Artificial Intelligence in Infection Prevention: What's New and What's to Come
Artificial intelligence offers seemingly limitless potential on its face. Reading headlines coming out of practically every industry on the planet over the past couple years or so, you’d be forgiven for thinking that we’re all just waiting for an AI development that will solve any intractable problem across any discipline. The truth is obviously more nuanced than that, but in a field like healthcare where mistakes can be life or death, demand for simple solutions for complex problems is only growing, and the problems themselves remain as nuanced and numerous as ever, you’d be forgiven for pinning a lot of hope on AI tools to help navigate the morass of keeping patients healthy and happy as efficiently as possible.
Luckily, there are a number of studies and testimonies from seasoned veterans in the field that posit we may be right to pin a lot of hope on artificial intelligence to help us make sense of complicated questions. One of them, published in the NIH’s PubMed database, specifically evaluated the potential for AI to help “prevent, detect, and manage” healthcare-associated infections. The researchers found that “AI models demonstrated high predictive accuracy for the detection, surveillance, and prevention of multiple HAIs, with models for surgical site infections and urinary tract infections frequently achieving area-under-the-curve (AUC) scores exceeding 0.80, indicating strong reliability. … Advanced algorithms, including neural networks, decision trees, and random forests, significantly improved detection rates when integrated with EHRs, enabling real-time surveillance and timely interventions. In resource-constrained settings, non-real-time AI models utilizing historical EHR data showed considerable scalability, facilitating broader implementation in infection surveillance and control. AI-supported surveillance systems outperformed traditional methods in accurately identifying infection rates and enhancing compliance with hand hygiene protocols.”
In addition, AI “played a pivotal role in antimicrobial stewardship by predicting the emergence of multidrug-resistant organisms and guiding optimal antibiotic usage, thereby reducing reliance on second-line treatments. However, challenges including the need for comprehensive clinician training, high integration costs, and ensuring compatibility with existing workflows were identified as barriers to widespread adoption.” This piece of the study further emphasizes the notion that these tools have enormous promise yet significant barriers to widespread adoption still exist. Regardless, those barriers don’t negate the promise these tools hold.
Another important piece of AI implementation is garnering and facilitating clinicians’ trust in the tools and models. This study also demonstrated that certain interpretability tools “increased clinician trust and facilitated actionable insights.” Following off of this, the authors of the study further emphasized that “successful implementation necessitates standardized validation protocols, transparent data reporting, and the development of user-friendly interfaces to ensure seamless adoption by healthcare professionals. Variability in data sources and model validations across studies underscores the necessity for multicenter collaborations and external validations to ensure consistent performance across diverse healthcare environments.” These conclusions and suggestions lend credence to the notion that it will be difficult to implement artificial intelligence across healthcare disciplines at a large scale, but the learning curve will likely lead to great improvements in the end to help with patient care.
Healthcare Purchasing News spoke with Dr. Rodney Rohde, PhD, MS, SM(ASCP)CM, SVCM, MBCM, FACSc, Regents’ Professor at Texas State University System, to get a better sense of how healthcare systems are using artificial intelligence on the ground, and to hear an expert’s accounting of the benefits, shortcomings, and promise these tools and technologies hold.
How is artificial intelligence being used in health systems and hospitals?
Rohde: Artificial intelligence is increasingly transforming healthcare by enhancing the way hospitals and health systems deliver care. For example, AI is used to assist in diagnosing diseases and detecting conditions early via medical imaging, pathology, and radiology. Using predictive analytics for patient outcomes such as diabetes, heart conditions, or which patients may be at risk for readmission. AI is also being used for personalized medicine, clinical decision support, personal health assistants, robotics (e.g., surgery), administrative automation like email management and brainstorming for process improvement, drug discovery, population health, mental health, supply chain logistics, and dosing medications.
Are there any particularly exciting advancements in artificial intelligence you're aware of?
Rohde: Personally, all of the answers are exciting. However, as someone who is involved in antimicrobial resistance, I am excited about drug discovery and managing the complicated world of selecting appropriate antimicrobial agents while reducing the risk of increasing resistance patterns. AI can also help the medical laboratory collaborate more effectively with physicians and pharmacists to maintain the interplay between “antimicrobial breakpoints" and the correct therapy.
What potential does AI have in infection prevention?
AI has significant potential in infection prevention by enabling early detection, monitoring, and response. It can analyze patient data, medical records, and environmental factors to predict outbreaks and identify high-risk individuals. AI-powered systems can also monitor hand hygiene compliance, track infection spread within hospitals and optimize infection control protocols. Additionally, AI can assist in designing targeted antimicrobial therapies and predicting resistance patterns, ultimately reducing healthcare-associated infections and improving patient safety.
What hesitancies, if any, do you have when it comes to incorporating AI in hospitals' operations?
Rohde: I think I would be most concerned about data privacy and security, as patient information is highly sensitive. There are also risks of AI bias, which could lead to unequal care or misdiagnoses, especially for underrepresented populations. Additionally, reliance on AI could diminish human oversight in critical decisions, and the technology may face regulatory hurdles or ethical challenges. Ensuring transparency, accountability, and proper training for healthcare professionals are crucial to addressing these concerns.
Are there any untapped areas where you could see AI helping?
Rohde: AI has untapped potential in areas like improving diagnostic accuracy for rare diseases, enhancing real-time surveillance of infectious disease outbreaks, and streamlining laboratory processes. AI could also play a key role in analyzing vast amounts of epidemiological data to predict future health trends, helping public health officials respond more effectively to emerging threats. Additionally, AI’s ability to process complex genomic data holds promise for personalized medicine and targeted therapies.
Does AI have any shortcomings in infection prevention?
Rohde: While AI offers significant promise in infection prevention, it has shortcomings, particularly in its reliance on high-quality, comprehensive data. Incomplete or biased data can lead to inaccurate predictions or misidentifications of risks. Additionally, AI systems may struggle to account for complex, real-world variables, such as human behavior and environmental factors, which are crucial in preventing infections. There’s also the challenge of integrating AI seamlessly into existing healthcare systems and ensuring that staff are adequately trained to use these tools effectively.
Are there any technologies for infection prevention, AI or not, you would like to spotlight?
Rohde: Wow, there are so many things coming out almost weekly now that are interesting. It’s difficult to keep track of them all. However, for me, technologies like advanced biosensors, machine learning-based predictive models, and automated disinfecting robots as key innovations in infection prevention. AI-powered predictive analytics can help forecast outbreaks and identify high-risk patients, while biosensors and wearables monitor real-time health data to prevent infections. Automated robots, equipped with UV light or chemical disinfectants, help reduce human error and ensure consistent sanitation in high-risk areas, such as operating rooms and intensive care units.
Anything else you'd like to tell our readers?
Rohde: AI is absolutely going to transform our world. I continue to tell my students, alumni, and colleagues that I hope “we” all embrace the idea of “augmented intelligence.” In other words, we must all ensure that AI is utilized and developed in the most ethical way possible going forward. It’s an exciting time.
