Online Triage for Patients: Implementing a Scalable and Cost-Effective Triage Platform Using Expert System, Machine Learning and Natural Language Processing Techniques Public Deposited

Downloadable Content

Download PDF
Last Modified
  • February 28, 2019
  • Oakkar, Oakkar
    • Affiliation: School of Information and Library Science
  • Keona Health, a spin-off company from the University of North Carolina (UNC) at Chapel Hill and in partnership with the National Institute of Health, has developed a triage platform called Online Triage. UNC Chapel Hill's Campus Health Services piloted the Keona platform by offering it to over 28,000 students. The pilot demonstrated significant nurse time savings, improved patient safety, and increased nurse satisfaction. Students used Online Triage from home using a link on the university website, as well as from kiosks in Health Services facilities. We tracked data on several variables including nurse review time and concordance between the nurse and our Clinical Decision Support (CDS) system. A total of 93% of students said they would use Online Triage to save time, but also to get reassurance and save money. Additionally, 77% of patients said they were moderately, very, or extremely likely to use it next time they had a health problem. Compared to traditional triage using Online Triage resulted in a 31% reduction in nurse time. The CDS system was able to safely screen patients for potential emergencies, and improve safety over patient's pre-dispositions. Nurses also reported a high degree of satisfaction with Online Triage and valued how it prepared them for patient interactions and reduced their documentation time. More work needs to be done improving the questionnaires, emergency screening, and display of information to nurses.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Mostafa, Javed
  • Master of Science in Information Science
Degree granting institution
  • University of North Carolina at Chapel Hill
  • 39
Deposit record
  • a03a4dce-6d08-44cf-8935-0346297c4e3f

This work has no parents.