Deep Learning Based Early Warning Score in Rapid Response Team Activation

Hospital Rapid Response Team Clinical Trial

Official title:

Comparison of Deep Learning Based Early Warning Score and Conventional Screening System in Rapid Response Team Activation in General Ward Patients

Clinical Trial Details — Status: Not yet recruiting

Administrative data

• NCT number: NCT04951973
• Other study ID #: DEWS_2021
• Status: Not yet recruiting
• Start date: August 1, 2021
• Est. completion date: April 30, 2022

Study information

• Verified date: June 2021
• Source: Seoul National University Hospital
• Contact: Yeon Joo Lee, MD
• Phone: 82-31-787-7082
• Email: yjlee1117[@]snubh.org
• Is FDA regulated: No
• Study type: Observational [Patient Registry]

Clinical Trial Summary

The objective of this study is to evaluate the safety and clinical usefulness of the Deep learning based Early Warning Score (DEWS).

Description:

SPTTS is the representative trigger tracking system. In addition to the conventional SPTTS, DEWS will be calculated at each time point by the previously developed algorithm. SPTTS and DEWS will be shown simulataneously on the screening board. The rapid response team performs the rescue activity as before, using both SPTTS and DEWS simultaneously.

The alarm threshold setting of DEWS will be changed to 70 points, 75 points, and 80 points every month.

The primary and secondary outcomes will be evaluated to compare SPTTS and DEWS (based on each threshold).

Recruitment information / eligibility

• Status: Not yet recruiting
• Enrollment: 50000
• Est. completion date: April 30, 2022
• Est. primary completion date: December 30, 2021
• Gender: All
• Age group: 18 Years and older

Eligibility

Inclusion Criteria:

• Patients admitted to general ward and monitored by in-hospital rapid response system

Exclusion Criteria:

• patients admitted to pediatric ward

• patients in emergency room, intensive care unit, and operating room

Related Conditions & MeSH terms

• Emergencies
• Hospital Medical Emergency Team
• Hospital Rapid Response Team

Intervention

Diagnostic Test:
• Deep Learning Based Early Warning Score (DEWS)
DEWS use 4 vital signs (systolic blood pressure, HR, respiratory rate, and body temperature) to predict in-hospital cardiac arrest. Deep-learning approach facilitates learning the relationship between the vital signs and cardiac arrest to achieve the high sensitivity and low false-alarm rate of the track-and-trigger system (TTS).

Locations

Country	Name	City	State
n/a

Sponsors (7)

Lead Sponsor	Collaborator
Seoul National University Hospital	Dong-A University, Inha University Hospital, Korea Health Industry Development Institute, Mediplex Sejong Hospital, Incheon, Sejong Hospital, Bucheon, VUNO

References & Publications (2)

Cho KJ, Kwon O, Kwon JM, Lee Y, Park H, Jeon KH, Kim KH, Park J, Oh BH. Detecting Patient Deterioration Using Artificial Intelligence in a Rapid Response System. Crit Care Med. 2020 Apr;48(4):e285-e289. doi: 10.1097/CCM.0000000000004236. — View Citation

Kwon JM, Lee Y, Lee Y, Lee S, Park J. An Algorithm Based on Deep Learning for Predicting In-Hospital Cardiac Arrest. J Am Heart Assoc. 2018 Jun 26;7(13). pii: e008678. doi: 10.1161/JAHA.118.008678. — View Citation

Outcome

Type	Measure	Description	Time frame	Safety issue
Primary	In-hospital cardiac arrest	Compare the predictability of in-hospital cardiac arrest between DEWS and SPTTS.	3 month
Secondary	Alarm coincidence	Evaluate the alarm coincidence between DEWS and SPTTS.	3 month
Secondary	Total alarm count.	Compare the total alarm count between DEWS and SPTTS.	3 month