MedEdPublish (May 2021)

Using a behavioural framework to optimize antibiotic prescribing by family medicine residents

  • Samantha Moe,
  • Tiffany Kan,
  • Charlene Soobiah,
  • Azy Golian,
  • Timothy Li,
  • Sumit Raybardhan

Journal volume & issue
Vol. 10, no. 1

Abstract

Read online

Background and objectives: Overprescribing of antibiotics in primary care is a prominent concern in the context of increasing antimicrobial resistance worldwide. Medical trainees are a key group to deliver thoughtful antimicrobial stewardship training. This study examined the factors influencing antibiotic prescribing for upper respiratory tract infections (URTI) by family medicine residents in order to identify educational interventions. Methods: Using purposive sampling of family medicine residents, semi-structured interviews were conducted until thematic saturation was reached. Interviews were coded into the domains of the Theoretical Domains Framework (TDF). Belief statements were created to characterize each domain and categorized as enablers or barriers to appropriate prescribing. Domains were plotted on the Behaviour Change Wheel (BCW) and intervention functions identified. Results: Twelve participants were interviewed. Nine domains of the TDF were relevant to antibiotic prescribing. Social influence was a prominent theme with the preceptor and patient being key influences on resident prescribing. Learning goals were also a key theme including the desire to strengthen independent clinical decision-making skills and improve antibiotic knowledge. Residents’ beliefs about capabilities were challenged when faced with diagnostic uncertainty. Additional domains included: professional role; environmental context and resources; intentions; beliefs about consequences and capabilities, and knowledge. Using the BCW, nine intervention functions were identified to change antibiotic prescribing behaviour. Conclusion: This study found nine domains of the TDF were relevant to family medicine resident antibiotic prescribing for URTI. Nine intervention functions could be used to guide intervention design.

Keywords