How AI-native tools support same-day urgent care visits

AI in urgent care: before, during, and after the visit

The clinician sees a walk-in patient they’ve never met. The chart may hold a decade of relevant history or almost nothing at all. Outside, the waiting area is filling up.

In the next several minutes, that clinician has to review the patient’s history, understand today’s concern, surface any medications or results that matter, conduct the encounter, document it, and close the chart.

It’s easy to misdiagnose this problem as one of visit length, as though the answer were simply to spend more minutes per patient. In practice, urgent care is constrained by the number of cognitive context switches a clinician has to make inside a single same-day encounter.

That is why one of the most useful questions about AI in urgent care is not whether it can write a clinical note. It’s whether it can reduce the number of times a clinician has to break concentration: before the visit, during, and after.

Same-day access is the point of urgent care. And the compression that comes with it—no pre-visit prep, no familiarity with the patient, a mixed bag of outside history—isn’t a flaw to be engineered away. Add high volume and a persistent need for clinically relevant data at the point of care, and the encounter becomes a series of interruptions instead of a continuous exchange. That’s where AI-native tools built into the clinical workflow can help, not by adding another system to check, but by reducing the number of times a clinician has to stop and switch context.

What follows is the encounter in the order a clinician actually experiences it. 

Before the visit: Preparing in minutes, not in advance

The pre-visit window in urgent care is measured in the time it takes to walk down a hallway. Whatever happens in that window determines how the encounter starts.

A shorter path to what's clinically pertinent

This is where the operational unlock is largest, and it is the part of the AI story that gets the least attention.

Tools like athenahealth's Chart Assistant and Intelligent Summaries work on the retrieval problem: the searching, scrolling, and cross-referencing required to answer a question the clinician already knows to ask, whether that concerns relevant history, active medications, or recent results. More information was never the difficulty. The gain is reaching the clinically pertinent information in fewer steps.

For a walk-in encounter, that changes the starting conditions. The clinician still takes the history but starts the conversation with the relevant background already in view instead of assembling it from scratch mid-visit. Fewer retrievals during the encounter mean fewer breaks in attention. 

Patient-reported information as context

Symptom information collected during intake arrives before the clinician does and gives the conversation a starting point instead of a blank one. Where AI surfaces clinically inferred diagnosis considerations from that information, they are presented for provider review. They serve as inputs to the clinician's thinking, which is what a human-in-the-loop framework requires.

During the visit: documentation that keeps the clinician in the room

Ambient clinical documentation addresses the most visible context switch of all: the one where the clinician turns to a keyboard mid-conversation.

Ambient Notes uses ambient listening and generative AI to capture the visit and produce a draft note that lands in the patient encounter. It lives inside athenaOne® instead of alongside it, and clinicians can choose among multiple ambient models to match their documentation style or specialty. The clinician reviews, edits, and signs. Nothing enters the record unreviewed.

The clinician-reported results are consistent. In an athenahealth survey of 153 Ambient Notes customers conducted in February and March 2026, 74% reported spending less time documenting¹ and 86% said they could give patients their full attention during visits.²

Independent research points the same direction. In a small prospective study at a single academic medical center, trained observers timed 169 consultations and found that documentation time fell from a mean of 5.3 minutes to 4.5 minutes, while the share of consultation time clinicians spent making eye contact rose from 69.6% to 77.1%.³ The sample is limited to nine clinicians and the setting is not urgent care, but it is one of the few studies to measure clinician behavior by direct observation rather than by survey or system timestamps. What clinicians report about their own attention, observers watching the room also measured.

Ambient models continue to develop toward analyzing the clinical dialogue to pre-populate diagnosis codes and orders, extending the same principle from capture into the administrative work that follows the visit.

After the visit: Review, close, and move on

The point of reducing rework at review is that the chart can close while the visit is still in working memory. In urgent care, where the next patient is already waiting, that window is narrow.

Among athenaOne clinicians evaluated after adopting Ambient Notes, three in five closed more encounters within one day of the visit, with an average improvement of 28%.⁴ That figure reflects providers who onboarded between February and May 2025, with three months of consistent use and at least 10 encounters per month. The population was not urgent care specific, and the result does not mean every provider improved by 28%.

Published research shows a similar pattern in other outpatient settings. A 2025 quality-improvement study of 46 outpatient clinicians across 17 specialties found ambient scribe use associated with 20.4% less time in notes per appointment and a 9.3% improvement in same-day appointment closure.⁵ The study had no control group, so the findings are associations rather than proof of cause. For urgent care teams, those are the measures that matter most: time spent in the note while patients are waiting, and whether the chart closes the same day it opens.

The closest operational analogue to urgent care comes from an emergency department study, where encounters documented with ambient AI required 28% less on-shift documentation time and 16% less total EHR time than those documented through standard workflows.⁶ Because physicians chose which encounters to use it for, and tended to use it for lower-acuity visits, some of that difference may reflect the visits themselves rather than the tool alone. Still, the emergency department shares much of what defines urgent care: high volume, unscheduled patients, and wide variation in acuity.

Taken together, these results support a realistic expectation of less documentation carried past the end of shift, and they point toward the same-day chart closure rates many urgent care organizations are working to improve. 

Oversight and integration by design

Everything above depends on a clinician’s review of what AI produces, which functions as a step inside the workflow. It applies at every stage where AI generates content.

Consent and notification practices for ambient capture work the same way. So does governance: how an organization sets policy, monitors output quality, and decides where these tools are appropriate. These are decisions made when the workflow is designed, not compliance steps added at the end.

That design is what makes the rest possible. Tools embedded in the EHR can draw on connected clinical information across the encounter. A standalone scribe sitting alongside the record can document a conversation, but it cannot prepare the clinician beforehand or reduce the rework afterward, because it has no view of the chart. 

Connected workflow, not a single feature

AI creates the most value in urgent care when it supports the connected work surrounding the encounter: preparing quickly, staying attentive, reviewing recommendations, and finishing documentation with less manual effort.

Each of those is a place where a clinician would otherwise have to stop, switch context, and start again. The actual gain is removing those breaks, not any one feature but the continuity across all of them.