Grant: CIHR Fellowship

The cluster-randomized trial is an important design for evaluating health service interventions in the outpatient hemodialysis (HD) setting. Cluster-randomized trials randomize intact groups of people (such as patients being treated in HD centres) rather than individuals. In the HD setting, a cluster-randomized trial design is a logical choice when the intervention is delivered at the centre-level, where nurses in each centre are trained to follow similar protocols for all patients. Also, when study outcomes are available from routinely collected sources (called "registries"), cluster randomization gives researchers an efficient way for following patients during the trial. This design and registries can reduce trial costs, time, and patient loss to follow-up. Moreover, this design can help avoid treatment contamination when patients in the same HD centre are on the same intervention compared to a setting where multiple interventions are in place at one centre (e.g., evaluating different advertising strategies to encourage smoking cessation). However, there are some challenges to conducting cluster-randomized trials. This fellowship addresses three methodological gaps in the existing literature to design, conduct, and analyze CRTs unique to the HD setting. The three proposed projects are embedded within a CIHR-funded platform with the primary objective of conducting large-scale, pragmatic, registry-based CRTs in the HD setting. Our results will directly inform the design and analysis of Canadian HD trials within our funded SPOR Innovative Clinical Trial Multi-Year Grant.