Outcome selection in primary care antimicrobial stewardship research.

Clinical and antimicrobial stewardship (AMS) outcomes are highly relevant to pragmatic primary care trials, reflecting aspects, such as persistent symptoms and relapses, or antibiotic use and antimicrobial resistance. Sometimes both can be equally important. We present evidence demonstrating the wide range of outcome measures used in previous primary care trials and observe that there are no agreed standards for their design. We describe AMS interventions and outcomes in terms of intervention types and targets, and we make recommendations for future research designs. Specifically, we argue that: (i) where co-primary outcomes are considered appropriate, investigators should pre-specify interpretation of conflicting results; (ii) intervention evaluation should ensure prescriptions from sources outside of the usual provider are included in any AMS effectiveness measure; (iii) where possible, outcomes should include antimicrobial resistance; (iv) in some contexts, it may be necessary to include the antibiotics used within the intervention as part of the outcome; and (v) patient involvement is needed to establish the principles investigators should use when deciding whether the AMS or clinical outcomes should be prioritized.