Potential explanations for control group benefit.

ABSTRACT

BACKGROUND: Estimating effectiveness of clinical interventions depends on detecting differences between the responses of intervention and control groups. The outcome, intervention, and moderating factors all may influence the between group change. The absence of a clinically or statistically meaningful difference may also result from control group improvement due to nonspecific factors such as participants' perception of attention, positive regard, expectations, desire to please, and therapeutic alliance with the care provider. PURPOSE: We examined perceived benefit and sources of benefit for control caregivers who participated in the CONNECT randomized controlled trial of a dementia caregiving intervention. METHODS: After the final scheduled data collection in CONNECT, control group participants were asked whether they believed they benefited from study participation. Those who reported benefit were asked to describe the benefit received. Data were analyzed qualitatively. RESULTS: Of 60 available control caregivers, 82% reported a perceived benefit from study participation in five areas getting information about dementia and caregiving; having someone to talk to and feeling supported; receiving understanding and validation of feelings; knowledge that others were in similar situations; and perceived appreciation of own abilities. Control caregivers who reported benefit were less burdened and depressed and spent less time on duty at baseline than those who did not report benefit. From caregivers' responses, we have identified the assessment battery, both content and time spent in data collection, as a possible mechanism of action for benefit. LIMITATIONS: Study limitations include the better baseline characteristics of the control caregivers who reported benefit, the sample size of benefit control caregivers, the possibility of perceptions of benefit being a function of social desirability, and the lack of a similar question about benefit being asked of intervention caregivers. CONCLUSIONS: These findings suggest that the outcome assessment questionnaires used in this and similar trials may be an intervention of sorts for caregiver participants in the control group. Caregivers who are less distressed may benefit more from the attention provided during assessment. These findings underscore the importance of asking caregivers how they are doing. For researchers, there is a need to develop innovative ways to assess outcomes to avoid unintended benefit of assessments.