CAPTION AI to Minimize Risk of COVID Exposure

Not Applicable

Withdrawn

• Conditions: COVID-19
• Interventions: Device: Caption AI

• Registration Number: NCT04336774
• Lead Sponsor: Duke University

Brief Summary

Participants scheduled for for an echocardiogram (echo) and being evaluated for, or is positive for COVID-19 will be asked if they would be willing to have their echo done using a new software program on one of the hand-held ultrasound scanners.

The new software program guides the investigator, or any other non-sonographer, to take the best possible pictures of the participants heart. The prior version of this software is already being used clinically and is FDA approved. The main reason for using the updated version is that it's faster and better in terms of guiding the user.

Detailed Description

To enable healthcare professionals that are not proficient in transthoracic echo (TTE) to acquire images in patients being evaluated for or positive for COVID-19. By leveraging the capabilities of the Caption AI which is designed to train novice users on how to acquire TTE, this would minimize the risk of sonographers to be exposed to COVID-19. Additionally, minimizing sonographer interaction with patients being evaluated for or positive for COVID 19 minimizes the risk of sonographers as vectors for transmission to other patients. Lastly, since the Caption AI device will be dedicated to the COVID wards and COVID ICU and not transported to other locations, use of the CAPTION AI device will help to limit viral transmission via the surfaces of the ultrasound machine. These images will be assessed by qualified medical professionals for diagnosis.

Study Timeline

• Study First Submitted: 2020-04-03
• Study First Submitted QC: 2020-04-03
• Study First Posted: 2020-04-07
• Study Start: 2020-07
• Status Verified: 2020-08
• Last Update Submitted: 2020-08-06
• Last Update Posted: 2020-08-10
• Primary Completion: 2020-11-01
• Study Completion: 2020-11-09

Recruitment & Eligibility

• Status: WITHDRAWN
• Sex: All
• Target Recruitment: Not specified

Inclusion Criteria

• Duke patients within the MICU and COVID overflow areas
• transthoracic echocardiogram ordered by their provider
• suspected or positive for COVID-19.
• Patients who consent to participating in the study or Physician discretion that information to be gained is important to the patient
• Patients ≥18 years old

Read More

Exclusion Criteria

• Unable to lie flat for study
• Patients unwilling to give consent

Read More

Study & Design

• Study Type: INTERVENTIONAL
• Study Design: SINGLE_GROUP

Arm && Interventions

Group	Intervention	Description
Echocardiogram patients	Caption AI	Patients scheduled to have an echocardiogram (echo) and who are also being evaluated for, or are positive for COVID-19.

Primary Outcome Measures

Name	Time	Method
Percent of patient echos that are not interpretable	Up to 1 hour	Images obtained through the point of care AI machine will be uploaded to the cardiology PACS system and read. If the images are felt to be not interpretable, the echo lab will send a sonographer with a regular echo machine to the patient's room to perform the study.
Time to acquire images as measured by time stamps	up to 1 hour
Percent of patient echos that provide an automated (AI) LVEF (left ventricular ejection fraction)	Up to 1 hour	There need to be enough images taken of sufficient quality to allow for calculation of an automated LVEF by the AI algorithm

Secondary Outcome Measures

Name	Time	Method
Percent of agreement between AI calculate LVEF and LVEF read by physician	Up to 24 hours

Trial Locations

Locations (2)

Duke Health; 🇺🇸 Durham, North Carolina, United States

Duke University Medical Center; 🇺🇸 Durham, North Carolina, United States