Structured antimicrobial stewardship programs (AMSP) should be implemented in all ICUs.

AMSP in the critical care setting should be led by an expert ICU staff member and implemented by a collaborative and multidisciplinary team including, infectologists, microbiologists, pharmacists, preventive medicine specialists and other specialists as needed.

Specialists of preventive medicine will form part of AMSP, for the coordination of outbreak control actions and other preventive measures to implant (patient hygiene, asepsis and antisepsis measures in taking samples, etc.) for invasive procedures device monitoring, for surveillance and monitoring of surgical prophylaxis, and for control of environmental biosafety in controlled environment rooms in order to minimize infection transmission and prevent the appearance of outbreaks of environmental origin.

The design and implementation of AMSP in critically ill patients should consider the specificities of this patient population. Infection severity, organ dysfunctions and changes in pharmacokinetics should be taken into account. Therapeutic drug monitoring, if available, is helpful for dose adjustments in ICU patients.

Collaborative, non-compulsory interventions including audit and feedback should be prioritized. De-escalation and shortening of treatments should be the main targets.

AMSP in critically ill patients should be fueled by real-time microbiological information and clinical microbiologists support.

Updated knowledge of the local epidemiology is mandatory to design the antibiotic policy.

AMSP should be developed by a multidisciplinary team including haematologists, infectious diseases experts, microbiologists, preventive medicine specialists and pharmacists.

Patterns of antimicrobial prescription and antimicrobial resistance should be monitored periodically.

The implementation of screening samples to detect colonization by multidrug-resistant organisms (MDRO) may be considered, at least in patients at highest risk.

Antimicrobial prophylaxis, empirical and streamlined therapy for febrile neutropenia should be protocolized considering local epidemiology and different patient risk profiles. Adherence to recommendations should be measured.

On-demand availability of infectious diseases experts, clinical microbiologists and pharmacists should be warranted, at least for the most complex cases, and also specialists in preventive medicine in order to control outbreaks or isolation.

Periodical, regular, multidisciplinary rounds are recommended.

Specific local guidelines for the treatment of the most frequent infectious syndromes in patients with solid tumors (not only febrile neutropenia) based on local epidemiology, must be developed by the AMSP team and oncologists. Implementation of clinical pathways is more complex but also useful.

The AMSP should include availability of infectious diseases specialists for consultation.

Continuous active training on appropriate antimicrobial use must be implemented.

Interventions specifically targeting febrile neutropenia, antifungal stewardship or biomarker-based algorithms may have a limited impact on patients with solid tumors and should not be prioritized.

Research is needed regarding optimal duration of antimicrobial treatments or the clinical impact of rapid diagnostic tests in this setting.

To develop local protocols with criteria helping in the identification of patients who will most probably benefit from antibiotics (i.e. improvement in survival or symptom relief while maintaining the quality of life) and those in which avoiding futile treatments, developed by palliative care experts and the AMSP team. The antibiotics and routes to be used in the most frequent syndromes and rules for early stopping should be included.

A multimodal approach targeting physicians, patients and care-givers’ perceptions is also recommended. Availability of consultation to palliative care and infectious diseases specialists is recommended.

Research is needed about the optimal approach for developing and implementing AMSP in palliative care.

The AMSP team should include an expert paediatrician in infectious diseases and antimicrobial use for the activities in paediatric population.

Overall, the principles of AMSP in adults are applicable to children; however, two aspects must be considered: (a) local protocols for antimicrobial use in children should not be simple dose-adjusted transcriptions of protocols for adults but be specifically developed, and (b) evidence for some interventions is more limited in children than in adults.

The fact that viral infections are more frequent as cause of acute respiratory infections in children should be outlined, and availability of appropriate diagnostic methods must be considered in order to help avoiding the unnecessary use of antibiotics.

For monitoring antimicrobial consumption, days of therapy (DOT) per 1000 patient-days is the preferred indicator.

Multidisciplinary antimicrobial stewardship team with knowledge of the specific barriers for implementation of AMSP in the local ED must be formed.

Local empirical treatment guidelines with input from local epidemiology must be available for ED faculty.

Appropriate training for the basic aspects of infectious diseases diagnosis and treatment must be provided.

Specific appropriate clinical outcome indicators and evaluations of quality of prescriptions must be measured.

The coordination activity of preventive medicine specialists with primary care doctors or community centres or residences, responsible for the continuity of antibiotic treatments, will be essential. This activity will make it possible to change the antibiotic when the patient has attended the ED where he has received the antibiotic of choice and has gone home.

If individualized data on antibiotic prescriptions are not automatically available, pooled data weighted by the number of patients attended at the department (daily defined doses [DDD] per 100 patients admitted in the observation area, DDD per 100 patients discharged from de ED) may be considered.

Increased adherence to CAP guidelines must be an objective of AMSP. Indicators for the measurement to adherence of the principles of treatment of CAP must be implemented.

Empirical therapy in CAP must be locally protocolized according to guidelines, considering patients’ severity and local epidemiology.

Adherence to early antibiotic de-escalation, switch to oral therapy and short courses of therapy must be objectives of AMSP. Barriers for the implementation of these two strategies may be overcome by appropriate educational interventions.

Research regarding the effectiveness of rapid diagnostic molecular tests as a stewardship tool in CAP is needed; however, these techniques should be considered in advanced AMSP with evaluation of results.

The use of procalcitonin is useful to improve antibiotic use in patients with CAP, and may be incorporated as an aid tool together with appropriate clinical workout.

Empirical antimicrobial therapy in UTI should be based on local guidelines. These guidelines should take into account local bacterial resistance rates, individual risk and severity of infections.

Antimicrobial stewardship teams should provide advice for optimizing directed therapy in UTI caused by MDRO.

Interventions to avoid the inappropriate use of antibacterials in asymptomatic bacteriuria must be implemented.

Interventions to improve adherence to short treatments according to specific type of infection and patients characteristics must be implemented.

Audits with feedback in selected patients or targeted microorganisms may be considered.

Promote the appropriate diagnosis and management of VCA-BSI by providing adequate training and adherence to guidelines.

Specifically, appropriate duration of therapy of VCA-BSI should be promoted.

Bacteraemia programs, in which unsolicited advice and follow-up is provided for all patients with bacteraemia (or at least for high risk patients and pathogens) should be implemented; measurement of adherence to quality-of-care indicators is recommended.

Local adaptations of guidelines for SAP in all procedures performed in the hospital, agreed with surgeons and anaesthesiologists, should be available. The local guidelines must include the antibiotics of choice and alternatives, as well as the appropriate timing, dosing and duration of antibiotic prophylaxis. It is important to enhance other prophylactic measures and their implementation carried out by preventive medicine specialists.

Training about the principles of SAP should be provided to all staff involved.

Regular audits about adherence to local guidelines using standardized indicators must be performed, and feedback provided.

Specific interventions to improve SAP must be considered according to local resources and problems detected.

The different routes for antimicrobial administration, and the different modes of intravenous perfusions should be considered in local protocols; AMSP must include activities aiming at providing the most appropriate way to administer antibiotics to the patients.

Timely switch to oral therapy when indicated must be an objective for AMSP, and activities encouraging this practice should be implemented; they should include actions to avoid longer treatments once the oral route is started.

Infusion pumps facilitate the correct administration of intravenous antimicrobials and must be used in circumstances in which extended or continuous perfusions are considered critical for the patient. Handling of these devices must be carried out by trained personnel.

Implication of nurses in AMSP is key in order to assure an appropriate administration of antibiotics.

AMSP should promote the inclusion of new drugs in hospital formularies according to local needs and best evidence, in order to provide patients with the best available therapy in infections caused by MDRO.

AMSP should include local consensus guidelines including specific indications of use of new drugs according to the best available evidence and local epidemiology.

AMSP should implement non-compulsory interventions as auditing and feedback to facilitate the adherence to the established guidelines.

AMSP should monitor and evaluate the data concerning the outcomes of patients treated with new drugs, development of resistance to them and their impact in the overall epidemiology of MDRO.

Antimicrobial stewardship interventions should be designed considering knowledge from behavioural and social sciences and should employ behaviour change techniques. Whenever possible, behaviour experts should be part of antimicrobial stewardship teams.

The prescribing behaviours that are intended to be changed should be precisely defined, and barriers and promoters should be analyzed locally.

Behaviour change techniques (BCTs) should be selected taking into account the main determinants of the desired and/or undesired behaviours. The interventions must combine several BCTs.

There are several behaviour's change theories and models that can be applied in the field of AMS. One of the most frequently adopted models for AMS is the behaviour change wheel.

It is advisable to conduct process evaluations of ongoing interventions in order to learn why they fail or succeed.

Whenever possible and feasible, e-tools should be included in AMS interventions to facilitate guidelines, clinical pathways and post prescription review.

The efficacy and safety of AMS e-tools should be validated through cluster randomized control trials or adequately controlled quasi-experimental designs.

AMS e-tools should be available for prescribers. Social media may be considered as tools to disseminate the AMSP resources.

Diagnostic stewardship principles should be part of any AMSP.

The adequacy of laboratory test requests and correct handling of specimens should be promoted among the activities of the AMSP.

Rapid microbiological diagnostic techniques for the adequate samples and patients, and rapid identification and susceptibility testing from positive blood cultures should be implemented and associated with rapid active reporting and antimicrobial stewardship interventions.

The use of rapid viral diagnostic testing for respiratory pathogen should be promoted in the appropriate settings.

Non-culture-based fungal markers in patients with haematological malignancies at risk of invasive fungal infections must be implemented and associated with appropriate interpretation criteria to both improve diagnosis and facilitate adequate use of antifungals.

Fast-track protocols for selected patients (critical or immunocompromised patients) must be designed and implemented.

Provide cascade or selective reporting instead of over-reporting all tested antimicrobials.

Provide periodical data on cumulative antimicrobial susceptibility for optimizing empirical therapy, with appropriate stratification of data whenever possible.

Surveillance programs of multidrug-resistant pathogens for the detection of outbreaks and emerging new clones must be implemented.