Automated Clinical Documentation
Conditions
Keywords
speech-to-text, Dragon Ambient eXperience (DAX), intervention evaluation
Brief summary
AI solutions like, Dragon Ambient eXperience (DAX) Copilot (Nuance/Microsoft), hold the potential to significantly enhance provider and patient interactions and alleviate pain points that contribute to burn out. Atrium Health was the first healthcare system in the world to pilot Nuance's DAX Copilot intelligence (AI) enabled scribe software, which synthesizes a draft clinic note by listening to the conversation between a provider and patient. After 180 days of use by primary care clinicians and advanced practice providers (APPs), along with a control group, the investigators assess provider satisfaction, patient experience, and provider efficiency through qualitative and quantitative methods.
Detailed description
In this non-randomized stepped-wedge trial, clinicians (including physicians and advanced practice providers \[APP\]) specializing in family medicine, internal medicine, and general pediatrics were recruited from outpatient clinics located in North Carolina and Georgia. The intervention group was divided into 5 waves between June and August 2023 based on clinic locations. Prior to the activation of accounts, clinicians underwent a one-hour training session on DAX Copilot. Meanwhile, a comparison group comprising clinicians not utilizing DAX was recruited in five waves from locations and specialties similar to those of the intervention group. Two types of primary outcomes, electronic health record (EHR) metrics and financial metrics, were assessed after 180 days of intervention. EHR metrics include time in EHR (EHR8), work time outside of work (WOW8), time in documentation (Note8), completed appointment rate, same day closure rate, and note length. Financial metrics include gross revenue per visit and work relative value units (wRVU) per visit. Secondary outcome include provider satisfaction survey, provider interview, and patient satisfaction survey.
Interventions
The tool (DAX) listens to the conversation between a clinician and patient/caregiver during their visit, provides transcriptions between different speakers for the clinician to review, edit, and included in the final EHR notes.
Sponsors
Study design
Eligibility
Inclusion criteria
* clinicians specializing in family medicine, internal medicine, and general pediatrics (including physicians and advanced practice providers \[APPs\]) * (intervention group) showed interests in participating in Dragon Ambient eXperience (DAX) Copilot * (control group) clinicians who opted out from Dragon Ambient eXperience (DAX) Copilot in vendor's baseline survey * (control group) clinicians who worked in the same clinic/region as the intervention group
Exclusion criteria
* preceptors * inactive DAX participants (i.e., had DAX license but never activate it) * have the same first use and last use DAX date * missing demographics
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| wRVU per visit | Month 10 | Total work relative value units (wRVU) in each day over total completed visits on the same day |
| Same day closure rate | Month 10 | Number of visits closed on the same day over total scheduled visits |
| Note length | Month 10 | Length of note in EPIC for each visit |
| Gross revenue per visit | Month 10 | Total charged amount in each day over total completed visits on the same day |
| Total EHR time | Month 10 | Total time on EHR (during and outside of clinic sessions) per 8 hour of patient scheduled time. |
| Work outside of work | Month 10 | Time on EHR outside of scheduled patient hours per 8 hour of patient scheduled time. |
| Time on encounter note documentation | Month 10 | Hours on documentation (note writing) per 8 hour of scheduled patient time |
| Completed appointment rate | Month 10 | Total completed visits over scheduled visits |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Provider interview | Month 1 | 12 semi-structured interview on clinicians in the intervention group. Each interview was facilitated using an interview guide with a semi-structured list of questions to help guide the conversation and answer the learning objectives. |
| Patient satisfaction | Month 6 | Patient satisfaction on the visit - patient satisfaction surveys are collected as part of routine care, which will be accessed through Medallia. Scores range between 0 (worst) to 10 (best) with one decimal point. |
| Provider satisfaction | Month 4 | REDCap survey to clinicians in both groups (intervention and control) and assess their satisfaction of the new tool. |
Countries
United States