Advertisement

The use of complexity theory to inform antimicrobial stewardship: a scoping review

Open AccessPublished:June 15, 2022DOI:https://doi.org/10.1016/j.jhin.2022.06.004

      Abstract

      Background

      Complexity theory has previously been used as a conceptual lens in human healthcare research. Antimicrobial stewardship (AMS) is an inherently complex healthcare intervention; however, the extent to which complexity has been operationalized in AMS is currently unclear.

      Aim

      To investigate if, and how, complexity theory has been used to inform AMS in human healthcare.

      Methods

      Scoping review methodology. Empirical research or policy specifically referencing complexity in relation to AMS were considered in any human healthcare setting and geographical location. Databases searched were: Cinahl, Cochrane Library, Embase, Medline, National Institute for Health and Care Excellence, PsycInfo, Scopus and Web of Science from inception to June 2020. Grey literature and other databases searched: EVIPNet, Google, Mednar, Proquest Theses, and the World Health Organization library of national antimicrobial resistance action plans. Non-English language articles were excluded.

      Results

      Of 612 records retrieved, 8 articles were included. Heterogeneity in study design and geographical location were noted. Three interventional studies evaluated AMS in hospital (n = 2) and long-term care (n = 1) settings. Remaining studies were non-interventional and proposed AMS strategies conceptualized through complexity theory. The importance of close engagement between researchers or policy administrators and the target population was emphasized in all studies, as a means of ensuring AMS relevance and success.

      Conclusions

      There is a paucity of AMS research informed by complexity theory, and no policy documents could be located using complexity as a guiding theory. Mixed methods, informed by complexity theory, is a potentially suitable strategy to develop, implement and evaluate AMS as a complex intervention.

      Introduction

      Complexity theory is a broad concept based on “…relationships, emerging patterns and interactions.” [] In its simplest form, complexity is the antithesis to traditional ‘cause and effect’ models, where the assumption is that addressing a historical event or issue will produce a predictable future response. Examples of this linear thinking can be found in manufacturing facilities. Such processes are arranged in sequence to yield predictable and standardized products, for example an aseptic compounding unit in a hospital pharmacy. Failure of a linear process can usually be addressed by deconstructing it to constituent parts to identify a dysfunctional component. Complex outcomes, on the other hand, are often unpredictable in nature and arise from multiple interactions between components in a non-linear fashion. [,
      • Plsek P.E.
      • Greenhalgh T.
      Complexity science: The challenge of complexity in health care.
      ] Examples of everyday complex processes are interactions between air and water to produce weather events or the interaction of plants and animals in an ecological network. []
      Complexity theory can also be applied to systems thinking, where a system can be thought of as an entity with multiple interacting components. [
      • Peters D.H.
      The application of systems thinking in health: why use systems thinking?.
      ] A complex adaptive system (CAS), therefore, occurs where the interaction between system components can be unpredictable, random, and not easily modifiable. [] Plsek and Greenhalgh define a CAS as:“…a collection of individual agents with freedom to act in ways that are not always totally predictable, and whose actions are interconnected so that one agent’s actions changes the context for other agents.” [
      • Plsek P.E.
      • Greenhalgh T.
      Complexity science: The challenge of complexity in health care.
      ]
      Antimicrobial resistance, a problem propagated from interaction between multiple different factors, within and outside of human healthcare, is inherently complex. [
      • Holmes A.H.
      • Moore L.S.
      • Sundsfjord A.
      • et al.
      Understanding the mechanisms and drivers of antimicrobial resistance.
      ] Antimicrobial stewardship (AMS) is a strategy aimed at addressing AMR by optimizing antimicrobial use. [
      • Dellit T.H.
      • Owens R.C.
      • McGowan J.E.J.
      • et al.
      Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship.
      ] It is a complex intervention in itself, and relies on the interaction between multiple actors in various healthcare contexts and settings to ensure safe and effective antimicrobial therapy. [
      • Dellit T.H.
      • Owens R.C.
      • McGowan J.E.J.
      • et al.
      Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship.
      ,

      Health Protection Surveillance Centre. Guidelines for antimicrobial stewardship in hospitals in Ireland. 2009. Available from: https://www.hpsc.ie/a-z/microbiologyantimicrobialresistance/infectioncontrolandhai/guidelines/File,4116,en.pdf.

      ] Despite the similarities of AMS with principles of complexity, there is a paucity of empirical research on the application of complexity science to rational antimicrobial use. Furthermore, although established action plans and policies to address AMR firmly advocate a One Health approach, there is little or no suggestion that complexity has been applied to the One Health Agenda.
      Thompson et als scoping review of complexity theory in health services research found a focus on relationships, self-organisation and diversity as aspects of complexity theory. [
      • Thompson D.S.
      • Fazio X.
      • Kustra E.
      • et al.
      Scoping review of complexity theory in health services research.
      ] However, none of the articles focused on either AMR or AMS specifically. Notably, Thompson et al excluded quality improvement (QI) studies and included articles published up until June 2015. Talkhan et al have recently published a systematic review on the use of theory in the development and evaluation of behaviour change interventions to improve antimicrobial prescribing. [
      • Talkhan H.
      • Stewart D.
      • Mcintosh T.
      • et al.
      The use of theory in the development and evaluation of behaviour change interventions to improve antimicrobial prescribing: a systematic review.
      ] Their review did not identify the use of complexity as an informative element to behaviour change interventions. However, the authors focused on primary studies and did not search the grey literature.

      Aim

      The aim of this scoping review was to describe if, and how, complexity theory has been used to inform AMS research or policy and to identify any evidence gaps.

      Methods

      A preliminary search for registered scoping or systematic reviews or protocols was conducted on the Joanna Briggs Institute (JBI) Evidence Synthesis database, the Cochrane Database of Systematic Reviews and the Prospero database, of which none were found. A scoping review was chosen as the most suitable method for several reasons. The extent to which complexity science has been used to inform AMS has not previously been explored. Furthermore, scoping reviews are a suitable method to identify if certain concepts are used, not just in research publications, but in policy or practice. Inclusion of such policy documents and sources are usually outside of systematic review parameters. Finally, as no systematic review was found on the use of complexity theory in AMS, a scoping review would determine if such a review is warranted in the future. The review was conducted according to scoping review guidelines from JBI. [

      Peters MDJ, Godfrey C, McInerney P, et al. Chapter 11: Scoping reviews (2020 Version). 2020. In: Joanna Briggs Institute Reviewer's Manual [Internet]. The Joanna Briggs Institute. Available from: https://reviewersmanual.joannabriggs.org/.

      ]

      Protocol

      An a-priori protocol (Supplementary material) was developed between study authors in accordance with the JBI scoping review manual. [

      Peters MDJ, Godfrey C, McInerney P, et al. Chapter 11: Scoping reviews (2020 Version). 2020. In: Joanna Briggs Institute Reviewer's Manual [Internet]. The Joanna Briggs Institute. Available from: https://reviewersmanual.joannabriggs.org/.

      ]

      Inclusion criteria

      Participants

      AMS research or policy involving clinicians or patients and the public of all ages, in any healthcare setting, as well as those involved in the management of healthcare delivery.

      Concept

      Explicit use of complexity theory to conceptualise, design, implement or analyse AMS in human healthcare. The framework described in Supplementary File (Table I) was used to conceptualize complexity for this study.

      Context

      Healthcare settings such as acute hospitals or primary care centres in any geographical or economic setting were included. Larger healthcare contexts, such as governmental organisations, where complexity theory was cited in a published report, guideline or policy document were also included.

      Evidence sources

      Primary research(for example quantitative, qualitative or mixed methods) were included as well as secondary research such as literature reviews from peer-reviewed journals. Policy or policy-related reports from healthcare organisations concerned with AMR or AMS were also included.

      Search strategy

      • 1.
        Two databases were searched initially (Embase and Medline). Keywords and index terms from relevant publications in this search were used to build the search strategy across all databases. A medical subject librarian constructed the search strategy.
      • 2.
        Research databases searched: Cinahl, Cochrane Library, Embase, Medline, National Institute for Health and Care Excellence, PsycInfo, Scopus and Web of Science from the date of archive inception to June 2020.
      • 3.
        Grey literature and other databases searched: EVIPNet, Google, Mednar, Proquest Theses, and the WHO library of national AMR action plans.
      • 4.
        Additional searching of reference lists of included publications.
      • 5.
        Authors of included studies were contacted to enquire if they are aware of any relevant additional or unpublished data.
      The search strategy is described in detail in Supplementary File (Table II).

      Data extraction, analysis and presentation

      Articles returned from searches were exported to Microsoft Excel ® where titles and abstracts were independently screened by two authors. Articles that met the inclusion criteria underwent a full text review. A third reviewer was available to resolve disagreements where necessary, through discussion with the other authors. Data were charted on summary tables in Microsoft Word ® and reported descriptively.

      Results

      Database searches yielded 612 initial records. After removing duplicates, titles and abstracts or executive summaries of 561 publications were screened, of which 528 were excluded. Full text review was conducted on 33 articles, of which 25 were excluded, leaving 8 articles included in the review [
      • Sturmberg J.
      Systems and complexity thinking in general practice. Part 2: application in primary care research.
      ,

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ,
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ,
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ,
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ,
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ,
      • Adamu A.A.
      • Gadanya M.A.
      • Jalo R.I.
      • et al.
      Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
      ,
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ] (Figure 1).
      Figure 1
      Figure 1PRISMA Flow Chart, adapted from Page et al 2021. Automation tools were not used to assist data collection or analysis; AMS = antimicrobial stewardship

      Study characteristics

      Included studies were heterogenous in terms of study design, healthcare context and geographical location. There were three interventional studies, [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ,
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ,
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ] one cross-sectional survey, [
      • Adamu A.A.
      • Gadanya M.A.
      • Jalo R.I.
      • et al.
      Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
      ] two case studies, [
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ,
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ] one literature review [
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ] and one short article describing a case vignette. [
      • Sturmberg J.
      Systems and complexity thinking in general practice. Part 2: application in primary care research.
      ] The contexts of these studies were varied: primary care, [
      • Sturmberg J.
      Systems and complexity thinking in general practice. Part 2: application in primary care research.
      ,
      • Adamu A.A.
      • Gadanya M.A.
      • Jalo R.I.
      • et al.
      Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
      ] acute care, [
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ,
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ] ambulatory care, [
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ] long term care [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ] and the remaining two from regional or national policy perspectives. [
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ,
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ] Geographical locations were also diverse with two studies from western Europe, [
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ,
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ] one from the US, [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ] two from Africa, [
      • Adamu A.A.
      • Gadanya M.A.
      • Jalo R.I.
      • et al.
      Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
      ,
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ] one from China, [
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ] one focusing on low/middle income countries (LMICs) [
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ] and one not specified. [
      • Sturmberg J.
      Systems and complexity thinking in general practice. Part 2: application in primary care research.
      ] Using the complexity theory characteristics previously described by Plsek and Greenhalgh [
      • Plsek P.E.
      • Greenhalgh T.
      Complexity science: The challenge of complexity in health care.
      ] (Supplementary File Table I), included articles were searched for incorporation of these characteristics to the studies. A summary of the findings is available in the Table I.
      Table IUse of complexity theory in included articles
      AuthorElements of complexity theory discussed
      Sturmberg [
      • Sturmberg J.
      Systems and complexity thinking in general practice. Part 2: application in primary care research.
      ]
      Identification of internalised patient and doctor rules which drive antibiotic use for sore throat

      Non-linearity identified through reinforcing feedback loops which drive doctor and patient behaviour

      Paradoxically, doctors may choose antibiotic prescription for sore throat (although not evidence based) as there is less workload involved than avoiding prescription
      Lanham et al [
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ]
      Self-organisation was intentionally supported and encouraged in the original RCT

      The mobile phone intervention supported existing interactional behaviour to produce new emergent behaviours

      Intervention effects spread outside the intervention group to the control group which was an unexpected occurrence
      Zimmerman et al [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ]
      Unexpected behaviour occurred in variable adoption of intervention components across study settings

      Physicians were identified as behaviour attractors in nursing homes
      McLellan et al [
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ]
      Feedback workshops facilitated optimal antimicrobial prescribing behaviour by junior doctors

      Proposition that junior doctors could foster prudent antimicrobial prescribing in hospital settings

      Prescribing behaviour emerges from interactions between junior doctors’ individual (e.g. knowledge) and social (e.g. workplace culture) variables
      Merrett et al [
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ]
      Identification of influencing factors, individuals and organisations which drive AMR and the interactions between these elements

      Fluid boundaries, such as those between public and private healthcare sectors, facilitates access to antibiotics

      Continuous evaluation of health systems required to observe intervention impact and identify unexpected consequences

      Tension, for example between mass antimicrobial administration campaigns and the potential for development of AMR

      Potential for inappropriate antimicrobial use to become the normal pattern within a healthcare system
      Wang et al [
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ]
      Mapped the emergent behaviour of multiple actors within the healthcare system as adaptive responses to antibiotic regulation

      Described overall pattern of antibiotic prescription and consumption based on this adaptive behaviour

      Highlighted unexpected outcomes from antibiotic regulation and policy such as reduced impact of regulation to decrease overall antibiotic consumption

      Tension between regulators ensuring financial health of hospital systems but also controlling antibiotic use

      Internalised rules held by patients and prescribers drive inappropriate antibiotic use

      Fluid boundaries between antibiotic access routes
      Cunney et al [
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ]
      Leveraged attractors within the hospital system to co-design intervention

      Preference of participants for written vs electronic feedback, rejection of reminder cards attached to reference material and rebuffing education opportunities were unexpected occurrences

      Participants identified simple rules to achieve study objectives

      Emergent behaviour occurred when junior doctors exiting their rotation informed incoming junior doctors of the principles of prudent antimicrobial prescribing, which sustained the intervention
      Adamu et al [
      • Adamu A.A.
      • Gadanya M.A.
      • Jalo R.I.
      • et al.
      Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
      ]
      Attractors/influencing factors on antibiotic consumption identified through causal loop diagrams
      AMR: antimicrobial resistance; RCT: randomised controlled trial

      Study findings

      The interventional studies used complexity as a guiding theory and included study participants as a co-design strategy. The non-interventional studies used complexity theory to propose potential interventions to guide antimicrobial prescribing behaviour change (Table II). However, there was variation in the extent to which complexity theory was operationalised in the studies. For example, Merrett et al [
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ] and Wang et al [
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ] refer to CAS in general terms and integrate it to their studies in a relatively unstructured format. Conversely, Adamu et al specifically used complexity theory as a framework to analyse their data. This approach enabled the construction of causal loop diagrams to illustrate feedback mechanisms within their system of antimicrobial use in primary care. The heterogenous nature of the studies proposed a challenge to perform synthesis on the data. However, two specific themes were notable: the importance of considered communication in the conduct of AMS and the utility of mixed methods as an AMS research strategy.
      Table IISummary of included articles
      AuthorSettingAim(s)Study population/sample sizeMethodsInterventionOutcomes and key findings
      Sturmberg et al [
      • Sturmberg J.
      Systems and complexity thinking in general practice. Part 2: application in primary care research.
      ]
      Primary careTo gain contextual understanding of known problems in primary careN/ANarrative conceptualization of a sore throat vignette through CASN/AHighlighted decision-making processes in prescribing antimicrobials for sore throat.

      Complexity science is a useful tool to inform AMS in primary care
      Lanham et al [
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ]
      Ambulatory care in KenyaTo examine the role of self-organisation in the scale up and spread of an antiretroviral adherence intervention538 ambulatory care HIV patientsRe-analysis of a previously published RCTN/AImportance of integrating intervention with local organisational infrastructure

      Close contact between investigators and participants key for intervention adoption

      Interventions are shaped by their environments and outcomes may spread outside study population
      Zimmerman et al [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ]
      US LTCSTo optimise antimicrobial prescribing in LTCSHealthcare professionals, residents/ resident familiesQI methodologyAntibiotic prescriber, resident/resident family education

      Communication form for healthcare staff to report infection

      Feedback to stakeholders
      Suboptimal antimicrobial use decreased in nursing homes, to a lesser extent in residential care

      Resident/family education did not result in change

      Use of CAS provided observations and guidance for further QI projects
      McLellan et al [
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ]
      UK acute care hospitalTo investigate if providing feedback to junior doctors optimized antimicrobial prescribing35 junior doctorsMixed methods nested in an RCTPharmacist-led antimicrobial prescribing feedback to doctorsLower suboptimal prescribing in intervention group

      Knowledge and awareness of suboptimal antimicrobial prescribing important to drive appropriate prescribing habits

      Mechanism for change suggested by placing junior doctors as positive influencers of antimicrobial prescribing
      Merrett et al [
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ]
      Governmental/policy level of LMICsIdentify interventions to optimize antimicrobial accessCitizens of LMICsLiterature reviewN/AInterventions need to account for the complex system in which antibiotic use occurs

      Synergies between multiple interventions (e.g. access to diagnostics, ensuring drug quality) are needed
      Wang et al [
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ]
      Governmental/policy level in ChinaTo investigate the implementation of regulations and strategies to control antimicrobial use in ChinaChinese citizensCase studyN/AReview identified routes to reverse the unexpected rise in Chinese antimicrobial use, despite regulations

      Heterogeneity of actors in the system need to be accounted for Complicated incentive schemes should be simplified
      Cunney et al [
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ]
      Irish paediatric hospitalTo improve documentation and compliance with local antimicrobial prescribing policyEmergency department doctorsQI methodologyFeedback sessions and plan, do, study, act cyclesMaintained 100% compliance rate with agent choice and documentation at 18 months follow up Antimicrobial consumption decreased

      Improvement in antimicrobial use quality indicator measurements

      Participant co-design helped foster frontline ownership of the project

      Goal setting and action planning identified as key components
      Adamu et al [
      • Adamu A.A.
      • Gadanya M.A.
      • Jalo R.I.
      • et al.
      Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
      ]
      Primary care in NigeriaTo describe the volume of non-prescription antibiotic sales and associated behavioural factors453 ‘medication retailers’Cross-sectional surveyN/AConstruction of causal loop diagrams to explain behavioural factors

      66.67% of participants sold antibiotics without a prescription

      Provider training decreased the likelihood of selling antibiotics without a prescription
      AMS: antimicrobial stewardship; CAS: complex adaptive system; LMIC: low/middle income country; LTCS: long term care setting; QI: quality improvement; RCT: randomized controlled trial; UK: United Kingdom; US: United States

      Importance of tailored communication strategies in AMS

      Sturmberg argued that merely educating doctors on prudent prescribing practices does not address other issues such as patient beliefs and perceptions of a satisfactory consultation. Instead, a focus on this communication is required to break the reinforcement of antibiotic prescribing. [
      • Sturmberg J.
      Systems and complexity thinking in general practice. Part 2: application in primary care research.
      ] This was exemplified by other studies where, for example, providing clinicians with prescribing feedback led to improvements in antimicrobial prescribing quality. [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ,
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ,
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ] Cunney et al [
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ] found that feedback was best delivered to junior doctors during scheduled ward rounds, while McLellan et al [
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ] provided additional time and space for participants to reflect on their prescribing practice.
      As part of considerate communication approaches, Zimmerman et al [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ] and Lanham et al identified the importance of close communication and regular exchanges between those seeking to optimise antimicrobial use, and the study population . [
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ] For example, when Zimmerman et al [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ] realised that healthcare providers were not engaging with online education sessions, they quickly switched to face-to-face sessions to increase impact and a team leader was nominated within the study population.

      Use of mixed methods

      The QI studies used mixed methods approaches to study antimicrobial prescribing and modify their interventions in real time, based on participant feedback. [

      Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

      ,
      • Cunney R.
      • Kirrane-Scott M.
      • Rafferty A.
      • et al.
      Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
      ,
      • Lanham H.J.
      • Leykum L.K.
      • Taylor B.S.
      • et al.
      How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
      ] McLellan et al used a more rigid approach of qualitative methods nested within an RCT to provide antimicrobial prescribing feedback to junior doctors and generate an interventional theory to inform future studies. [
      • McLellan L.
      • Dornan T.
      • Newton P.
      • et al.
      Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
      ] .

      Discussion

      In an explicit way, complexity theory has not been extensively used to inform AMS in human healthcare, despite previous calls for complexity-driven approaches to healthcare interventions. [
      • Plsek P.E.
      • Greenhalgh T.
      Complexity science: The challenge of complexity in health care.
      ,
      • Greenhalgh T.
      • Papoutsi C.
      Studying complexity in health services research: desperately seeking an overdue paradigm shift.
      ] The relatively low number of articles included in this review and their heterogenous nature suggests a fragmented and disparate application of complexity theory to AMS. Indeed, this posed a challenge to synthesise the evidence for this review. However, this could also be considered as a strength of complexity theory, in that it has been integrated as a conceptual framework in a variety of research methods and contexts. Papoutsi and Greenhalgh have previously highlighted this point, that complexity offers “…a flexible and emergent approach…” to conducting healthcare research. [
      • Greenhalgh T.
      • Papoutsi C.
      Studying complexity in health services research: desperately seeking an overdue paradigm shift.
      ]
      One of the potential challenges of capturing the integration of complexity theory and AMS is how this is reported in the literature. In our review, Adamu et al, [
      • Adamu A.A.
      • Gadanya M.A.
      • Jalo R.I.
      • et al.
      Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
      ] Merrett et al [
      • Merrett G.L.
      • Bloom G.
      • Wilkinson A.
      • et al.
      Towards the just and sustainable use of antibiotics.
      ] and Wang et al [
      • Wang L.
      • Zhang X.
      • Liang X.
      • et al.
      Addressing antimicrobial resistance in China: policy implementation in a complex context.
      ] discussed the unexpected impact of financial incentives and disincentives on rational antimicrobial prescribing. However, this has previously been termed a “squeezing the balloon” effect, where introducing a restrictive measure may result in an adaptive, compensatory response. [
      • Burke J.P.
      Antibiotic resistance-squeezing the balloon?.
      ] This effect has recently been reported in aUK primary care, where a financial incentive resulted in a sustained reduction in antimicrobial prescribing for uncomplicated respiratory tract infections, but with an unpredictable reduction in appropriate prescribing for lower respiratory tract infections. [
      • Bou-Antoun S.
      • Costelloe C.
      • Honeyford K.
      • et al.
      Age-related decline in antibiotic prescribing for uncomplicated respiratory tract infections in primary care in England following the introduction of a national financial incentive (the Quality Premium) for health commissioners to reduce use of antibiotics in the community: an interrupted time series analysis.
      ] These two studies describe one facet of complexity theory (unexpected occurrences), but neither of these studies specifically reference complexity. Therefore, the apparent lack of complexity theory integration to AMS research as an informative framework to AMS may be due to semantics. This is likely also true of policy and policy-related publications. A 2014 Department of Health (England) publication on factors influencing AMR describes multifactorial problems as “messy, complex situations” [] An detailed systems map in this report elegantly describes the myriad influences on antimicrobial prescribing in the style of a causal loop diagram but, again, without specifically mentioning complexity as an overarching theory.
      The recently updated Medical Research Council guidelines on developing and evaluating complex interventions advocates considering intervention development in a system context, where the intervention itself interacts with the local environment an emergent, unpredictable fashion. As AMS is, by definition from the MRC, a complex intervention, frequently deployed in complex settings such as acute hospitals, future AMS endeavours will likely benefit from using complexity science as a guiding theory. [
      • Shahsavari H.
      • Matourypour P.
      • Ghiyasvandian S.
      • et al.
      Medical Research Council framework for development and evaluation of complex interventions: A comprehensive guidance.
      ]

      Strengths and limitations

      This review adds to the literature on theory-informed AMS research. The search strategy widened the explorative scope in comparison to previous reviews. Although the number of included articles was relatively small, they transcended across different contexts and settings and provided recommendations for AMS interventions applicable to these settings. In keeping with scoping review methodology, this review did not appraise the quality of the included articles. However, it is clear from the limited and heterogenous publications available, a systematic review on the application of complexity theory to AMS in human healthcare is not currently warranted.

      Conclusions and future work

      Antimicrobial prescribing and consumption behaviours are part of an overall complex network of behaviours within healthcare settings. While there is extant literature on this subject, little has been conducted from a complexity theory perspective.
      Understanding the drivers and reinforcements of these behaviours is important for healthcare systems to foster cultures of prudent antimicrobial use. Complexity theory is a practical and useful way to conceptualise and design AMS interventions in human healthcare. It is unclear whether those concerned with addressing AMR are unaware of complexity as a practical theory to inform AMS, whether they have not embraced it as an informative concept or whether aspects of complexity are being utilized but just not explicitly. Future research on the design, implementation and evaluation of AMS interventions in healthcare should consider complexity as an informative theory to guide study designs. Equally, policy makers and regulators concerned with prudent antimicrobial use should consider complexity in the administration and monitoring of their programmes.

      Funding

      This research did not receive any specific grant funding from agencies in the public, commercial or not for profit sectors.

      Uncited reference

      [
      • Page M.J.
      • Moher D.
      • Bossuyt P.M.
      • et al.
      PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews.
      ].

      Conflicts of Interest

      None to declare

      Appendix A. Supplementary data

      References

      1. The Health Foundation. Evidence scan: Complex adaptive systems. 2010. Available from: https://www.health.org.uk/publications/complex-adaptive-systems#:∼:text=This%20evidence%20scan%20collates%20more,or%20development%20in%20this%20area.

        • Plsek P.E.
        • Greenhalgh T.
        Complexity science: The challenge of complexity in health care.
        BMJ. 2001; 323: 625-628
        • Peters D.H.
        The application of systems thinking in health: why use systems thinking?.
        Health Research Policy and Systems. 2014; 12
        • Holmes A.H.
        • Moore L.S.
        • Sundsfjord A.
        • et al.
        Understanding the mechanisms and drivers of antimicrobial resistance.
        Lancet. 2016; 387: 176-187
        • Dellit T.H.
        • Owens R.C.
        • McGowan J.E.J.
        • et al.
        Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship.
        Clin Infect Dis. 2007; 44: 159-177
      2. Health Protection Surveillance Centre. Guidelines for antimicrobial stewardship in hospitals in Ireland. 2009. Available from: https://www.hpsc.ie/a-z/microbiologyantimicrobialresistance/infectioncontrolandhai/guidelines/File,4116,en.pdf.

        • Thompson D.S.
        • Fazio X.
        • Kustra E.
        • et al.
        Scoping review of complexity theory in health services research.
        BMC Health Serv Res. 2016; 16
        • Talkhan H.
        • Stewart D.
        • Mcintosh T.
        • et al.
        The use of theory in the development and evaluation of behaviour change interventions to improve antimicrobial prescribing: a systematic review.
        J Antimicrob Chemother. 2020; 75: 2394-2410
      3. Peters MDJ, Godfrey C, McInerney P, et al. Chapter 11: Scoping reviews (2020 Version). 2020. In: Joanna Briggs Institute Reviewer's Manual [Internet]. The Joanna Briggs Institute. Available from: https://reviewersmanual.joannabriggs.org/.

        • Sturmberg J.
        Systems and complexity thinking in general practice. Part 2: application in primary care research.
        Aust Fam Physician. 2007; 36 4: 273-275
      4. Strategies to reduce potentially inappropriate antibiotic prescribing in assisted living and nursing homes [Internet]. 2014 [cited 20/02/2020]. Available from: https://www.ahrq.gov/hai/patient-safety-resources/advances-in-hai/hai-article8.html.

        • McLellan L.
        • Dornan T.
        • Newton P.
        • et al.
        Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials.
        J Antimicrob Chemother. 2016; 71: 1415-1425
        • Merrett G.L.
        • Bloom G.
        • Wilkinson A.
        • et al.
        Towards the just and sustainable use of antibiotics.
        J Pharm Policy Pract. 2016; 9
        • Wang L.
        • Zhang X.
        • Liang X.
        • et al.
        Addressing antimicrobial resistance in China: policy implementation in a complex context.
        Global Health. 2016; 12
        • Cunney R.
        • Kirrane-Scott M.
        • Rafferty A.
        • et al.
        Start smart': using front-line ownership to improve the quality of empiric antibiotic prescribing in a paediatric hospital.
        BMJ Open Qual. 2019; 8e000445
        • Adamu A.A.
        • Gadanya M.A.
        • Jalo R.I.
        • et al.
        Factors influencing non-prescription sales of antibiotics among patent and proprietary medicine vendors in Kano, Nigeria: a cross-sectional study.
        Health Policy Plan. 2020; 35: 819-828
        • Lanham H.J.
        • Leykum L.K.
        • Taylor B.S.
        • et al.
        How complexity science can inform scale-up and spread in health care: Understanding the role of self-organization in variation across local contexts.
        Soc Sci Med. 2013; 93: 194-202
        • Page M.J.
        • Moher D.
        • Bossuyt P.M.
        • et al.
        PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews.
        BMJ. 2021; 372
        • Greenhalgh T.
        • Papoutsi C.
        Studying complexity in health services research: desperately seeking an overdue paradigm shift.
        BMC Med. 2018; 16
        • Burke J.P.
        Antibiotic resistance-squeezing the balloon?.
        JAMA. 1998; 280: 1270-1271
        • Bou-Antoun S.
        • Costelloe C.
        • Honeyford K.
        • et al.
        Age-related decline in antibiotic prescribing for uncomplicated respiratory tract infections in primary care in England following the introduction of a national financial incentive (the Quality Premium) for health commissioners to reduce use of antibiotics in the community: an interrupted time series analysis.
        J Antimicrob Chemother. 2018; 73: 2883-2892
      5. UK Department of Health. Antimicrobial resistance (AMR) systems map. 2014. Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/387746/Microbial_Maps.pdf.

        • Shahsavari H.
        • Matourypour P.
        • Ghiyasvandian S.
        • et al.
        Medical Research Council framework for development and evaluation of complex interventions: A comprehensive guidance.
        J Educ Health Promot. 2020; 9