Use AI to improve Nurse Workflow
In Winter 2019 and at Stanford’s Designing AI to Cultivate Human Well-Being class, interdisciplinary teams of students worked together to explore solutions to several important societal problems through the application of technology and AI. In this post and in the few following ones, the class teaching team will be highlighting the top 3 teams and how they defined their problem statement as well as their projects’ outcome.
Nurses are the primary hospital caregivers. Yet, a recent study finds that nurses spent just 19.3% of their with patients and only 7.2% of their assessing the patient and reading of vital signs. Most of their time was spent on documentation (35.3%), medication administration (17.2%), and care coordination (20.6%). On average, nurses walk 4–5 miles during a 12-hour shift (most Americans walk just 2.5–3 miles during the course of an 18-hour day). Health care leaders and nurses alike would prefer to reduce the number of steps a nurse takes during a shift in order to increase efficiency, decrease fatigue and increase time at the bedside.
So, How can we use AI to optimize nurses workflow?
Here is how the team approached the problem in their own words:
Hendrich et al. (2008) report in a study of nurse workflows that nurses spend less than 20% of their time on patient care activities. Documentation is particularly time-consuming, and takes place across seven different activities: admission paperwork, assessment, transcribing orders, writing care plan, medications paperwork, teaching, and discharge paperwork.
We propose a human-centered AI approach to tackle the problem of nurse workflow. Our vision is for nurses to snap a quick photo of observations they make and actions they perform; a computer vision algorithm will understand the type of situation presented in the photo, extracts key information, and logs it in an EHR.
Applications may include:
● Assessment/Vitals: automatically logging vitals, blood pressure, etc.
● Medication Administration: automatically logging key information (e.g. type and dose) about nurse’s medication administration.
Automated logging with computer vision helps improve nurses’ workflow and effectiveness in multiple ways:
● It saves time by automating the mundane tasks of documentation.
● It reduces error rates that stem from human mistakes in manual logging.
During the last class session, the team delivered their final presentation with their key insights and takeaways. Let us take a look: