# Duke University Hospital made rapid response nurses the users of its sepsis deep-learning tool: they triage its alerts and call physicians, at up to four high-risk alerts an hour

Page: https://flyvolo.ai/en/changes/ev-20181105-registered-nurse-7
Date: 2018-11-05 · Stage: Deployment
Can this record move a task judgement? yes
Occupation: https://flyvolo.ai/en/careers/registered-nurse

## Scope

North Carolina, United States. The hospital's own implementation study says Sepsis Watch launched on 5 November 2018, that the rapid response team nurse triages patients at risk of sepsis using it and communicates with emergency department clinicians about treatment, that those nurses called physicians about every patient with or at high risk of sepsis, and that up to four high-risk alerts per hour was agreed as the ideal volume for a single nurse user. One hospital and one specialised nursing team; it reports no workload measurements.

## Source

- Sendak M et al. (Duke University Health System) — 'Real-World Integration of a Sepsis Deep Learning Technology Into Routine Clinical Care: Implementation Study', JMIR Medical Informatics 8(7) (15 July 2020) — https://pmc.ncbi.nlm.nih.gov/articles/PMC7391165/ (primary source)

## What this means

Watching and acting on a deep-learning model's alerts has become a defined nursing job in at least one hospital.

## What it does not show yet

One hospital and one specialist team, with no measured workload.

## How to verify it yourself

Open the JMIR Medical Informatics study of Sepsis Watch (PMC7391165) and find "Up to 4 high-risk alerts per hour".


More: https://flyvolo.ai/llms.txt