THE GLOBAL antimicrobial resistance (AMR) crisis is a major global health challenge, threatening 10 million lives annually by 2025, with emerging research presented at ESCMID Global 2026 highlighting its ever-growing burden.
Rasha Elshenawy, Public Health and Patient Safety Unit, University of Hertfordshire, Hertfordshire, UK, presented her research at ESCMID Global 2026, exploring how hospitals and clinicians can leverage AI in antimicrobial stewardship (AMS).1
After studying the use of AI across two UK NHS Foundation Trust hospitals, Elshenawy found that machine leaning algorithms were useful in predicting prescribing appropriateness, intervention timing, and antimicrobial risk.1
How AI Can Help Fight the AMR Crisis
She explained: “AI can help us to produce, based on patterns of antibiotic use, a warning sign when we have a prediction of high resistance at certain places in the hospital or for a specific condition in our patient, and help us to take the corrective action.
“So, AI can be used as acceleration, it can be used in decision support, as guidance for doctors to prescribe properly, and also for all healthcare professionals, such as pharmacists who dispense, clinical pharmacists in ward rounds, and nurses.
“Using AI properly is very promising with regard to antibiotic resistance and stewardship.”
In the study, AI prediction accuracy for prescribing appropriateness was 84.7%.1
An AI system also identified 156 drug interactions and 89 dosing adjustments requiring intervention.1
Following the success of AI in aiding clinical care, technical validation confirmed 99.2% data accuracy and expert validation demonstrated strong clinical relevance.1
The observed improvements remained robust throughout pandemic waves, despite a 40% increase in workload.1
Implementation in Low- and Middle-Income Countries
AMR disproportionately impacts low- and middle-income countries (LMICs), yielding high mortality in particular regions.
In another study, Elshenawy and the research team found that AI-enabled stewardship is technically feasible in LMICs, but implementation gaps are severe.2
She said: “Low- and middle-income countries have severe constraints with regard to the supply chain, with regard to resources, with regard to policy, guidelines, education, access – they maintain all of these barriers.
“Focusing on the facilitators that help implement AI is very important.”
The team called for mandatory cost-effectiveness analysis, patient-centred outcomes, co-design requirements, sustainable domestic financing, and capacity-building for autonomous LMIC-led innovation in the field.2
Elshenawy added: “Based on real data from these countries, we can implement tailored, targeted interventions for antibiotic stewardship.
“Another important issue is education – how we could use these stewardship strategies to educate health professionals in patient care, to promote the effective use of antibiotics.”
Responsible AI Use
Notably, Elshenawy warned that AI must be properly validated prior to implementation in clinical settings.
She cautioned: “If we use AI properly in antibiotic stewardship implementation, in decision support, in prediction, in measuring antibiotics, it will have a significant effect on the patient – the most important effect of AI use.
“With regard to this: I’m promoting AI use, but I’m promoting the responsible use of AI.
“We need, first of all and before applying AI, to validate it.
“It is very important to make sure this AI will give appropriate recommendations.”
Elshenawy emphasised that AI may only be leveraged against the AMR crisis if hospitals implement robust testing protocols.
She concluded: “When hospitals or leaders are implementing AI, they need to consider validation and how appropriate the AI is as a tool and as a guide for decision support systems.”
References
1 Elshenawy R. Machine learning-enhanced real-time antimicrobial stewardship dashboard for predictive prescribing and pandemic-resilient quality improvement. P2998. ESCMID Global 2026, 17-21 April, 2026.
2 Aser M et al. Artificial intelligence applications in antimicrobial stewardship: a systematic review of effectiveness, implementation challenges, and equity considerations in low- and middle-income countries (LMICs). P3063. ESCMID Global 2026, 17-21 April, 2026.
Featured image: I Viewfinder on Adobe Stock
