Children's Health, a paediatric health system in North Texas, has implemented clinically oriented generative AI solution Pieces within its inpatient clinical operations.

The solution generates a real-time 100-word summary of a patient's course, spanning diverse clinical disciplines.

It uses contextual modelling, collaborative and adversarial AI, human-in-the-loop supervision, and automated clinical prompt engineering.

According to the vendor, this results in an ‘extremely low rate’ of serious errors.

Integrated into the Epic electronic medical record, the solution is used by healthcare professionals for aiding various tasks including clinical handoffs as well as length of stay management.

The tool can also keep track of discharge barriers, facilitate team communication, enhance care team alignment, and save time that would have otherwise been spent in manual chart reviews.

The vendor claims that it has been successful in showcasing clinical summarisation across varying age ranges.

Children's Health chief medical information officer Dr. Philip Bernard said: “We are encouraged by the positive responses to this technology from our care teams.

“Our clinicians recognise that Generative AI will allow our teams to enhance high-quality, innovative care. Pieces is delivering meaningful and generative AI for our patients, ensuring the safe use of this technology today.”

Pieces CEO Ruben Amarasingham said: “Working with Children's Health is an honour and privilege for Pieces.

“Our work with Children's Health is an opportunity to craft an AI that matches their reputation as a national leader in pediatric health care.”

Last month, Pieces integrated Amazon Web Services (AWS) generative AI services, including Amazon Bedrock, Amazon Elastic Compute Cloud (Amazon EC2), and Amazon SageMaker Canvas, into its healthcare solutions.

This integration allowed Pieces to offer Sculpted AI, a technology customised to the requirements of health systems, down to individual units, specialties, or physician level.