Evaluating AI in real-world clinical settings

In healthcare, AI promises to transform clinical decision-making processes as it has the potential to harness the vast amounts of genomic, biomarker, and phenotype data that are being generated across the health system to improve the safety and quality of care decisions. Today, AI has been successfully incorporated into variety of clinical decision support systems for detecting clinical findings in medical imaging, suggesting diagnoses and recommending treatments in data-intensive specialties like radiology, pathology and ophthalmology. Future systems are expected to be increasingly more autonomous, going beyond making recommendations about possible clinical actions to autonomously performing certain tasks such as triaging patients and screening referrals.

Little is known about the use of the present generation of AI in clinical settings, with evaluation of these systems primarily focusing on examining the performance of algorithms within laboratory settings. To date, there have been limited published observational studies in the contemporary literature that have investigated the use of AI-based decision support systems within a clinical setting or within standard of care. Furthermore, very few studies have examined the use of AI in real-world clinical settings to demonstrate its effects on care delivery and patient outcomes. As such further research in this clinical area is warranted, to ensure that such technologies are appropriately applied and validated for clinical use and safety.

The goals of this project are:

• To investigate approaches to safely implement and evaluate the impact of AI systems on clinical decision-making, care delivery and patient outcomes.
• To develop requirements for safe implementation and use of AI in real-world clinical settings.

We welcome enquiries about approaches for safe and effective integration of AI in real-world clinical settings, and are happy to assist individuals and organisations study the impact of AI on decision-making, care delivery and patient outcomes.