Jefferson Health ambient AI users closed charts up to 39% faster

Philadelphia-based Jefferson Health has published one of the larger real-world looks at ambient AI documentation so far. According to a study published Oct. 23 in NEJM Catalyst Innovations in Care Delivery, as reported by Becker's Hospital Review, the analysis covered 5,234 clinicians with access to Abridge's ambient documentation tool. The rollout was associated with up to 16.6% less after-hours documentation and up to 39.3% faster chart closure among the most active users.

Those two measures sit at the center of the problem most health systems are trying to solve. After-hours documentation is the work physicians and advanced practice clinicians carry home, and it shapes how sustainable their jobs feel. Chart closure speed affects how quickly encounters move to coding and billing, which matters to the revenue cycle. A tool that moves both at once touches the Chief Medical Officer's retention concerns, the CFO's cash timing and the CIO's case for scaling or renewing a contract.

The evidence deserves a careful read. As reported, the study describes associations, not a controlled trial, so it does not establish that the tool alone caused the improvements. The largest figures, 16.6% and 39.3%, are described as upper bounds concentrated among the heaviest users. That leaves open how much benefit reaches clinicians who use the tool occasionally, and whether the most active users differed from others in ways that would have shown up regardless. Access across 5,234 clinicians also is not the same as consistent use across 5,234 clinicians.

That distinction is the practical lesson. If gains cluster among heavy users, value depends on adoption, not on licenses purchased. Paying for broad access while a minority uses the tool regularly would dilute any return, and enterprise-wide averages could hide both strong results and weak ones.

The move: set after-hours documentation time and chart closure speed, segmented by usage level, as the scorecard for your own ambient AI pilot or renewal. Establish a baseline before go-live, track the same metrics by user tier afterward, and tie expansion decisions to what each tier actually shows. Pair the numbers with support for clinicians who try the tool and stop, since moving more users into the active group is where the reported gains appear to live. Done this way, a renewal conversation rests on your own clinicians' time returned, not on a vendor's headline.

You are an operations advisor to the executive team of [your organization], preparing for an ambient AI documentation pilot or contract renewal.

Context: A study published Oct. 23 in NEJM Catalyst Innovations in Care Delivery, as reported by Becker's Hospital Review, covered 5,234 Jefferson Health clinicians with access to Abridge's ambient documentation tool. The rollout was associated with up to 16.6% less after-hours documentation and up to 39.3% faster chart closure among the most active users. The study reports associations, not a controlled trial. The largest gains are limited to the heaviest users. Access across 5,234 clinicians is not the same as consistent use across 5,234 clinicians.

Task: Build a scorecard for our pilot or renewal that tracks after-hours documentation time and chart closure speed, segmented by usage level. Include a baseline taken before go-live, the same metrics tracked by user tier afterward, and decision rules that tie expansion to what each tier actually shows. Add a measure for clinicians who try the tool and stop, and note what support they would need.

Output: A one-page scorecard for the executive team with these sections: Metrics and definitions, Baseline, Usage tiers, Decision rules for expansion, Owners and data sources, Risks and open questions.

Use my figures: [our ambient AI vendor], [number of clinicians with access], [our current after-hours documentation hours per clinician per week], [our current median time to chart closure], [how we define light, moderate and heavy users], [our pilot end or renewal date].

Guardrails: Use only the facts above and the slots I have filled. Cite NEJM Catalyst Innovations in Care Delivery and Becker's Hospital Review by name whenever you use their figures. Do not present the Jefferson Health figures as results we should expect. Flag anything uncertain or that needs verifying, including any slot I left blank. Use no patient data. Nothing here is medical advice. Label the result as a draft for the team to check.