted for in HEOR and highlight potential issues and pitfalls. DESCRIPTION: Cost-effectiveness and cost-utility analysis alongside clinical trials or observational studies using individual-level data (ILD) are a crucial component of the economic evaluation of healthcare interventions. However, almost invariably, ILD are affected by missingness, i.e. data may be only partially observed or not recorded at all. While there is extensive literature on missing data in the field of clinical statistics, this issue tends to be potentially more complex in HEOR because of: 1) the bivariate nature of the outcome (costs and benefits); 2) the fact that simplifying distributional assumptions (e.g. normality) are unlikely to hold; and 3) that the main interest is in aiding decision-making, rather than simply making inference about unobservable parameters. This workshop will discuss how missing data are routinely reported and dealt with in cost-effectiveness/utility analyses and highlight the potential issues arising from inadequate consideration and reporting of missing data and statistical modelling used, as well as identify areas in which guidance is most needed. Because ultimately models for ILD subject to missingness are likely to be based (at least partly) on un-testable assumption, these are crucial issues for modellers, regulators and sponsors, who all need to work collaboratively in order to validate the models and their resulting outcomes in terms of decision-making.