UToledo cuts open charts to under 30

Hospital charting is one of those problems that sounds boring until you see what it does to a health system. It eats physician time, pushes note-writing into nights and weekends, and slows down billing and follow-up. UToledo Health says it found a real fix — not by asking clinicians to type faster, but by putting ambient AI inside the visit itself. In an update published May 1, the system said its documentation backlog fell from more than 400 open charts to fewer than 30 after rolling out Nabla’s ambient documentation tool, building on an eight-week evaluation announced in late March. ### What is the actual thing they deployed? This is ambient clinical documentation software. It listens to the conversation between clinician and patient, turns that into a draft note, and drops the result into Epic, the health system’s electronic record platform. The point is simple — documentation happens during the encounter instead of becoming homework afterward. ### Why do “open charts” matter so much? An open chart is basically unfinished clinical paperwork. If too many stack up, doctors end up doing “pajama time” after hours, and the organization gets slower at coding, billing, and closing the loop on care. That is why the jump from 400-plus open charts to under 30 matters more than it might look at first glance — it signals that unfinished work stopped piling up. ### How big was the pilot? The evaluation ran for eight weeks. UToledo Health used it across 40 providers and more than 3,000 encounters, then decided the results were strong enough to expand more broadly. That matters because this was not a tiny demo with one enthusiastic department — it was large enough to test whether the workflow held up in routine care. ### Was it just backlog, or did speed improve too? Speed improved too. UToledo Health said time to chart closure fell 29% during the evaluation. The system also said documentation became more complete and coding more specific, which is the part administrators care about because better notes can translate into cleaner claims and more accurate capture of higher-acuity visits. ### Why is the workflow piece the real story? Because ambient AI only helps if clinicians do not have to leave their normal routine to use it. UToledo’s pitch is that the software worked because it sat inside Epic and because clinicians and vendor engineers kept tuning it together. Basically, they are arguing that “AI scribe” is not a standalone gadget purchase — it is an operations project. ### Does this mean ambient AI is solved? Not really. One health system’s results do not settle the whole category, and vendor case studies always highlight the upside. The more careful takeaway is narrower — if the tool is embedded well, adopted by clinicians, and measured against real operational pain, the gains can be large enough to show up in backlog, turnaround time, and revenue cycle performance at the same time. ### Why is this showing up now? Because health systems have spent the last two years moving from “AI sounds interesting” to “show me the operational metric.” UToledo’s numbers fit that shift. The headline is not that a model wrote notes. The headline is that a hospital tied AI use to fewer unfinished charts, faster closure, and cleaner documentation inside a live EHR workflow. ### Bottom line The interesting part is not the AI magic trick. It is the boring operational win. UToledo Health is treating ambient AI like infrastructure for clinical workflow, and the under-30 chart backlog is the proof point it wants the rest of healthcare to notice.