Artificial intelligence (AI) is being used to provide same-day assessments of antimicrobial resistance for patients in intensive care units, which is crucial in preventing life-threatening conditions like sepsis. Antimicrobial resistance, the process by which microorganisms develop defenses against treatment, is a significant challenge in healthcare worldwide. It is estimated to cause 1.2 million deaths globally and costs the NHS at least £180 million per year. Infections in the bloodstream can become resistant to antibiotics, leading to sepsis, which can result in organ failure, shock, and even death.
Patients in intensive care units may have varying levels of antimicrobial resistance due to factors such as previous exposure to antibiotics, genetics, and diet. To address this issue, researchers from King’s College London and clinicians at Guy’s and St Thomas’ NHS Foundation Trust have collaborated to develop an AI-based system for assessing antimicrobial resistance and identifying sepsis-causing bloodstream infections. This interdisciplinary study aims to improve outcomes for critically ill patients by providing same-day triaging, particularly in environments with limited resources. The technology is more cost-effective than manual testing and could help clinicians make quicker, more informed decisions about patient care.
Current assessments of ICU patients require time-consuming laboratory tests, with bacteria needing to be cultured over several days. This delay can have a significant impact on care outcomes, particularly for critically ill patients. Quicker access to information on antimicrobial resistance would enable clinicians to make more timely decisions, including the appropriate use of antibiotics, which is crucial for positive patient outcomes. This study demonstrates the potential benefits of AI in healthcare, particularly in addressing issues related to antimicrobial resistance and bloodstream infections. By utilizing machine learning, researchers hope to provide clinicians with a valuable tool for making important decisions, especially in the ICU setting.
The use of AI to speed up the diagnosis of infections and enable the prescription of the correct antibiotics could have a profound impact on patient survival and care outcomes. It could also help preserve existing antibiotics and prevent further development of antimicrobial resistance. By accelerating the diagnosis process, clinicians may be able to avoid prescribing broad-spectrum antibiotics that can harm the patient’s microbiome and potentially make the pathogen more resistant to the drug. The study involving data from over 1,000 patients at Guy’s and St Thomas’ NHS Foundation Trust has paved the way for ongoing research with larger datasets of more than 20,000 individuals.
The researchers are exploring the potential for advanced approaches to this study, particularly within a multi-hospital setting using Federated Machine Learning technology. This innovative machine learning approach shows promise for widespread implementation, offering a robust solution to address critical healthcare issues on a larger scale. The ultimate goal is to improve patient outcomes by providing clinicians with more efficient tools for diagnosing and treating antimicrobial-resistant infections. By deploying this AI approach in the front line of the NHS, regulatory requirements could be fulfilled, leading to better patient care and the preservation of existing antibiotics.
