Attrition in Randomized and Preference Trials of Behavioural Treatments for Insomnia

Preferences for treatment contribute to attrition. Providing participants with their preferred treatment, as done in a partially randomized clinical or preference trial (PRCT), is a means to mitigate the influence of treatment preferences on attrition. This study examined attrition in an RCT and a PRCT. Persons with insomnia were randomly assigned (n = 150) or allocated (n = 198) to the preferred treatment. The number of dropouts at different time points in the study arms was documented and the influence of participant characteristics and treatment-related factors on attrition was examined. The overall attrition rate was higher in the RCT arm (46%) than in the PRCT arm (33%). In both arms, differences in sociodemographic and clinical characteristics were found between dropouts and completers. The type of treatment significantly predicted attrition (all p <= .05). The results provide some evidence of a lower attrition rate in the PRCT arm, supporting the benefit of accounting for preferences as a method of treatment allocation.