Behaviour change interventions to influence antimicrobial prescribing: a cross-sectional analysis of reports from UK state-of-the-art scientific conferences

In the United Kingdom (UK), about one third of all hospital inpatients receive antimicrobials during their admission with a significant proportion of these identified as inappropriate [1‐3]. This accounts for a large amount of unnecessary antimicrobial exposure. Antimicrobial resistance (AMR) is a leading patient safety issue that requires urgent interventions to curb its exponential growth. One target of interventions to address the problem of AMR is the promotion of the appropriate use of antimicrobials in humans, which is thought to be a leading driver for the growth of AMR [4].

To address this and promote the appropriate use of antimicrobial agents a number of national and international antimicrobial stewardship (AMS) initiatives have been implemented [5‐8]. A key facet of these interventions targets improving and sustaining individual prescribing behaviours. Implementation of AMS programmes have been demonstrated to reduce rates of AMR and improve health and economic outcomes [9‐11]. However, despite these positive steps forward, several challenges appear to remain in promoting the sustainable use of antimicrobials across clinical practice [12].

Firstly, there is a growing body of evidence to describe the cultural and social factors that influence antimicrobial prescribing across healthcare settings as well as qualitative data that supports the role of behaviour change interventions in improving antimicrobial prescribing [13‐16]. Despite this, very little evidence exists to describe the current landscape of behaviour change interventions being implemented within this field [12, 13, 15‐20]. Secondly, despite evidence to support engagement of infection specialists with the AMS-AMR agenda, there appears to be poorer engagement across other clinical specialties in terms of formal training and awareness at state-of-the-art scientific conferences [21‐23]. Finally, although there are described frameworks and taxonomy’s available from which to begin mapping behaviour change methods [24, 25], very little data is currently available to describe the appropriateness of these specifically for AMS interventions.

In this cross-sectional study we aimed to explore antimicrobial stewardship interventions reported at major cross specialty UK state-of-the-art scientific conferences in 2015, which contained behaviour change interventions. We aimed to determine the number and type of behaviour change interventions reported by different specialties and compare these to currently available behaviour change taxonomies to identify potential gaps and highlight potential targets for future interventions.

Clinical state-of-the-art conferences provide an opportunity for medical professionals to participate in research and reporting. They also allow us to gain an insight into different levels of research being undertaken within the field; from small scale research undertaken at the local level, to large scale studies being performed by key opinion leaders and organisations. This provides a window into the activities within specialties that is less influenced by publication bias than can often be observed through systematically reviewing peer-reviewed publications. Within this study, we observed a low rate of behaviour change intervention reporting across the majority of specialty state-of-the-art conferences in 2015. Infection specialties reported a significantly greater number and broader variety of AMS-AMR interventions with the majority of interventions reported by non-infection specialties falling into a narrow band of intervention and policy based functions.

These observations are concerning given the recent focus placed upon the need for cross-specialty engagement with AMS-AMR and behaviour change interventions. This has been supported broadly in the literature [4, 13, 16, 21, 22, 39]; by national organisations including Public Health England (PHE) [18], the British Society for Antimicrobial Chemotherapy (BSAC), and government [40]; and major international governments and organisations [6, 7]. These findings highlight the need to broaden the focus of AMS campaigns beyond infection specialties to promote leadership from within cohorts, which can drive behaviour change towards antimicrobial use.

This must be supported by clear and defined tools to help specialties engage and design AMS-AMR behaviour change interventions and assess the impact of these on patient outcomes. Whilst the use of behaviour change taxonomies allows content of interventions to be coded and categorised, facilitating the analysis of behaviour change interventions [24, 41], they are not appropriate for determining the effectiveness of interventions. Furthermore, for the field of AMS-AMR there remains no definition on what an appropriate behaviour change intervention is. This makes evaluating behaviour change challenging as certain functions of any framework used may be irrelevant or actually have a negative impact on behaviours within this context. [24, 41, 42] Kok and colleagues argue that as behavioural determinants are often specific to behaviours, populations, and contexts characterisations should thus be individualised and tailored for such [24, 41]. This will require engagement and drive from within clinical specialties to review current practices, define the context in which AMS-AMR interventions need to be implemented, and then tailor behaviour change interventions to optimise their effect within their local environment.

Finally, in a previous study “high risk” specialties were identified that currently use large amounts of antimicrobials and also experience high levels of healthcare associated infections [21]. This study reported that certain “high risk” specialties such as infection and intensive care had relatively high levels of engagement with AMS-AMR, whilst other specialties such as haematology and nephology tended to have a low apparent engagement at scientific conferences [21]. Within this study, we have observed an overall low rate of behaviour change interventions across all high risk specialties reported previously with only infection [29] and nephrology [43] reporting any behaviour change interventions in 2015. Furthermore, specialties with relatively high coverage of AMS-AMR at scientific conferences, such as Genitourinary Medicine, failed to report on behaviour change interventions despite having 9% coverage of AMS-AMR topics in 2015. This highlights the need for greater pan-specialty promotion of behaviour change interventions for AMS-AMR given the significant lack of focus on reporting such interventions currently.

There are several limitations to this study. Firstly, we only selected one leading state-of-the-art conference for each major clinical specialty in the UK. This makes our findings difficult to generalise across other countries and also may have introduced bias through excluding smaller, conferences and meetings, where specialties may have had more of an AMS-AMR focus. However, this method was selected as we aimed to generate a representative picture of current behaviour change interventions and the importance placed upon this by different specialties. By selecting leading state-of-the-art conferences we hoped that this would reflect the current overall importance of this subject within the specialty as well as allow for a more representative view of work being undertaken in the field. Secondly, we did not review invited talks and seminars provided by conferences given that they often were not presenting original data or results. Furthermore, as the taxonomy used to describe behaviour change does not allow for evaluation of the effectiveness of interventions it is not possible to evaluate whether interventions described were “appropriate” for the context in which they were described. Finally, as only a relatively small number of specialists from a specific geographical area were contacted to seek opinions on defining the largest UK conference in their field this may have introduced bias in our conference selection. To address this we ensured that two specialists from each field were contacted for their opinions. When there was discrepancy in responses from the individuals, conference attendance size was considered as the determinant with the conference with the largest attendance selected.

In conclusion, despite evidence to support the role of behaviour change interventions for improving antimicrobial prescribing, very few studies reported implementing them at UK state-of-the-art conferences in 2015. Future research must focus on providing appropriate frameworks and mechanisms to allow clinical specialties to engage with AMS-AMR and design and evaluate the impact of behaviour change interventions within their specific contexts.