Can nurses use speech recognition for charting?

Yes. Nurses can use speech recognition for charting, and it works best for free text: a shift summary, a handoff or a narrative about a conversation with a family. It works least well for flowsheets, which are grids of tick boxes and short fields. The evidence that voice charting saves nurses time is small. A 2024 literature review kept seven studies and suggested that speech recognition can speed up documentation by up to 15 percent ¹, particularly when it is integrated with artificial intelligence.

Free text suits speech, and flowsheets resist it

Speech recognition handles sentences well, and most free-text notes are sentences. A flowsheet is a set of discrete fields, where a nurse taps a value or selects an option. Speaking a value into a grid is often slower than tapping it, and a wrong number or unit in a clinical record is costly to find and fix. Free text is where dictation is most likely to help, and flowsheets are where it is least likely to.

Some electronic health record (EHR) setups already allow dictation into notes. One health system's guidance says that Epic's Haiku and Canto apps support Nuance Dragon dictation directly into progress notes started in the mobile app ². That describes one system's 2020 setup, so check what your own EHR and device support now.

Templates are the main risk in free text. A 2024 review found that excessive use of DotPhrases can make a clinical note excessively long and cumbersome ³. Dictate into a template that holds only the fields you must record.

The nursing studies show small, unproven savings

The nursing evidence is thin. The 2024 review searched PubMed, CINAHL and Google Scholar, found 156 studies, and kept seven that met its inclusion criteria ⁴. Its 15 percent is a ceiling rather than an average, because the review reports the gain as "up to" 15 percent.

A 2023 survey of 73 ICU nurses in Iran asked about the benefits of speech recognition ⁵. The most common benefits they named were paperwork reduction, performance improvement and cost reduction ⁶. The scores were close together, so the ranking means little, and the survey measured opinion, not time saved.

Accuracy takes time to learn, and the best data comes from radiology

Accuracy is the cost that a new user pays first. In a 2013 hospital study of radiology reports, new users of speech recognition corrected 11 characters per report, on average ⁷. That study is about radiologists, with specialized terms, and one speech system, so it shows the learning curve, not a nursing result. A first week of voice charting can feel slower than typing for that reason, so judge a tool after several weeks, not the first shift.

It depends on the note, the EHR and the unit

Speech is a good fit when the note is free text, the EHR accepts dictation into that note, and you can review the text before you sign. It is a poor fit when most of the shift is flowsheet entry, when your EHR does not accept dictation into that screen, or when your unit is too loud for a clear voice. Each of these cases is a reason to test in your own setting, not a reason to skip the tool.

What to do before you start

1. Check your EHR and privacy policy first. Confirm what the tool sends, where it is processed and whether your employer approves it for patient information.
2. Start with one free-text note type, such as the shift summary or handoff, rather than the whole shift.
3. Read back every dictated number, name and unit before you sign. Those are the errors that matter most in a record.
4. Measure a baseline for a week, then repeat it for several weeks with the same method, so you can judge the change.

For the wider picture, see dictation for nurses, therapists and allied health. For the time nurses spend charting, read how much time nurses spend on documentation. For the format of therapy notes, see how to dictate SOAP notes. For the general method, see the guide to how to dictate.

What the evidence does not show

We found no randomized trial of speech recognition for nursing charting in the sources we searched. The studies that exist are small, one of them is a survey of opinion, and the time figure is an upper bound from a review of seven studies. Until a larger study reports time saved on real shifts, the honest answer is that nurses have reason to try dictation for free text and no measured basis to expect large savings on flowsheets.