ME patient experiences: Sampling bias limits the external validity of findings.

In a recent paper in , Kielland et al. present a study with the first objective of documenting how helpful or unhelpful persons with ME perceive common services and interventions. The authors recruited participants by respondent driven sampling, a method that aims to produce estimates that correct for sampling bias. However, we argue that the main assumptions of the method are not met, and that the results of the study thus cannot be generalised to the intended target population.