Patient preferences and performance bias in a weight loss trial with a usual care arm

This qualitative study examines performance bias, i.e. unintended differences between groups, in the context of a weight loss trial in which a novel patient counseling program was compared to usual care in general practice.

14/381 consecutive interviewees (6 intervention group, 8 control group) within the CAMWEL (Camden Weight Loss) effectiveness trial process study were asked about their engagement with various features of the research study and a thematic content analysis undertaken.

Decisions to participate were interwoven with decisions to change behavior, to the extent that for many participants the two were synonymous. The intervention group were satisfied with their allocation. The control group spoke of their disappointment at having been offered usual care when they had taken part in the trial to access new forms of help. Reactions to disappointment involved both movements toward and away from behavior change.

There is a prima facie case that reactions to disappointment may introduce bias, as they lead the randomized groups to differ in ways other than the intended experimental contrast.

In-depth qualitative studies nested within trials are needed to understand better the processes through which bias may be introduced.