Community participatory learning and action cycle groups to reduce type 2 diabetes in Bangladesh (D:Clare trial): study protocol for a stepped-wedge cluster randomised controlled trial

Background: An estimated 463 million people globally have diabetes, with the prevalence growing in low-and middle-income settings, such as Bangladesh. Given the need for context-appropriate interventions to prevent type 2 diabetes mellitus (T2DM), the 'Diabetes: Community-led Awareness, Response and Evaluation' (D:Clare) trial will rigorously evaluate the replication and scale-up of a participatory learning and action (PLA) cycle intervention in Bangladesh, to inform policy on population-level T2DM prevention and control. Methods: This is a stepped-wedge cluster randomised controlled trial, with integrated process and economic evaluations, conducted from March 2020 to September 2022. The trial will evaluate a community-based four-phase PLA cycle intervention focused on prevention and control of T2DM implemented over 18 months, against a control of usual care. Twelve clusters will be randomly allocated (1:1) to implement the intervention at project month 1 or 12. The intervention will be evaluated through three cross-sectional surveys at months 1, 12 and 24. The trial will be conducted in Alfadanga Upazila, Faridpur district, with an estimated population of 120,000. Clusters are defined as administrative geographical areas, with approximately equal populations. Each of the six unions in Alfadanga will be divided into two clusters, forming 12 clusters in total. Given the risk of inter-cluster contamination, evaluation surveys will exclude villages in border areas. Participants will be randomly sampled, independently for each survey, from a population census conducted in January 2020. The primary outcome is the combined prevalence of intermediate hyperglycaemia and T2DM, measured through fasting and 2-h post-glucose load blood tests. A total of 4680 participants provide 84% power to detect a 30% reduction in the primary outcome, assuming a baseline of 30% and an ICC of 0.07. The analysis will be by intention-to-treat, comparing intervention and control periods across all clusters, adjusting for geographical clustering. Discussion: This study will provide further evidence of effectiveness for community-based PLA to prevent T2DM at scale in a rural Bangladesh setting. However, we encountered several challenges in applying the stepped-wedge design to our research context, with particular consideration given to balancing seasonality, timing and number of steps and estimation of partial versus full effect.