Cerner, Amazon Web Services team to tackle readmissions

Register now

Cerner and Amazon Web Services are collaborating on ways to reduce patient readmissions and the time it takes for physicians to document patient visits.

Cerner brings its client base and technology, which it will provide to the partnership in efforts to reduce unnecessary follow-up work to doctor appointments. The initiative will use de-identified patient data to help make early determinations of what is causing return hospitalizations.

The insights Cerner hopes to find will give physicians resources to make more informed decisions on treatment approaches that enable reduced readmissions.

For example, a large Cerner client asked for help predicting patients at risk of being readmitted from a rehabilitation center back to the hospital. By applying machine learning technology to historical data migrated to the Amazon Web Services Cloud, Cerner built a model that helped the delivery system reach its lowest readmission rate in over a decade and sent more patients directly from rehabilitation to their homes.

At the 2019 Amazon Web Services conference in Las Vegas, Cerner CEO and Chairman Brent Shafer addressed some of the industry’s biggest challenges and ways to fix them. “As we enter into a new era of digital transformation, our work with AWS will lead a wave of breakthrough innovations focused on making healthcare better and clinicians’ work easier,” Shafer predicted.

To address physician burnout, Cerner is building a virtual scribe in work with Amazon Transcribe Medical to reduce documentation burden.

“Working with AWS will allow us to capture doctor-patient interaction and integrate it directly into the electronic workflow of the physician,” Shafer explained. Further, a new patent issued to Cerner supports voice-activated documentation, captures details shared between the patient and provider, and leverages voice inputs and natural language processing to reduce documentation burden.

For reprint and licensing requests for this article, click here.