Medication errors cause preventable patient harm with annual costs of USD$42 billion globally. In England, 237 million errors occur annually, accounting for 1700 deaths, with trainee doctors being responsible for the highest proportion of errors. The UK introduced the national Prescribing Safety Assessment (PSA) in 2013 to ensure prescribing competency. Imperial College School of Medicine (ICSM) sought to increase learning opportunities and reduce errors through a monthly Prescribing Practice Questions (PPQs) programme. PSA scores from 2020 (prior to PPQ introduction) to 2024 (first cohort to receive entire PPQ programme), the prescription writing skills (PWS) subsection and applied knowledge test (AKT) scores in 2020 and 2024 were analysed. 1505 students sat the PSA between 2020 and 2024. PSA fails significantly reduced from 2.51% in 2020 to 0% in 2024 (p=0.0054). Median PSA scores significantly improved from 78.5% in 2020 to 84.0% in 2024 (p<0.0001). Median AKT scores decreased between 2020 and 2024 (78.0% vs 74.5%, p<0.0001). Absolute increase in PSA scores from lowest to highest quintiles was 10 percentage points (pp), 7pp, 5pp, 4pp and 1pp, respectively, between 2020 and 2024. Median PWS scores significantly improved from 76.3 % (61/80) in 2020 to 87.5% (70/80) in 2024 (p<0.0001). Since introducing PPQs, PSA and PWS scores have statistically significantly improved, with progressively greater impact on students at the bottom of the performance distribution. Early, targeted and repeated opportunities for authentic prescribing activities, such as the PPQs, in undergraduate training may lead to significant improvement in prescribing competency as determined by the PSA.

Objectives: To identify knowledge and attitude factors influencing primary care clinician decision-making in diagnosing, managing, and treating urinary tract infections. Design: A qualitative think-aloud study. Methods: Semi-structured qualitative interviews were conducted with primary care clinicians in England over Microsoft Teams. Interviews were transcribed and coded in two ways. First, clinicians’ responses for each scenario were coded as either following (optimal) or not following (suboptimal) evidence-based national guidelines. Second, the knowledge and attitude factors that influenced decision-making were coded according to an empirically-informed umbrella framework. Clinicians external to the study team reviewed the findings to promote their trustworthiness and utility. Setting: English primary care clinicians with the right to prescribe medications in England. Sample: Ten clinicians with prescribing rights in primary care took part. Results: Despite clinicians’ expressing high awareness of relevant evidence-based guidelines (a knowledge factor) and high confidence (an attitude factor), more than half of their decisions were suboptimal in some way. Our framework analysis suggests that knowledge could impede adherence, e.g., where local guidelines conflicted with national guidelines. Conclusions: Suboptimal prescribing decisions could result from a combination of different knowledge and attitude factors. Most clinicians relied on their experiential knowledge rather than using evidence-based guidelines. To optimise antibiotic prescribing, policy-level interventions could increase concordance across local and national guidelines, or more tailored individual-level interventions could help clinicians recognize where their experiential knowledge causes deviations from evidence-based guidelines when diagnosing, treating, and managing urinary tract infections.