Protocol for a Randomised Trial of Higher Versus Lower Intensity Patient-Provider Communication Interventions to Reduce Antibiotic Misuse in Two Paediatric Ambulatory Clinics in the US

Introduction

Children with acute respiratory tract infections (ARTIs) are prescribed up to 11.4 million unnecessary antibiotic prescriptions annually. Inadequate parent-provider communication is a chief contributor, yet efforts to reduce overprescribing have only indirectly targeted communication or been impractical. This paper describes our multisite, parallel group, cluster randomised trial comparing two feasible interventions for enhancing parent-provider communication on the rate of inappropriate antibiotic prescribing (primary outcome) and revisits, adverse drug reactions and parent-rated quality of shared decision-making, parent-provider communication and visit satisfaction (secondary outcomes).

Methods/Analysis

We will attempt to recruit all eligible paediatricians and nurse practitioners (currently 47) at an academic children's hospital and a private practice. Using a 1:1 randomisation, providers will be assigned to a higher intensity education and communication skills or lower intensity education-only intervention and trained accordingly. We will recruit 1600 eligible parent-child dyads. Parents of children ages 1-5 years who present with ARTI symptoms will be managed by providers trained in either the higher or lower intensity intervention. Before their consultation, all parents will complete a baseline survey and view a 90 s gain-framed antibiotic educational video. Parent-child dyads consulting with providers trained in the higher intensity intervention will, in addition, receive a gain-framed antibiotic educational brochure promoting cautious use of antibiotics and rate their interest in receiving an antibiotic which will be shared with their provider before the visit. All parents will complete a postconsultation survey and a 2-week follow-up phone survey. Due to the two-stage nested design (parents nested within providers and clinics), we will employ generalised linear mixed-effect regression models.