In this week’s round of insights, we discuss three of the latest findings in the application of artificial intelligence (AI) in healthcare.
Machine learning can help predict heart failure risk in CKD
A recent study from the Journal of the American Heart Association found that machine learning (ML) models can predict heart failure risk in patients with chronic kidney disease (CKD), a population at high cardiovascular risk.
In the study, the model outperformed traditional approaches in identifying patients at risk of developing heart failure. Key predictors included haemoglobin levels, kidney function markers, and comorbidities such as hypertension.
The findings suggest that ML-driven tools could support early intervention and personalised risk stratification, thus potentially improving patient outcomes.
AI-assisted readings may reduce radiologist liability
Emerging evidence has suggested that AI-supported ‘double reading’ in radiology could reduce diagnostic errors and, consequently, malpractice liability.
An article from the American Journal of Medical Care discusses a recent study by Bernstein et al., which placed AI as a second reader alongside radiologists to improve detection rates and consistency in image interpretation. By providing an additional layer of review, AI may help reduce missed diagnoses, one of the leading causes of malpractice claims in radiology.
However, questions remained around accountability, particularly in cases where AI recommendations conflict with clinician judgement.
Large language models can potentially improve drug safety
The growing use of large language models (LLMs) in pharmacovigilance is opening new possibilities for managing safety data. According to a piece published in Pharmaceutical Executive, LLMs can improve efficiency across the drug safety lifecycle.
These tools have proven particularly valuable in handling unstructured data, such as clinical narratives and adverse event reports. However, challenges regarding data accuracy, hallucinations, and maintaining compliance with regulatory standards remain. Human oversight remains essential to protect data integrity and patient safety.
For more industry insights visit elion.nz/insights/
Want to get in touch? Contact us today at https://elion.nz/get-in-touch/
