Cleaning up contamination with statistics: Experiences from a trial of a nurse-led intervention to support self-management in type 2 diabetes, and statistical approaches to address the problem.
Treatment contamination is defined as the receipt of treatment amongst patients in the control arm of a trial. It is a significant problem in trials of complex interventions in mental health, where it is often difficult to confine the treatments within the trial arms.
Khalida Ismail will describe her experiences of designing a trial of psychological treatment in diabetes where contamination was anticipated and mitigated for, but where there was still some receipt of active intervention in the control arm.
Nick Magill will then describe how contamination can be adjusted for in the analysis of treatment efficacy in a trial and will compare two options for addressing the problem when designing a trial.