ABSTRACT
Most randomized controlled trials assume that the outcomes for each participant are independent-that is, it is unrelated to the outcome for any other participant. For a trial involving nonpharmacological treatments (NPTs), this assumption is at least open to debate and in some cases untenable. For trials involving skill-dependent interventions such as those delivered by a therapist or surgeon, it is realistic to expect participants under the care of the same health professional to be in°uenced in a similar manner as a result of the health professional practice, skill, and experience in delivering the intervention (see Chapter 4 for further discussion of the impact
CONTENTS
5.1 Introduction ..................................................................................................55 5.2 How to Measure the Clustering Effect .....................................................56
