Abridge AI Pilot Test

We are piloting a new technology (Abridge) that uses artificial intelligence to listen to clinician-patient office visits and then document the interaction in the electronic health record.

Study Overview

• Status: Active, not recruiting
• Conditions: Clinician Experience; Clinician Burnout
• Intervention / Treatment: Other: Ambient AI for clinician documentation

Detailed Description

This study will evaluate pilot implementation of the Abridge Artificial Intelligence (AI) Platform (Abridge) in UCHealth clinics. Abridge licenses will be deployed to 116 physician and advanced practice clinicians to accomplish the following goals:

1. Determine if Abridge meaningfully impacts clinician documentation time and burnout or stress
2. Determine whether to scale Abridge beyond the pilot

We will evaluate clinician experiences using Abridge, particularly with regard to documentation time and clinician burnout or stress. Study findings will explore value added in integrating Abridge into clinician practice, and guide decision-making for entering into a commercial agreement with Abridge. This pilot will also inform best practices for scaling Abridge to a broader user base across UCHealth outpatient practices.

• Study Type: Interventional
• Enrollment (Estimated): 232
• Phase: Not Applicable

Contacts and Locations

Study Locations

• United States
  • Colorado
    • Aurora, Colorado, United States, 80045
      • UCHealth

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

-clinician (physician, advance practice professional, or licensed mental health professional) employed by UCHealth

• See clinic patients at least three half-day sessions per week. (Half-day sessions seeing or supervising patients with residents and fellows will NOT count toward the weekly total. Half-day sessions seeing patients with a medical student CAN count toward the weekly total, however, Abridge can only capture conversations where you are physically present with patients.)
• Own an iOS smartphone and be willing to download and use the Epic Haiku smartphone app during clinic visits

Exclusion Criteria:

• Trainees including residents and fellows
• PT/OT, ED, maternal/fetal, pediatric, dental, eye specialties

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Other
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Assigned Abridge license; Providers assigned a license	Other: Ambient AI for clinician documentation; Ambient AI for clinician documentation; Other Names: Abridge
No Intervention: Control; Providers not assigned a license or waitlisted

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Burnout - High emotional exhaustion	Validated measure collected by survey. Reference: https://pmc.ncbi.nlm.nih.gov/articles/PMC3475833/	Baseline, 2, 4 and 8 months
Burnout - High depersonalization	Validated measure collected by survey. Reference: https://pmc.ncbi.nlm.nih.gov/articles/PMC3475833/	Baseline, 2, 4 and 8 months
Burnout - emotional exhaustion and depersonalization	Validated measure collected by survey. Reference: https://pmc.ncbi.nlm.nih.gov/articles/PMC3475833/	Baseline, 2, 4 and 8 months
Time in notes per appointment	Collected via EHR signal data	Baseline, 2, 4 and 8 months
Adoption	Percent of eligible visits in which intervention is used. Collected via EHR data	2, 4 and 8 months

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Reach	Proportion and representativeness of eligible patients and visits intervention was used for	2, 4 and 8 months
Clinician experience	Likelihood to recommend via validated NPS, qualitative drivers of experience, likelihood to change jobs, Weiners feasiblity/implementability/acceptability, system usability scale	Baseline, 2, 4 and 8 months
Time documenting	Time spent documenting beyond appointment, average time in the EHR, time of last log out of the day, pajama time documenting	Baseline, 2, 4 and 8 months
Operational efficiency	Same day note closure, average note star rating, average note turn around time	Baseline, 2, 4 and 8 months
Adoption	Abandonment rate and representativeness	2, 4 and 8 months

Collaborators and Investigators

• Sponsor: University of Colorado, Denver

Study record dates

Study Major Dates

• Study Start (Actual): January 13, 2025
• Primary Completion (Estimated): July 13, 2025
• Study Completion (Estimated): December 31, 2025

Study Registration Dates

• First Submitted: January 10, 2025
• First Submitted That Met QC Criteria: January 10, 2025
• First Posted (Actual): March 25, 2025

Study Record Updates

• Last Update Posted (Actual): March 25, 2025
• Last Update Submitted That Met QC Criteria: January 17, 2025
• Last Verified: January 1, 2025

More Information

Terms related to this study

• Keywords: Ambient AI; Clinician burnout
• Additional Relevant MeSH Terms: Behavioral Symptoms; Stress, Psychological; Burnout, Psychological
• Other Study ID Numbers: 24-1623

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: Unsure if the third party owner of Abridge will allow sharing

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No