Key Takeaways
- University Hospital at Downstate is piloting Sunrise Thread AI, an ambient artificial‑intelligence scribe integrated directly into its Sunrise EHR, with 50 physicians across family medicine, internal medicine, cardiology and the emergency department.
- Since May the tool has generated notes for more than 900 patient encounters; physicians review and approve each note before it becomes part of the official record, eliminating most post‑visit documentation work.
- Early feedback highlights the system’s robustness and ease of use, with CIO Dilip Nath noting it is “more robust than others while offering greater ease of use.”
- Across health systems, ambient AI scribes are cutting documentation time—examples include a 27% reduction in note‑taking per appointment at Intermountain Health and an estimated 15,000+ clinician‑hour savings reported by the AMA.
- Vendors are expanding the role of AI scribes beyond note‑creation to include pre‑bill review and coding, aiming to close the gap between bedside encounters and billing submissions.
- Downstate plans to broaden the rollout to inpatient and emergency‑department physicians as custom templates become available, reflecting an industry‑wide shift toward ambient AI as a core EHR function.
Ambient AI Scribe Pilot at Downstate
University Hospital at Downstate has launched a pilot of Sunrise Thread AI, an ambient artificial‑intelligence scribe built natively into the Sunrise electronic health record (EHR) by Altera Digital Health. The initiative involves 50 physicians representing family medicine, internal medicine, cardiology and the emergency department. According to Altera, the tool has already drafted notes for more than 900 patient encounters since its launch in May. The system listens to the clinician‑patient conversation, converts speech into a structured clinical note, and places that note directly inside the EHR for physician review and approval. Once the clinician signs off, the note flows into official documentation, coding and billing streams, dramatically reducing the need for after‑the‑fact charting.
Physician Feedback and Usability Metrics
Downstate’s chief information officer, Dilip Nath, praised the technology, saying the hospital found the tool “more robust than others while offering greater ease of use.” He added that demand from clinical staff remains high, indicating strong acceptance among the pilot physicians. The health system is currently tracking usability, transcription quality and the impact on documentation workflows, while Altera develops specialty‑specific templates for inpatient and emergency‑department providers. This feedback loop aims to refine the AI scribe’s performance before a broader deployment.
Impact on Documentation Time Across Health Systems
The Downstate pilot aligns with a growing body of evidence showing that ambient AI scribes cut clinicians’ documentation burden. PYMNTS reported in February that the University of Chicago Medicine experienced reduced documentation time and lower burnout after adopting similar technology, and that Sharp HealthCare and MaineHealth observed fewer after‑hours charting sessions. The American Medical Association (AMA) analyzed multiple AI scribe programs and estimated that participating health systems saved roughly 15,000 clinician hours.
Quantified Savings from Large‑Scale Deployments
Further data illustrate the scale of possible efficiencies. The Permanente Medical Group logged more than 2.5 million uses of its ambient scribe over one year and estimated a savings of 15,700 hours compared with non‑users, according to NEJM Catalyst. Intermountain Health reported a 27% drop in time spent on notes per appointment for clinicians who used Microsoft’s Dragon Copilot for ten or more encounters, a figure highlighted by the American Hospital Association. These findings suggest that even modest per‑encounter time reductions can accumulate into substantial workforce relief when scaled across large organizations.
Variability in Real‑World Outcomes
Not all studies show uniform benefits. A JAMA investigation of 1,800 physicians across five academic medical centers found that doctors saved an average of about 16 minutes of documentation time per eight hours of patient care, with notable inconsistency in how individual clinicians adopted the technology. STAT’s coverage of the study emphasized that while ambient AI can reduce charting load, its effectiveness depends on factors such as workflow integration, user training and specialty‑specific nuances. This variability underscores the importance of tailored implementation strategies, like the custom templates Downstate is developing for its inpatient and emergency teams.
From Note‑Creation to Billing Integration
The role of ambient AI is evolving beyond simple transcription. Abridge unveiled a pre‑bill review tool on September 14 that audits inpatient claims before submission, aiming to close the lingering gap between bedside encounters and billing—a process that often leaves coders chasing physicians weeks after discharge. As noted by HIT Consultant, the tool provides clinical documentation and coding teams with an early‑stage check, reducing denials and rework.
EHR Vendors Embedding Scribes Directly
Major EHR platforms are now embedding AI scribes as native features rather than add‑ons. Athenahealth, for example, added Microsoft Dragon Copilot alongside Abridge and Suki in its marketplace, per Fierce Healthcare. Altera’s Thread AI follows the same model, residing inside the Sunrise EHR where it can capture encounters from a desktop or mobile device and push the approved note directly into downstream coding and billing workflows. This tight integration distinguishes a true ambient scribe from a faster transcription service, as the approved note becomes the single source of truth for clinical, financial and regulatory purposes.
Future Expansion at Downstate
Looking ahead, Downstate intends to expand the ambient AI scribe rollout to its inpatient and emergency‑department physicians once the custom templates are finalized. The health system serves roughly three million Brooklyn residents with more than 800 physicians across 53 specialties, so even a modest increase in adoption could yield significant time savings and burnout reduction. As the industry continues to shift toward ambient AI as a core EHR function, Downstate’s experience offers a practical case study of how thoughtful piloting, clinician feedback and vendor collaboration can drive meaningful change in clinical documentation.
