Stream 4

GP Registrars – developing and evaluating targeted interventions

The problem: Ingrained habits of GPs are one contributor to high antibiotic use. But GP registrars have been shown by CI van Driel and AI Magin to also be high prescribers.  Current role models, and other educational effects, are failing to adequately influence GP registrars. Yet it is essential to change their prescribing behaviour while it is still adaptable. Qualitative work has identified several barriers to reduced use of antibiotics, such as interactions with supervisors, practice culture and dealing with clinical uncertainty. Among enablers, our recent (not yet published) work shows that GP registrars are responsive to using delayed prescribing, which has also shown to be successful overseas. We identified an important barrier to reduced prescribing: clinical uncertainty, which especially worries GP registrars (publication under editorial review). We recently demonstrated some interventions successfully reduced antibiotic prescribing for acute bronchitis by 16%, but not undifferentiated ARIs (other specific ARIs were not measured).

The solution: Our team (CIs van Driel, Del Mar, Glasziou; AI Magin) will work collaboratively with Australian GP Regional Training Organisations, to expand the set of interventions developed or identified in CREMARA for registrars, and include new resources developed in Stream 2 to address other common infections (SSTIs and UTIs), and from Stream 3 for use in RACFs.

The team will thus develop an Antimicrobial Stewardship Program for use in GP training, informed by the barriers and enablers identified above. To ensure ongoing commitment, the team will work with the RACGP, which is responsible for curriculum and assessment of GP registrars. This stream emphasises the concept of Stewardship and provides early career GPs with tools to implement it in their own future practices.

Evaluation of these strategies will be by quasi experiments with pre-post intervention vignette-based evaluation of intended prescribing, and non-randomised trials with a non-equivalent control group nested within the ReCeNT cohort study. Pure RCTs are impracticable in this context. A mixed methods approach, which includes qualitative and participatory (including registrars as investigators) designs, will be used and this is expected to enhance the sense of “ownership” of the problem of antimicrobial resistance and solution in the next generation of GPs.