Voice Typing for Respiratory Therapists: Chart Faster Between Treatments

Short answer: Voice typing lets respiratory therapists dictate the narrative parts of documentation at speaking speed instead of typing them. Speech runs 130 to 150 words per minute against roughly 40 for comfortable typing, so treatment notes, vent check comments and weaning summaries get written while the details are still fresh.

Respiratory therapy has an unusual documentation shape. The work is spread across floors, the patient list changes hour to hour, and a meaningful part of the job happens standing next to equipment rather than sitting at a desk. You do not get one long charting block. You get a treatment, a walk, another treatment, a rapid response, and whatever minutes are left over.

The numbers and dropdowns in a flowsheet are fast to fill in. What eats time is the prose: the assessment between pre and post, the reason a treatment was held, the comment explaining why a setting changed, the education you gave a family who is about to take a nebuliser home. That writing is where voice typing earns its place, because it is the only part of the note where your speaking speed is actually the binding constraint.

What actually has to be written down

Across acute care, pulmonary diagnostics, and home or sleep services, the narrative load tends to cluster into a handful of recurring kinds of writing:

• Treatment notes with pre and post assessment. Breath sounds before, breath sounds after, tolerance, effort, cough quality, what the patient produced, whether anything changed.
• Ventilator check comments. Not the numbers, which the flowsheet takes, but the reasoning: why a setting was adjusted, what the patient was doing when the alarm fired, how synchrony looked.
• Weaning and spontaneous breathing trial summaries. Start and stop conditions, how the patient tolerated it, the judgement call about continuing.
• Airway and secretion management. Suctioning, cuff checks, tracheostomy care, what was recovered and how the patient responded.
• Held or refused treatments. Often the most important narrative you will write that shift, and the one most likely to get written hours late from memory.
• Escalations and events. Your contribution to a rapid response or code, intubation assists, what you observed and when.
• Patient and family education. Inhaler and spacer technique, incentive spirometry, airway clearance devices, home equipment setup and troubleshooting.
• Diagnostics and reports. Pulmonary function testing comments, quality of effort, anything that qualifies how the results should be read.

None of that is long. Each item is two to five sentences. The problem is volume and timing, not length. Fifteen short narratives at three minutes of typing each is forty-five minutes of your shift, and those minutes get borrowed from the end of the day when your recall of the third patient is at its worst.

Why typing is the bottleneck, specifically

A comfortable adult typing speed is around 40 words per minute. Skilled typists reach 80 to 100. Ordinary conversational speech runs 130 to 150, and you already produce it without training.

On a respiratory floor, the practical gap is wider than those figures suggest, because your typing conditions are bad in ways an office worker's are not. You may be gloved. You may be at a shared workstation on wheels with a keyboard at the wrong height. You may have just walked in from an isolation room. Every one of those drags your effective typing speed below your real one, while your speaking speed is unaffected.

There is a quality effect too, and it is the one clinicians tend to notice first. When writing is slow, people compress. A held treatment becomes "held per patient request" instead of the two sentences that actually explain the clinical picture. Dictating the same note takes twenty seconds, so the full version gets written because the full version is no longer expensive. That is the same pattern we saw writing about voice to text for nurses: the gain shows up as completeness at least as much as speed.

The terminology problem, and how to solve it once

The first objection any respiratory therapist raises about dictation is vocabulary, and it is a fair objection. The RT lexicon is dense with abbreviations that are pronounced as letters, as words, or as some hybrid nobody agrees on. FiO2. PEEP. SIMV. PRVC. APRV. HFNC. BiPAP. NIF. RSBI. ABG values with four components. Drug names that are said one way and written another.

Two things make this manageable.

First, modern transcription handles clinical speech far better than the desktop dictation software many clinicians gave up on years ago. If your mental model of voice typing was formed by a tool that needed you to speak in a flat, deliberate cadence, that model is out of date.

Second, and more usefully, you do not have to rely on the engine getting your local conventions right. Smart Vocabulary in Voice Keyboard Pro is a personal dictionary with replacement rules, which means you teach it your written forms once and it applies them from then on. You say the thing the way you actually say it out loud, and the text that lands is the thing you would have typed:

• Say "peep" and get PEEP, capitalised, every time, rather than the English word.
• Say "fee oh two" and get FiO2 in your house style.
• Say "duo neb" and get DuoNeb with the capitalisation your department uses.
• Say your unit's shorthand for a protocol and get the full written phrase it expands to.

The point is that this is configuration, not hope. Spend fifteen minutes on the twenty terms you use hourly and the vocabulary objection mostly stops being a real one. Build the list from your own notes rather than from a generic medical word list, because the abbreviations that matter are the ones your facility actually writes.

A shift, reworked

Here is what changes in practice, described concretely rather than as a promise.

Start of shift

You take handoff and want your own working notes, separate from the chart: who is on which mode, who is trialling, who is likely to extubate today, who the family concerns are about. Dictating that list into a notes app takes a minute or two instead of ten, and because it is fast you actually make it instead of trying to hold it in your head.

Between treatments

This is the core win. The narrative for a treatment gets dictated in the corridor or at the workstation immediately after you leave the room, while the breath sounds are still clear in your memory. Twenty to thirty seconds of speaking replaces two to three minutes of typing, and the gap between doing and recording shrinks from hours to seconds. Accuracy of recall is the underrated benefit here. A note written at the moment is simply a better note than the same note reconstructed at 18:45.

After a vent change

The flowsheet captures the settings. What you dictate is the reasoning, which is the part that a physician reading the chart tomorrow actually needs and the part most likely to be skipped when writing is slow.

After an event

Following a rapid response or an intubation, you have a sequence to record and a memory that is degrading fast. Dictating a rough chronological account first, then cleaning it into the formal note, is far more reliable than composing the formal version directly at a keyboard while still catching your breath.

End of shift

The goal is that there is nothing waiting. Most of the narrative was written when it happened, so what remains is review rather than composition.

Fixing a note without retyping it

Dictated text needs editing, and editing is where a lot of voice workflows fall apart: you save two minutes writing and then spend three minutes clicking around fixing a number.

On iPhone, Voice Edit addresses this directly. You speak the change you want rather than navigating to it by hand. Correcting a value or rephrasing a sentence becomes another spoken instruction instead of a precision cursor operation on a small screen with one free hand. For anyone charting from a phone between rooms, that is the difference between a workflow that survives contact with a real shift and one that does not.

Handoff, huddles, and in-services

Respiratory therapists sit in more structured conversations than the job description suggests: shift handoff, ventilator rounds with the intensivist team, care conferences, competency in-services, departmental meetings.

Meeting Mode on Mac handles speaker detection and produces AI notes, and calendar meeting detection means it recognises when a scheduled meeting is beginning rather than waiting for you to remember. For an in-service or a departmental meeting, that turns an hour of half-attention note-taking into a usable summary you can actually search later. It is a meetings feature rather than a charting tool, and it is worth being clear about that distinction, but the category of "structured conversation I need a record of" covers a real part of the week.

Where voice is the wrong tool

An honest assessment matters more than an enthusiastic one, because the fastest way to abandon a tool is to use it for the thing it is worst at.

• Structured numeric fields. Tabbing through a flowsheet entering values is a keyboard and keypad job. Dictating into a grid of small numeric cells is slower and riskier than typing them, and the risk is the kind you do not want in a vitals column.
• Dropdowns, checkboxes, and pick lists. These are selection, not text entry. Voice adds nothing.
• Anything you have not read back. Dictated text needs the same review a typed note needs. A transposed number reads perfectly fluently, which is exactly what makes it dangerous. Proofread before you sign, every time.
• Conversations you should be having instead. Charting in front of an anxious family while they wait is a worse use of the time saved than sitting with them for an extra minute.
• Anywhere your facility's policy says no. Which brings us to the next section.

Working inside your facility's rules

Clinical documentation sits inside a compliance framework, and this is not a topic to be casual about.

What we can tell you precisely is how the product behaves: the server stores only operational pings. No audio and no transcript content is kept on it. That is a deliberate design choice about what the service is allowed to know, and it is the right question to be asking of any tool you put near clinical work.

What we cannot do is tell you what your employer permits. That decision belongs to your organisation, and the answer varies enormously between a hospital system with a locked-down EMR and a home care or sleep practice where you control your own machine. Before you use any dictation tool for patient documentation, check your facility's policy on third-party software and on device use, and ask rather than assume. Many therapists start with the parts of the job where the answer is unambiguous: their own working notes, education materials, continuing education writing, competency documentation, and email. Those build the habit with none of the ambiguity.

Accuracy in a loud room

Respiratory environments are noisy by nature. Ventilator alarms, high-flow systems, suction, monitors, and a busy corridor all sit in the background. A few practical habits help:

• Step out of the immediate noise. Two metres away from a high-flow setup is a meaningfully different acoustic environment.
• Speak at a normal conversational pace. Over-enunciating and slowing down is a habit from older dictation software and it makes results worse now, not better. Speak the way you speak.
• Dictate in complete sentences. Surrounding context is what disambiguates a clinical term. Word-by-word dictation removes the very signal that resolves ambiguity.
• Use one mic setup consistently. Whatever you use, use the same thing. Consistency is worth more than an upgrade.
• Say punctuation deliberately. It becomes automatic quickly and it is what separates a clean note from a wall of text.

None of this is specific to respiratory therapy, and the general advice we gave in voice to text for physical therapists applies here, since both roles document in motion between patients rather than at a fixed desk.

How the two apps fit a respiratory workflow

Voice Keyboard Pro exists in two places, and they cover different halves of the job.

On Mac, it lives in the menu bar. You hold a hotkey, speak, and release, and the text appears at your cursor in whatever application you are already using, system-wide. There is no separate window to open and nothing to copy across, which is what makes it usable in short bursts rather than only in dedicated writing sessions. Smart Vocabulary, Meeting Mode with speaker detection and AI notes, and calendar meeting detection are all on this side.

On iPhone, it is a custom keyboard with a microphone button built in, so dictation is available in any iOS app rather than only the ones that chose to support it. Voice Edit, two-way translation while dictating across 24 languages, and swipe typing are on this side. The translation feature is worth a specific note for therapists working with patients and families who do not share a first language: it is genuinely useful for informal communication and for preparing written education material, and it is not a substitute for a qualified medical interpreter when clinical instruction or consent is involved. Treat it as a convenience, not a clinical service.

There is a free tier with daily limits, which is enough to find out whether this fits your shift. Pro is $4.99 a month or $34.99 a year.

Getting started in about ten minutes

1. Install and set a hotkey you will not hit by accident. Something you can reach one-handed, since one hand is frequently occupied.
2. Practise on something unimportant. Dictate an email or a shopping list. The goal is to get past the initial self-consciousness of talking to a computer, which takes most people a day and then never returns.
3. Build your first vocabulary list from real notes. Open three of your own recent narratives and add every abbreviation and drug name in them. Twenty entries covers most of what you write.
4. Start with one note type. Treatment narratives are the best first target: high frequency, short, and formulaic enough that you will feel the difference within a shift.
5. Expand only once the first one is automatic. Changing everything at once is how tools get abandoned in week two.

What to expect in the first two weeks

The first day is slower. That is normal and it is not a sign the tool is wrong. You are learning to compose out loud, which is a different skill from composing at a keyboard, and it takes a few days before your spoken sentences come out already shaped like a note.

By the end of the first week, the short narratives are faster than typing and your vocabulary list has stopped growing much. By the second week, the change most therapists report is not a stopwatch figure but a structural one: documentation happens next to the patient encounter instead of accumulating into a backlog you clear at the end of the shift.

That is the real argument. The speed difference between 40 and 140 words per minute is what makes it possible, but the thing worth having is closing the gap between doing the work and recording it, while you still remember it exactly.

The best time to write a treatment note is the thirty seconds after you leave the room. Voice typing is what makes that window long enough to be useful.

If you want to see whether it holds up on a real shift, the free tier is enough to test it against your own note types. Start with treatment narratives, build the vocabulary list from your own writing, and check your facility's policy before it goes anywhere near a chart.