Clinical trial design in the era of comparative effectiveness research

Abstract

Clinical trials are one of the key study designs in the evolving field of comparative effectiveness research. Evaluating the effectiveness of interventions in real-world settings is complex and demands a rethinking of the traditional clinical trial approach as well as transformation of the clinical trial landscape. Novel strategies and refinement of existing approaches have been proposed to generate evidence that can guide health care stakeholders in their decision process. The purpose of this review is to discuss clinical trial design approaches in the era of comparative effectiveness research. We will focus on aspects relevant to the type of clinical trial, study population and recruitment, randomization process, outcome measures, and data collection.

Translational Science

“…In this case, the units need to be similar, and then one unit uses the new protocol, and the other uses the usual protocol 13 . Thus, rather than assigning the intervention to a specific patient, it is assigned to a specific unit 13 . If the change is found to be better than usual care, then implementation research is the next step.…”

“…Effectiveness research can take the form of comparative research designs (Figure 2), where the change is compared to the current practice 2 . This type of study is best designed as a randomized controlled trial 2,11–13 . However, other designs are also relevant, such as comparing longitudinal cohort designs.…”

“…2 This type of study is best designed as a randomized controlled trial. 2,[11][12][13] However, other designs are also relevant, such as comparing longitudinal cohort designs. Cohort designs do not use random assignment.…”

“…13 Thus, rather than assigning the intervention to a specific patient, it is assigned to a specific unit. 13 If the change is found to be better than usual care, then implementation research is the next step. If the change is not found to be better, but perhaps it is faster or cheaper, and there is no difference in outcome, it may also be a reason to implement.…”

Comparing the Effects of Single‐ and Multiple‐Component Therapies for Insomnia on Sleep Outcomes

“…Preference-based allocation consisted of (a) having participants complete the Treatment Acceptability and Preference (TAP) scale (Sidani, Epstein, Bootzin, Moritz, & Miranda, 2009), which had a concluding question about participants' preferences for the therapies under evaluation; (b) randomly assigning participants indicating no preferences; and (c) allocating participants expressing preferences to the chosen therapy. The partially randomized preference trial has the advantages of improving enrollment rate and reducing attrition (Swift, Callahan, Ivanovic, & Kominiak, 2013) and is considered appropriate for comparative effectiveness research (Winter & Colditz, 2014). The limitation of this design is the possible non-comparability of participants assigned to the therapy groups on baseline variables, resulting from differences associated with preferences (Sidani, 2015); analysis of covariance was applied to control for these differences, if present.…”