Making Antibiotic Guidance Work in the Emergency Department

The emergency department (ED) is where antibiotic stewardship confronts its most difficult test [1]. Clinicians must decide whether infection is present, choose an empirical regimen, and start treatment quickly, often before diagnostic results are available. The study by Kwon and colleagues, published in this issue of Infection & Chemotherapy, makes an important point: ED clinicians do not simply need more antibiotic knowledge [2]. Existing guidance is often designed in a form that does not match the work of emergency care [1,3].

The authors conducted focus group interviews with 16 physicians involved in antibiotic prescribing at four tertiary hospitals in Korea. Their findings identify a practical gap between how guidelines are written and how antibiotics are prescribed at the bedside. Participants described diagnostic uncertainty, overlapping infection syndromes, and practical prescribing details—renal dose adjustment, infusion information, institutional habits—as recurring challenges. Only 11 of 16 participants were aware of existing guidelines, and even those rarely used them because the documents were hard to access and poorly suited to real-time ED decisions [2]. This low awareness echoes a nationwide survey reporting that only 12.9% of Korean hospitals had institution-specific empirical antibiotic guidelines as of 2020 [4].

This study deserves attention because it moves the discussion of ED stewardship beyond whether clinicians follow guidelines, to a more fundamental question: can clinicians realistically use guidelines in the circumstances in which the decision must be made?

Traditional antibiotic guidelines are usually organized around infection sites and written as comprehensive reference documents [1]. That structure is valuable for detailed consultation, education, and standardization, but less suitable when a clinician is assessing several unstable patients, working with incomplete information, and preparing an order within minutes [3].

The distinction between a reference document and a decision-support tool is central. A reference document explains what should be done across a broad range of situations; a decision-support tool helps a clinician decide what to do now, with the information at hand, without ignoring that uncertainty remains.

The participants' concerns illustrate this difference. They did not ask for shorter guidelines simply because shorter is easier to read. They wanted guidance that could answer practical questions: Is antibiotic treatment indicated at this stage? Which features require broader coverage? How should the regimen be adjusted for renal dysfunction? Which diluent and infusion rate should be used? Is the drug stocked at this institution? These are not peripheral details–they are part of antibiotic selection in the ED.

The findings suggest that ED-specific guidance should be organized around the sequence of clinical decisions rather than isolated disease labels alone [3]. A useful tool might begin with severity, likely source, host factors, resistance risk, and competing or coexisting syndromes, then lead to a limited set of treatment options that clearly distinguish standard therapy from situations requiring escalation or specialist consultation.

Such an approach would not replace clinical judgment; it would make the boundaries of that judgment more visible. Algorithms should identify when a simplified pathway is unsafe—severe sepsis, major immunocompromise, substantial renal dysfunction, suspected polymicrobial infection, or uncertainty about the infectious focus. The goal is not to force every patient into a narrow flowchart, but to reduce avoidable cognitive work while preserving room to pause and broaden the assessment.

The study also highlights the importance of including prescribing information alongside the recommendation. Dose, renal adjustment, administration fluid, infusion rate, formulary availability, and contraindications should not require a second search; separating these details from the recommendation creates friction precisely when the risk of error is highest.

A guideline can be evidence-based and still fail at the point of care if it does not reflect local practice [5]. Participants described a gap between general recommendations and the preferences of receiving departments, local prescribing routines, and institutional formularies–particularly important in the ED, where decisions are made during a transition of care and may later be modified elsewhere.

Institution-specific adaptation should therefore be a core component of guideline development, not a final editorial step [5]. Local resistance data, formulary restrictions, diagnostic resources, consultation pathways, and admission practices should shape the final tool, developed with ED physicians, infectious disease specialists, pharmacists, microbiologists, nurses, and information technology staff, together with departments that receive admitted patients. Without this broader participation, a guideline may reflect its authors' perspective rather than its users' workflow.

Local adaptation also requires maintenance. Recommendations lose credibility when clinicians cannot tell whether they are current. Each institutional guideline should identify a responsible owner, review interval, trigger events for interim revision, and a visible date of update.

Making a guideline available is not the same as making it usable. Participants favored one- or two-page summaries, point-of-care materials, and integration into electronic prescribing systems–a layered model: a brief algorithm for immediate decisions, expandable prescribing details, and a full reference document for complex cases [1].

Electronic integration could strengthen this model, displaying renal dose adjustments, formulary information, or administration instructions at the moment an antibiotic is selected. However, prompts should be selective and clinically relevant; excessive alerts create fatigue and encourage automatic dismissal.

Implementation should be evaluated with more than download counts or self-reported awareness. Relevant outcomes include time to appropriate therapy, guideline-concordant prescribing, unnecessary broad-spectrum use, dosing errors, de-escalation after diagnostic clarification, and clinician experience [1]. Because this study examined perceptions rather than prescribing outcomes, the next step is to test whether user-centered guidance changes practice without delaying urgent treatment or increasing inappropriate restriction.

The study appropriately identifies audit and feedback as part of the intervention, not an optional educational supplement [6]. In the ED, clinicians may not see what happened after admission–whether cultures changed the regimen, or the initial choice was later judged inappropriate–so the link between decision and outcome remains weak.

Feedback should be timely, specific, and non-punitive [6], focused on a small number of actionable issues such as unnecessary broad coverage, missed renal adjustment, or failure to reassess therapy when new information emerges. Education and data feedback have been shown to improve antibiotic prescribing in the ED [7], and combining feedback with clinician education can further strengthen guideline-concordant prescribing. The purpose is not to blame clinicians for decisions made under uncertainty, but to improve the conditions in which those decisions are made.

The central message of this study is straightforward: ED antibiotic stewardship will not be strengthened by more comprehensive documents alone. Guidance must be designed around emergency decision-making, adapted to each institution, integrated into prescribing workflow, and reinforced through feedback. The most useful ED guideline may be shorter than a conventional one, but more deliberate about uncertainty, practical prescribing, local context, and escalation. In Korea, the nationwide expansion of antimicrobial stewardship policy in 2024 has created a new institutional context for such efforts [8]. In line with this direction, the Korea Disease Control and Prevention Agency developed and distributed practical guidelines in 2026 for the appropriate use of antibiotics in five common infections—acute upper respiratory tract infection, urinary tract infection, intra-abdominal infection, community-acquired pneumonia, and skin and soft tissue infection—together with an integrated handbook [9]. These field-oriented materials are a welcome step toward usable guidance; their impact in the ED will depend on how well they are adapted to local practice, embedded in prescribing workflows, and evaluated against actual prescribing outcomes. Developing and evaluating user-centered tools with the clinicians who use them should be a priority for antimicrobial stewardship programs in Korea and elsewhere [10].