How do you solve a problem like… missing data? Lessons from the PAin SoluTions In the Emergency Setting (PASTIES) trials

Abstract

When studies measure the effect of a treatment longitudinally over time, multiple data points are generated, but are sometimes incomplete. This may affect studies in a range of different settings, including studies on military patients. When analysing the data from such a study, one of the challenges is how to deal with missing data points in the final analysis. This article describes the possible different ways of dealing with missing data points in such a context, and outlines some pragmatic methods used for handling missing pain scores in the recent PAin SoluTions In the Emergency Setting (PASTIES) trials.