Engaging rural communities in Bangladesh to address antimicrobial resistance via the community dialogue approach: a protocol for a cluster-randomised controlled trial

Rebecca King¹Joseph Paul Hicks^1*Fariza Fieroze²Md. Badruddin Saify²S M Abdullah²Dani Barrington³Prudence Hamade⁴Helen Hawkings⁴Tim Ensor¹Sophia Latham⁵Jessica Mitchell¹Amam Zonaed Siddiki⁶Rumana Huque²

• ¹Nuffield Centre for International Health and Development, University of Leeds, Leeds, United Kingdom
• ²ARK Foundation, Dhaka, Dhaka, Bangladesh
• ³School of Population and Global Health, University of Western Australia, Perth, Australia
• ⁴Malaria Consortium, London, United Kingdom
• ⁵Department of Livestock and One Health, University of Liverpool, Liverpool, United Kingdom
• ⁶Faculty of Veterinary Medicine, Chattogram Veterinary and Animal Sciences University, Chattogram, Bangladesh

Introduction To effectively tackle antibiotic resistance (ABR) a One Health approach is required, focusing on the human, animal and environmental sectors together, and that public education and engagement programmes must be part of the overall approach. However, there has been limited research on such programmes in low-/middle-income countries (LMICs). Here we describe our plans to evaluate a community-engagement programme, known as the community dialogue approach, that takes a One Health approach to tackling ABR in rural communities in Bangladesh, and involves community-led and community-based education and discussion forums. Members of our team previously developed this approach and used it to address other health issues in other LMIC contexts, while our team has previously adapted it for this topic and setting. Methods We will use a pragmatic, non-blinded, two-arm, parallel-group, cluster-randomised, controlled trial to primarily evaluate whether the intervention can improve 1) the level of correct and appropriate knowledge about antibiotics, ABR, and antibiotic usage from a One Health perspective, 2) levels of awareness about the existence of antibiotics and ABR, and 3) the relative frequency of self-reported and observable indicators of best practices related to antibiotic usage. Within Cumilla district, we will randomise 50 clusters of villages in a 1:1 ratio. In intervention community clusters trained community volunteers will deliver a set of 11 health education and discussion forums across a 12-month period, while control community clusters will receive no inputs. We will collect outcomes at baseline (pre-randomisation) and endline (following the final community dialogue) via two repeated cross-sectional household surveys (each aiming to survey 2,200 participants across all clusters). We will also conduct nested process evaluation and costing studies. Discussion Community engagement approaches have successfully addressed other health issues in low resource settings, but there is limited evidence on using community engagement approaches to address ABR in low resource contexts, particularly in Bangladesh. We will closely involve the Bangladeshi health system in this research to ensure feasibility and facilitate scale-up via an embedded approach.