Consensus-based antimicrobial resistance and stewardship competencies for UK undergraduate medical students

Abstract Background In the UK there is limited coverage of antimicrobial stewardship across postgraduate curricula and evidence that final year medical students have insufficient and inconsistent antimicrobial stewardship teaching. A national undergraduate curriculum for antimicrobial resistance and stewardship is required to standardize an adequate level of understanding for all future doctors. Objectives To provide a UK national consensus on competencies for antimicrobial resistance and stewardship for undergraduate medical education. Methods Using the modified Delphi method over two online survey rounds, an expert panel comprising leads for infection teaching from 25 UK medical schools reviewed competency descriptors for antimicrobial resistance and stewardship education. Results There was a response rate of 100% with all 28 experts who agreed to take part completing both survey rounds. Following the first-round survey, of the initial 55 descriptors, 43 reached consensus (78%). The second-round survey included the 12 descriptors from the first round in which agreement had not been reached, four amended descriptors and 12 new descriptors following qualitative feedback from the panel members. Following the second-round survey, a total of 58 consensus-based competency descriptors within six overarching domains were identified. Conclusions The consensus-based competency descriptors defined here can be used to inform standards, design curricula, develop assessment tools and direct UK undergraduate medical education.


Introduction
Antimicrobial resistance (AMR) is one of the greatest threats to the future of healthcare. 1 AMR occurs when microorganisms are exposed to antimicrobial drugs, with the misuse and overuse of antimicrobials accelerating the development of resistance. 1 Infection with resistant microorganisms can have severe consequences, increasing mortality, prolonging hospital stays, adding a significant economic burden and threatening to undermine the global health improvements made over recent decades. 2,3 Antimicrobial stewardship (AMS) is a coherent set of actions which promote responsible use of antimicrobials and is recognized as essential for limiting AMR. 4 The WHO Global Action Plan on AMR emphasizes the importance of including AMS and antimicrobial prescribing in the training of health workers, calling for increased awareness and training by making AMR a core component of professional education. 2 A variety of factors can result in injudicious use of antimicrobials by health workers, including fundamental lack of knowledge and awareness of AMR. In the UK there is inadequate coverage of AMS across the majority of postgraduate clinical training curricula, including specialities responsible for the largest volumes of antimicrobial usage (e.g. primary care) and hospital specialities which have high rates of broad spectrum antimicrobial use and healthcare-associated infections. 5 The UK Foundation Programme curriculum, General Medical Council outcomes for graduates and the recently produced Royal College of Pathologists undergraduate curriculum are helpful guidance, but are unable to address AMR/S in sufficient detail to guide educators or standardize competencies. [6][7][8] Education on AMR/S must be improved for all prescribers, including pharmacists, midwives, nurses, allied health professionals and doctors. [9][10][11] Medical school prepares doctors for clinical practice, however there is evidence that final-year medical students have insufficient prescribing competencies and lack confidence in correctly prescribing antibiotics, despite these being among the most common medications junior doctors prescribe. [12][13][14][15][16][17][18] Competencies represent a combination of knowledge, attitudes and skills and are designed to define the minimum standards that should be reached to practise responsibly and safely. 9 Whilst AMR/S prescribing competency frameworks exist, [9][10][11]19 there are no consensus-based AMR/S competencies for UK medical student education.
As part of the Keep Antibiotics Working (KAW) undergraduate programme, a joint BSAC, Health Education England and Medical Schools Council initiative was developed with the aim of providing a national consensus of competencies on AMR/S for UK undergraduate medical students.

Ethics
The research was conducted in accordance with the Declaration of Helsinki and national and institutional standards. Ethical approval was granted by the School of Healthcare Sciences Research Governance and Ethics Committee, Cardiff University (reference number 427).

Delphi method
Using the modified Delphi method, 20 the opinions of experts were gathered over two rounds of data collection. In collaboration with the Medical Schools Council, 28 experts (12 female; 16 male) were identified from 25 UK medical schools in England, Wales and Northern Ireland (Appendix S1, available as Supplementary data at JAC-AMR Online). After discussion with the Scottish Antimicrobial Prescribing Group, the decision was made to not include Scottish medical schools to avoid multiple simultaneous quality improvement initiatives. An expert in this study was defined as an individual from each institution with expertise in antimicrobial prescribing and medicines management, AMR/S and leading infection education for undergraduate medical students. Experts came from backgrounds in infectious diseases or microbiology and all were involved in leading undergraduate teaching. All experts were sent a participant information sheet and given the opportunity to discuss any queries with a researcher. All 28 experts agreed to participate.
A comprehensive list of core competencies was generated from available frameworks, 9,11,19 allowing use of the modified Delphi method due to the availability of pre-existing information. 20 These competencies were split into six domains, with overarching competency statements and a list of 55 descriptors designed to represent the knowledge, skills, attitudes and values that are required for undergraduate medical students (Table 1). As the competency frameworks used to generate the first-round survey had not been created for UK medical students it was important to include all descriptors for review in the first round.
The Joint Information Systems Committee online survey tool was used to develop each round of the online survey, with each survey round open for 3 weeks between May and July 2020. Each expert was emailed a link to complete the first round of the survey. Participants were asked to rank each descriptor on a six-point Likert scale (1 " strongly disagree; 6 " strongly agree) as to the extent to which they assessed it was important to be part of the undergraduate curriculum. An additional open-ended question was included at the end of each domain for experts to include comments, provide feedback and to identify any additional descriptors.
Following the first round the results were analysed by a steering group, ensuring all qualitative feedback was addressed and that no descriptors were unnecessarily excluded. Following analysis of the first-round results, a report was circulated to respondents detailing the quantitative results and inviting further interpretation and feedback. Descriptors for which there was a lack of agreement, descriptors that were amended following feedback and additional descriptors identified by respondents were included in the second-round questionnaire. The second-round questionnaire was sent to all experts who had responded to the first-round survey. Follow-up reminder emails were sent at weekly intervals across the two survey rounds.

Data analysis
The most frequently used and robust method to determine consensus in Delphi studies is medians and IQRs. 21 Medians and IQRs were calculated for each descriptor; responses where the median was 5 (i.e. experts agreed or strongly agreed with the importance of including) with an IQR 1.5 (i.e. there was minimal spread between expert answers) were considered important descriptors that had reached expert consensus.

First round
Of the 28 individuals who agreed to participate in the expert panel there was a 100% response rate for completion of the first-round questionnaire. There were high levels of agreement (i.e. median 5) on the importance of 51 descriptors (Table 1). Two descriptors were viewed as less important (i.e. median ,5) with an IQR of 1, indicating there was a high strength of agreement with minimal spread between expert answers; these two descriptors were excluded: '2.4 Describe the mechanisms of antimicrobial resistance, including: intrinsic or acquired resistance and the importance McMaster et al.  2.4 Describe the mechanisms of antimicrobial resistance, including: intrinsic or acquired resistance; the importance of selection advantages (e.g. the greater ability for some to colonize) and how this can be an amplification process for antimicrobial resistance 4 a 1 3.9 Demonstrate an understanding of how to interpret microbiology results/reports from the laboratory 5 1 3.10 Describe and demonstrate switching to the correct antimicrobial when susceptibility testing indicates resistance, or to a cheaper or more cost-effective antimicrobial that is also compatible with the clinical presentation 4.6 Aware of local and national targets for immunization uptake and why vaccine uptake data is important. If appropriate, know where to find data for their area of practice 4 a 1 Domain 5: Person-centred care 5.1 Support participation of patients/carers, as integral partners when planning/delivering their care 5.5 1 5.2 Share information with patients/carer in a respectful manner and in such a way that is understandable, encourages discussion, and enhances participation in decision-making 6 1 5.3 Ensure that appropriate education and support is provided by learners to patients/carer, and others involved with their care or service 5 1

Continued
McMaster et al.
of selection advantages' and '4.6 Aware of local and national targets for immunization uptake and why vaccine uptake data is important. If appropriate, know where to find data for their area of practice'. Additionally, two descriptors were viewed as less important but with a low strength of agreement between experts: '1. 16 Demonstrating knowledge and awareness of international/national strategies on infection prevention and control and antimicrobial resistance (e.g. Global Action Plan for AMR; WHO SAVE LIVES: Clean Your Hands; UK Government 5-year AMR strategy)' and '2.1 Describe the modes of action of antibiotics and other antimicrobials'. These 2 descriptors, and a total of 10 descriptors with a low strength of agreement between experts (i.e. IQR .1.5) were included in the second-round survey. Of the initial 55 descriptors, 43 reached consensus (i.e. median 5 and IQR 1.5) by the expert panel (78%). The 12 descriptors which had disagreement, 4 amended descriptors and 12 new descriptors identified following qualitative feedback formed the second round of the survey ( Table 2).

Second round
All 28 members of the expert panel were invited to complete the second-round questionnaire; there was a 100% response rate to the second round. Three descriptors were viewed as less important, with minimal spread between expert answers (i.e. median ,5 and IQR ,1.5) and were excluded (

Summary statement
We have identified 58 competency descriptors within six domains that have reached expert consensus by a group representing UK medical schools from England, Wales and Northern Ireland. These competencies form an AMR and AMS framework for undergraduate medical student education (Appendix S2).

Discussion
Utilizing expert opinion from across the UK, we present here the first set of specific AMR/S competencies for UK undergraduate medical student education. This framework can be used to inform standards for education and prescribing and will help standardize a high-level of antimicrobial knowledge for tomorrow's doctors. The competencies defined here have been developed to address a gap in UK undergraduate medical student education and to ensure that all new graduates are trained in the principles of evidence based AMS. In the EU alone, it is estimated that infections from MDR bacteria result in 25 000 deaths annually. 3 Hence, it is not surprising that medical students perceive misuse of antibiotics as unethical. 13 Castro-Sánchez et al. 22  Evidence shows that many medical students lack self confidence in choosing the correct antibiotics, deciding when to use  Awareness of factors contributing to AMR including inappropriate prescribing by healthcare workers and the sale of antimicrobials without prescription (e.g. over the counter in some parts of the world; online sales) 5 1 Understand the link between antimicrobials and the human microbiome and how this facilitates spread of resistant organisms 5 2 b Aware of which vaccinations healthcare workers should receive in addition to standard UK immunizations 4 a 1 Domain 3: Antimicrobial prescribing and stewardship Describe key features of specific infections and the best narrow spectrum antibiotics to prescribe and length of antibiotic course in these scenarios (e.g. UTI, pneumonia, cellulitis) 6 1 Understand how to request and interpret basic diagnostic tests that can guide antimicrobial therapy (e.g. microbiology, radiology, immunology)  McMaster et al.
combination therapy and choosing the correct dose and interval of administration, with up to 98% of students wanting more training on antibiotic use during medical school. [12][13][14][15][16][17][18][23][24][25][26][27] A study of selfreported preparedness for prudent antibiotic use among final-year medical students in 29 European countries reported that UK students felt least prepared on selecting initial empirical therapy without using guidelines. 27 This further highlights gaps in UK undergraduate medical education, with an overreliance on guidelines rather than a fundamental understanding of how to select appropriate antibiotics for common infections. A survey of junior doctors in the UK and France identified gaps in knowledge on the prevalence of antibiotic resistance and antibiotic misuse, despite 98.6% of those surveyed having prescribed an antibiotic within the last 6 months. 28 To improve AMS in the UK a coherent and consistent national approach must be taken to create high-quality educational resources to train current and future healthcare workers and improve individual practice. Using predefined competency statements with associated descriptors, experts representing medical schools in the UK formed a consensus on core AMR/S competencies for UK undergraduate medical students. There was a high response rate to both rounds of the Delphi process, with consistently high levels of agreement for many descriptors. Within the overarching domains of: 'Infection prevention and control', 'Antimicrobials and antimicrobial resistance', 'Antimicrobial prescribing and stewardship', 'Vaccine uptake', 'Person-centred care' and 'Interprofessional collaborative practice' we have reached consensus on 58 competency descriptors. These competencies can be used by professional bodies, regulators and education providers to inform standards, design curricula, create teaching materials and assess learning outcomes.
Strengths of the study include the use of a robust methodology, a high response rate and the opinions of a defined panel of experts. Some may consider the professional background of the experts in this study, including mostly specialists in infectious diseases or microbiology, a limitation in reaching consensus on relevance to all prescribers. However, experts in this group were also selected due to their role in undergraduate infection teaching and are therefore likely to have a pragmatic understanding of expectations and limitations of what can be included within undergraduate medical student education. In addition, the competency frameworks used to form the first-round survey had involved much broader input from other healthcare professionals (e.g. dentists, nurses, midwives, pharmacists), allowing the specialist expert group here to prioritize specific competencies for future doctors.
Preventing the rise of AMR requires multifactorial interventions and collaboration between healthcare professionals. It is essential and urgent that AMR and AMS competencies are embedded into the curricula of all healthcare professionals, including medical students, and we encourage those that develop curricula to adopt a similar process to develop competencies specific to their professional group.