AI use case

AI voice charting and end of shift note drafting for nurses

AI that lets a nurse document patient observations by voice at the bedside directly into the electronic health record, and that drafts a structured end of shift care plan note from the patient's chart for the nurse to review, edit and sign, so documentation happens during the shift instead of after it.

By Len Debets · Last verified 29 September 2026 · 2 public deployments

At least 50%
Reported cost reduction
Cedars-Sinai Medical Center, vendor claim.
81%
Reported cycle time reduction
Cedars-Sinai Medical Center, vendor claim.
USD 275,000 to USD 1.3 million
Indicative value per year
A hospital with 500 nursing full time equivalents documenting at the bedside. Worked example, see how it is calculated.

What problem does it solve?

Nurses spend a large share of every shift on documentation rather than at the bedside. Cedars-Sinai says studies show they spend up to 40 percent of their shift on documentation alone, which contributes to burnout and staffing pressure. Much of it happens after the fact: a nurse finishes a round of observations, moves to a workstation to type them in, and often finishes the shift's paperwork late, on overtime hours. An end of shift note written from memory at the end of a long shift is also less complete and less useful to the next shift than one built while the patient's status is fresh.

How does it work?

  1. Voice capture at the bedside. A nurse presses a button on a hospital issued phone and speaks an observation in plain language, such as a pain score or an intake amount.
  2. Structured mapping. The system parses the speech and maps it to the right fields and flowsheet rows in the electronic health record, sometimes filling several fields from one sentence.
  3. Nurse confirmation. The nurse reviews the parsed entry on screen and confirms it before it is written to the record; nothing is saved without that confirmation.
  4. End of shift note drafting. Separately, the system reads the shift's vitals, medication administration, orders and flowsheet entries and drafts a structured end of shift care plan note built around the goals set at the start of the shift.
  5. Review and sign. The nurse reviews, edits and signs the draft note before it becomes part of the permanent record and is handed to the next shift.
Audience
Employee facing
Autonomy
Copilot
Adoption
Early adopters
Channels
Internal tools, Mobile app

What is it worth?

Benchmarks are computed from the public deployments below: one data point per organization per KPI, with who made each claim.

Value benchmarks for AI voice charting and end of shift note drafting for nurses
KPIMedianReported rangeData pointsClaimed by
Cost reductionToo few to pool
at least 50%
11 vendor
Cycle time reductionToo few to pool
81%
11 vendor
Handling time reductionToo few to pool
85%
11 vendor
Satisfaction upliftToo few to pool
37%
11 vendor

Value drivers: Employee productivity, Lower cost to serve, Customer experience.

Indicative value

A hospital with 500 nursing full time equivalents documenting at the bedside

USD 275,000 to USD 1.3 million

Annual nursing time cost avoided from faster end of shift documentation per year

How this is calculated

Formula: nurses * minutesPerShiftSaved * shiftsPerYear / 60 * costPerNurseHour. The low scenario uses every low input, the high scenario every high input.

InputLowHighBasis
Nursing full time equivalents using the tool nurses, FTEs500500The reference size.
Minutes saved per nurse per 12 hour shift on end of shift notes minutesPerShiftSaved, minutes per shift410Epic's own average is about one minute saved per note; the low end applies that to an editorial assumption of about 4 end of shift notes per nurse per shift, replace with your own note volume. The high end uses Epic's own upper figure, 8 to 10 minutes per 12 hour shift, which Epic reports only "for some nurses", not as the typical result. Source
12 hour shifts worked per nurse per year shiftsPerYear, shifts per year150180Editorial assumption for a full time nurse working about three 12 hour shifts a week; replace with your own roster.
Fully loaded nurse cost per hour costPerNurseHour, USD per hour5585Editorial assumption; replace with your own fully loaded nursing cost.

What it leaves out: Counts only the end of shift note saving Epic reports. It leaves out any time saved on bedside flowsheet charting during the shift, where Cedars-Sinai's pilot with Aiva reports a separate reduction in the time between an observation and its documentation, and in incidental overtime, the cost of the software itself, and any change in note quality.

Who already uses it?

2 public deployments, strongest evidence first. Grades: A regulator or audit, B the organization itself, C vendor case study, D anonymous or estimate.

Mercy

United States · Healthcare · 2026

ProductionGrade C

Mercy, one of the 15 largest health systems in the U.S., rolled out Epic's Art AI assistant so nurses can generate end of shift care plan notes from the patient's chart instead of writing them from scratch. Art pulls vitals, medication administration records, flowsheets, orders and prior notes into a structured draft built around the goals set for the shift, which the nurse reviews and edits before signing.

  • Handling time reduction: 85%, per end of shift note
    "At Mercy, one of the 15 largest health systems in the U.S., average end-of-shift documentation time fell from 3.5 minutes per note to about 32 seconds—an 85% reduction—while the number of notes completed fully and on-time increased by 225%."
    Claimed by: vendor

Cedars-Sinai Medical Center

United States · Healthcare · 2025

PilotGrade C

Cedars-Sinai piloted the Aiva Nurse Assistant, a HIPAA compliant mobile app that lets nurses on a 48 bed surgical unit document patient observations by voice directly into Epic. Aiva's founder and CEO, Sumeet Bhatia, says Cedars-Sinai was the first health system to launch Aiva Nurse Assistant, developed with its own nurses through its Accelerator Program. Aiva Health's case study of the pilot, which involved more than 120 nurses, reports large reductions in the time between an observation and its documentation and in nurses' incidental overtime, alongside a rise in patient experience scores on nursing related questions. Cedars-Sinai's own newsroom separately confirms the pilot, the 48 bed surgical unit, voice dictation into 50 of the most commonly used Epic fields, and that a clinician validates each entry before it is filed to the record.

  • Cycle time reduction: 81%, pilot, more than 120 nurses
    "A pilot at Cedars-Sinai, involving over 120 nurses, reported an 81% reduction in time between nurse interventions or observations and documentation to the correct flowsheet rows, directly attributed to real-time ambient charting via Aiva's application."
    Claimed by: vendor
  • Cost reduction: at least 50%, pilot
    "The ability to complete charting during shifts significantly reduces the necessity for nurses to stay late, cutting incidental OT by more than half."
    Claimed by: vendor
  • Satisfaction uplift: 37%, pilot
    "37% Higher Patient Satisfaction: Increased nurse presence at the bedside and reduced computer time directly correlate with improved patient perception of care quality."
    Claimed by: vendor

How do you implement it?

A model agnostic playbook: what to prepare, the order to build in, and what goes wrong.

Data you need

  • A defined set of flowsheet fields and note templates the tool is allowed to write to
  • Access to the shift's vitals, medication administration record, orders and prior notes for the end of shift draft

Systems to integrate

  • Electronic health record, for flowsheets, the medication administration record and notes
  • Hospital issued mobile device or workstation on wheels for voice capture

Complexity: Medium

Voice to field mapping only works well when it is tightly scoped to the flowsheet rows and note templates the hospital already uses in its electronic health record. The practical work is agreeing that mapping, testing it against real bedside speech including accents and background noise, and building nurses' trust that confirming a parsed entry is fast rather than another chore.

  1. 1

    Start with one unit and a narrow field set

    Pilot on a single medical surgical or similar unit with a defined list of flowsheet rows, as Cedars-Sinai did on a 48 bed unit, before expanding hospital wide.

  2. 2

    Require confirmation before anything is written

    Show the nurse the parsed entry and require an explicit confirmation before it is saved to the record; never write directly from speech without that step.

  3. 3

    Build the end of shift note around the shift's own goals

    Structure the draft note around the patient goals set at the start of the shift, not only a list of tasks completed, so it is useful to the next shift.

  4. 4

    Track rejection and correction rates

    Monitor how often nurses reject or heavily edit a parsed entry or a drafted note, and use that to fix the mapping or the prompt, not only to judge adoption.

  5. 5

    Extend only after the first unit proves out

    Expand flowsheet coverage and additional units only once documentation timeliness, overtime and nurse feedback show the first unit benefited.

Guardrails

  • Nothing is written to the patient record without the nurse confirming the parsed entry or signing the drafted note
  • The tool only writes to a defined, agreed list of flowsheet fields and note types, not free text clinical judgment
  • Voice recordings and transcripts are handled under the same privacy and security rules as the rest of the electronic health record

KPIs to instrument

  • Documentation timeliness, the time from observation to charted entry
  • Incidental overtime hours per nurse per month
  • Rejection or heavy edit rate on parsed entries and drafted notes
  • Nurse reported satisfaction and patient experience scores on the pilot unit versus a comparable unit

Human in the loop

A nurse confirms every voice captured entry before it reaches the record and reviews, edits and signs every end of shift note. The tool never finalizes a clinical entry on its own.

Common failure modes

Misheard entries in a noisy ward
Background noise or an accent causes a wrong field or value to be parsed. Always show the parsed entry for confirmation and track correction rates by unit and shift.
A note that reads like a task list, not a story
A drafted note that only lists tasks completed is less useful to the next shift than one built around the patient's goals and trajectory. Structure prompts around goals and flag when a patient is not progressing as expected.
Coverage gaps push nurses back to the keyboard
If the tool only covers some flowsheet rows, nurses end up using two systems and lose the time saving. Prioritize the rows nurses use most before expanding breadth.

What are the risks and rules?

EU AI Act

Depends on design

Bedside voice charting only transcribes and maps a nurse's own observations for the nurse to confirm, which is not listed in Annex III and is usually minimal risk. The end of shift note drafter generates a structured document from the patient's chart, so the provider of that generative text can owe the Article 50(2) duty to mark the output as AI generated, unless an exception such as an assistive function for standard editing applies. Neither deployment makes a clinical decision or profiles the patient today, but a version that summarized or flagged clinical risk, rather than only structuring what already happened, could need assessment as medical device software under Article 6(1) and Annex I of the EU Medical Device Regulation. Health data captured by voice or generated in a note falls under GDPR Article 9 in every case.

Controls to put in place

  • Confirmation required before any parsed entry is written to the record
  • A defined, agreed list of fields and note types the tool may write to
  • Nurse review, edit and signature required on every drafted note before it is final

Frequently asked questions

Does AI voice charting replace manual documentation entirely?
No. At Cedars-Sinai, a nurse confirms each parsed voice entry before it is filed to the record. At Mercy, a nurse reviews, edits and signs each Art drafted end of shift note before it becomes part of the record. In both deployments the tool changes when and how the entry is made, not who is accountable for it.
How much time does it actually save?
Reported savings vary by what is measured. Epic reports that nurses using its Art tool for end of shift notes at Mercy cut average documentation time from 3.5 minutes to about 32 seconds, an 85% reduction. Aiva Health reports an 81% reduction in the time between a nursing observation and its documentation, and a 50% reduction in incidental overtime, in a pilot at Cedars-Sinai Medical Center.
Does it affect patient experience scores?
Aiva Health reports that patient satisfaction on Cedars-Sinai's pilot unit rose by more than a third on Press Ganey's nursing related questions, which it attributes to nurses spending more time at the bedside and less time at a computer.
What should a hospital pilot before expanding widely?
Start on one unit with a narrow, agreed set of flowsheet fields, as Cedars-Sinai did on a 48 bed surgical unit, and track documentation timeliness, overtime and the rate at which nurses reject or correct what the tool captures before expanding further.

How to cite this page

Blits.ai AI Use Case Library, "AI voice charting and end of shift note drafting for nurses", last verified 29 September 2026, https://www.blits.ai/ai-use-cases/nursing-voice-documentation. Licensed under CC BY 4.0. Method: how we verify use cases.

Changelog
  • 29 September 2026: First published

Related use cases

Healthcare

AI ambient scribe for clinical documentation

An AI scribe that listens, with the patient's consent, to the conversation between a clinician and a patient and drafts the clinical note, and often the letter or after visit summary, for the clinician to review, edit and sign in the health record. It documents; it does not diagnose or decide on treatment.

Deployments
3 public, best grade B
Reported handling time reduction
8.2%
Great Ormond Street Hospital for Children NHS Foundation Trust, organization claim
InsuranceHealthcare

AI for health insurance prior authorization and claims adjudication support

AI that reads prior authorization requests, medical claims and appeals with their clinical and billing documents, extracts diagnoses, treatments and costs, checks them against the policy and published clinical criteria, and prepares a summary and recommendation for a clinician or adjudicator, who makes every adverse decision.

Deployments
5 public, best grade B
Reported handling time reduction
about 50%
Acentra Health, vendor claim
Cross industryBanking

AI agent for IT service desk resolution

An AI agent in Microsoft Teams, Slack or the intranet that takes the high volume IT support queue, such as password and MFA resets, account unlocks, VPN, device and software requests, and resolves common requests by acting in the identity and IT service management systems, handing the rest to the right resolver group with the context attached.

Deployments
6 public, best grade B
Reported employee adoption
94%
Mercari US, vendor claim
HealthcarePharma and life sciences

AI clinical trial patient matching and prescreening

AI that reads structured data and clinical notes in the health record, compares each patient with the inclusion and exclusion criteria of open clinical trials, and gives research staff and treating clinicians a ranked list of likely eligible patients with the evidence for each criterion, so that people confirm eligibility and invite the patient.

Deployments
3 public, best grade B
Reported accuracy
100%
Cleveland Clinic, organization claim