Over recent decades, the progressive increase in antimicrobial resistance has had a strong impact on healthcare systems worldwide.1 Associations have been found between infections caused by multidrug-resistant microorganisms and increased mortality, increased hospital stay, and significant increases in healthcare costs.2,3 Excessive and/or inappropriate use of antimicrobials has been found to contribute to the emergence, rapid spread, and perpetuation of these multidrug-resistant strains.4 Thus, in recent years, antimicrobial stewardship programmes (ASPs) have been developed to optimise their use while minimising the spread and number of infections caused by multidrug-resistant bacteria.5 The implementation of these programmes has been recognised as a priority by health administrations and scientific societies.5,6

To date, most of the reported experiences of these programmes have focused on hospitalised patients, particularly critical patients, and more recently, on outpatient settings.7,8 Emergency departments (EDs) are one of the most relevant services for implementing ASPs. These departments are where the first doses of antibiotics are prescribed to patients who are to be hospitalised and to those returning to primary care; they are also where large numbers of antibiotics are prescribed to patients discharged directly to their homes or other healthcare centres. Several studies have reported significant increases in the number of infections caused by multidrug-resistant bacteria in EDs.9,10 Although guidelines on ASPs recognise that EDs are preferential sites for their implementation, the participation of multidisciplinary teams in EDs is still limited.11 On the other hand, there is a lack of uniformity in the type of indicators used to monitor the use of antimicrobials in this setting,12,13 which hinders the implementation of ASPs. However, the indicators typically associated with ASPs may not be directly applicable to their implementation in EDs as they primarily focus on inpatient management.

The aim of this study was to create a panel of indicators that could be used to monitor the correct utilisation of antimicrobial agents in EDs. An expert panel employed a modified Delphi methodology to achieve consensus during the development process.

The study was designed following a modified Delphi methodology. Initially, a coordination committee comprising 4 members from the Spanish Society of Hospital Pharmacy (SEFH) and the Spanish Society of Emergency Medicine (SEMES) utilised their knowledge, experience, and a literature review to propose a set of indicators. These indicators took into account criteria of evidence, intervention outcomes, ease of implementation (resource requirements), and priority level (essential or advanced). Subsequently, an evaluation panel was formed comprising 20 Spanish experts with proven experience in the use of antimicrobials and ASPs. The panel included 5 specialists with experience in EDs, 5 specialists in infectious diseases, 5 pharmacists with more than 3 years' experience in EDs, and 5 other specialists in ASPs from hospital pharmacy, intensive care medicine, and microbiology.

The proposed indicators were evaluated using a RAND/UCLA Appropriateness Method combining the Delphi technique with the Nominal Group Technique. This approach involves 2 evaluation rounds: in the first round each panel member makes an evaluation; in the second round the evaluations are pooled.14 In the first round, the members of the evaluation panel were e-mailed a document containing the list of indicators, a basic description of the indicators, and relevant bibliographical references. The experts could comment on each indicator and propose new indicators.

Each panel member rated the relevance of each of the proposed indicators on 2 dimensions (care outcomes and ease of implementation) and 2 attributes (priority level and recommended measurement frequency for each indicator [quarterly, half-yearly, or annual]). The 2 dimensions were scored on a scale ranging from 0 (complete disagreement) to 10 (complete agreement). The priority level was scored using a scale ranging from 1 (high) to 3 (low).

The coordinating committee evaluated the panel members' scores and comments and prepared a second questionnaire which retained, modified, or eliminated the indicators. Regarding care outcome and ease of implementation, each indicator was classified as appropriate (median score >6), inappropriate (median score <4), or inconclusive (median score 4–6). The criterion for excluding an indicator from moving to the second round was that it was classified as inappropriate according to its median score on care outcome or ease of implementation. Regarding the priority level, each indicator was classified as high (median = 1), medium (median = 1.5–2), or low (median = 2.5–3) priority.

In the second round, the experts were sent the modified questionnaire, which included the median scores and ranges from the first round, the anonymous comments, and an analysis of the results for each indicator to aid them in the next evaluation round. The experts reassessed the indicators and their priority level. The purpose of this round was to give the experts the opportunity to review their own assessments in the light of the other experts' assessments. They were also asked to assess the new indicators that had been proposed in the first round. These new indicators were scored in the same way as in the first round.

Fig. 1 shows the study flow chart with the overall results obtained in each round. The coordinating group put forward an initial list of 61 potential indicators to the evaluation panel. These indicators were divided into 4 blocks: consumption indicators, microbiological indicators, process indicators, and outcome indicators.

Fig. 1.

Study flowchart.

(0,3MB).

After analysing the scores and comments, 31 indicators were classified as high priority, 25 as medium priority, and 5 as low priority. None of the initially proposed indicators were excluded from the second round, because they all scored more than 6 points on care outcome and ease of implementation. In addition, 19 new potential indicators were proposed to monitor these activities. The evaluators' anonymised opinions were included in the document mailed out in the second round.

After the second round, 18 of the 19 new indicators were accepted. Of the 79 final indicators, 42 (53.2%) were scored as high priority, 28 (35.4%) as medium priority, and 9 (11.4%) as low priority. The indicators were grouped into 4 blocks: 50 (63.3%) antimicrobial use indicators, 7 (8.9%) microbiological indicators, 13 (16.5%) process indicators, and 9 (11.4%) outcome indicators. To ensure a uniform measurement frequency, we calculated the median frequency according to the first and second rounds for each group of indicators and then assigned this frequency to all the indicators in the corresponding group. Tables 1–3 show the final indicators and their measurement frequencies.

Antimicrobial stewardship programs have proven to be essential tools for improving the use of antimicrobials in hospital settings and are now perceived as a growing need within healthcare teams.6,15 Thus, there is a need to describe the indicators that aid in monitoring the use of antimicrobials in specific departments. The current investigation shows, for the first time, the indicators deemed significant by experts from various specialities for monitoring the performance of ASPs in EDs. It provides a potential starting point from which to prioritise activities for optimising antimicrobial usage in such departments. The approach yielded 79 indicators divided into 4 blocks (consumption, microbiological, process, and outcome indicators) and 3 priority levels, half of which were considered high priority.

Previous authors have discussed the lack of ASP indicators in EDs.13 A recent study on the use of antimicrobials in these departments suggested that although the described experiences were encouraging, there remains a need for well-planned studies using relevant indicators.12 The present study followed a methodology similar to that used by Schoffelen et al.16 They selected 22 indicators associated with ASP activities in EDs that comprised process indicators covering ASP prerequisites, empirical therapy, documentation of information, and post-discharge educational activities, but without including indicators relating to antimicrobial use, resistance monitoring, or clinical outcomes. To date, we have not found any publication that has established a set of consumption and microbiological indicators that can be used to infer the quality of antimicrobial use in EDs.

The indicators obtained from this consensus were grouped into 4 blocks, addressing the need to monitor different aspects of ASP activities. Although these programmes have been shown to optimise antimicrobial use while reducing direct costs and hospital stays,17,18 their effect on the emergence and spread of multidrug-resistant strains remains uncertain.19 The implementation of ASPs in EDs has been shown to reduce overall antimicrobial prescribing by 10%–40%, while increasing adherence to clinical practice guideline recommendations.20 However, there remains a pressing need for microbiological indicators related to resistance profiles and adequate culture collection,21,22 as they would allow treatment protocols to be updated to the resistance profiles and ensure the adequate identification of the possible pathogens responsible for infections. The indicator panel included 14 process indicators (Table 3), including key elements, such as designating the individuals responsible for applying ASP policies in EDs, and other elements essential to coordinating ASP activity and achieving the objectives of optimising consumption and improving clinical outcomes.

The methodology used in this study has certain limitations. Firstly, the final results do not reflect the indicators supported by clinical evidence. However, to date, there have been no comparative studies on the impact of these indicators on the clinical outcomes of patients or the emergence of antimicrobial-resistant strains.12 Furthermore, although the cut-off points used to select the priority levels were based on previous experience,21 they are not based on rigorously established criteria. It should be noted that the experts were selected according to their experience in the field of study as well as their involvement in ASPs. However, the expert panel was not analysed to ensure sufficient representativeness (in terms of gender, age, years of experience, etc), as suggested by the Delphi methodology,23 thereby resulting in potential bias in the obtained evaluations.

During the 2 evaluation rounds, several members of the expert panel noted that it can be difficult to extract antimicrobial use data from data systems. This barrier is widely documented in the literature24 and can hinder the development of ASPs. Despite this aspect, all the evaluated indicators for ease of implementation had a median score of more than 6 points. This study is also limited by the lack of evaluators specialised in information systems, and therefore it remains unclear how much time and resources are required to access these data with accuracy.

The final document of this study includes the appropriate indicators for monitoring antimicrobial use, resistance profiles, and the secondary clinical outcomes related to infectious processes in EDs within the scope of ASPs. Given the scarce information available on indicators for monitoring ASPs in EDs, as well as the multidisciplinary nature and experience of the expert panel, we suggest that the indicator panel obtained represents a turning point for the implementation of these programmes in EDs. The monitoring of these indicators will prove valuable in promoting the implementation of ASPs in EDs, and in comparing the use of antimicrobials and their association with resistance profiles.

This project was conducted with funds from the SEFH Call for Project Grants for Working Groups 2022–2023.

Jesús Ruiz, M. Rosario Santolaya, Javier Candel, and Juan González participated in developing and designing the study, data collection, data analysis and interpretation, writing the article, and approving the final version for publication.

The other authors participated in the evaluation of the proposed indicators, including their prioritisation, as well as approval of the final document.