Voice Dictation for Session Notes: A Workflow

By Tanner Oliver, LCSW ·March 25, 2026

You just finished a 50-minute session. Your next client is in 10 minutes. You have exactly enough time to use the bathroom, check your phone, and not write a session note.

So you don't. You tell yourself you'll write it later. And later becomes the end of the day, when you're sitting at your desk trying to reconstruct five sessions from memory. The details blur. The clinical nuance fades. By the time you're writing the note for your 2pm client at 7pm, you're documenting what you remember, not what happened.

Voice dictation breaks this cycle. Instead of writing, you talk. Instead of sitting at your desk, you use your phone. Instead of reconstructing from memory hours later, you capture the session while it's fresh — in the time it takes to walk to the bathroom and back.

But here's the thing nobody tells you: voice dictation isn't a separate piece of software you install. It's already in your EHR, in your phone's keyboard, and in your operating system. The right question isn't what voice-dictation app should I buy? — it's how do I work with the dictation tools I already have?

This post is about that workflow.

Where voice dictation actually lives

There are three places worth using.

Inside your EHR's mobile app. SimplePractice, TherapyNotes, Jane App, and most modern EHRs have a session-note field on their mobile app that accepts dictation through your phone's built-in voice-to-text. Tap into the note field, tap the microphone on your iOS or Android keyboard, talk. The transcription lands directly in the note that's tied to the right session. No copying, no exporting, no separate app. The note ends up in the same place your billing, scheduling, and clinical chart live.

iOS Dictation (or Android Voice Typing) in any text field. Every text input on your phone supports dictation by tapping the microphone key. Use it in Notes, in your EHR, in a Google Doc you'll paste later. The transcription quality on iOS 16+ is good enough for clinical work — it handles assessment names (PHQ-9, GAD-7), proper nouns, and natural pauses better than any third-party tool.

Dedicated dictation services (Otter, Rev, Dragon). Generally not worth it for session notes. They're built for meetings and journalism, charge per-minute, and rarely have a BAA available for a solo therapist. Stick to the free tools that are already on your phone.

The one workflow that doesn't make sense: dictating into a tool that isn't your EHR, then copy-pasting the result into your EHR. That's two extra steps and an audit-trail gap. Just dictate into the EHR directly.

What good dictation sounds like

You don't need to dictate in a formal clinical voice. The transcription handles capture; you handle structure on review. What you need to do is cover the key elements.

What the client reported. Their mood, their week, significant events, symptoms, concerns. This becomes the Subjective section.

What you observed. Affect, engagement, behavioral changes. This becomes Objective.

Your clinical thinking. What's working, what isn't, how the client is progressing. This becomes Assessment.

What comes next. Interventions to continue, homework, referrals, next assessment schedule. This becomes Plan.

A good dictation covers all four in about 60–90 seconds of talking. It doesn't need to be polished. It doesn't need to be in order. You'll edit on review.

Here's what a raw dictation might sound like:

> "Session with Jordan today. He came in talking about the argument with his partner over the weekend — said it triggered some of the old patterns we've been working on. Affect was flat, more withdrawn than usual. GAD-7 came back at 16, up from 12 two weeks ago. We spent most of the session on the argument — used the thought record to identify the catastrophizing pattern. He was able to see it by the end of the session but expressed frustration that he can't catch it in the moment yet. Plan is to continue cognitive restructuring, daily thought record for the next two weeks, revisit the GAD-7 at next session. No safety concerns."

That's about 40 seconds of speaking. You go back when you have a quiet five minutes, edit the transcription into proper SOAP/DAP/BIRP structure in your EHR, save the note, done.

Where Theracharts fits

Notice the part of that dictation that says "GAD-7 came back at 16, up from 12 two weeks ago." That's the part of the note that's actually data — the client's measured score, the trend, the clinical interpretation.

That part doesn't need to be dictated. It's already in Theracharts.

If you're using outcome tracking, you have the measured data sitting in your client's chart already. Score, trend, severity band, alerts when something crosses a threshold. The data-summary part of your session note is one Generate Clinical Update click away — it produces a 2- or 3-paragraph narrative grounded in measured data that you paste into your EHR's note field. You dictate the human-narrative part; the measurement layer fills in the rest.

The split looks like this:

Dictate: subjective content, observations, your clinical reasoning, plan
Paste from Theracharts: assessment scores with trend interpretation, completion patterns, goal progress, alerts triggered

Two minutes of dictation + one click of Copy + one paste = a session note that's grounded in real data without you having to reconstruct (or fudge) the numbers from memory.

The review step still matters

Voice dictation saves capture time. It does not save you from reviewing the output.

Transcription errors happen. A clinical term gets misheard. A number gets transposed. A sentence gets garbled because you were walking and the microphone picked up background noise. These are normal and expected.

The review step — reading the drafted note, correcting errors, adjusting language, adding nuance — is where your clinical expertise shows up. It takes two to three minutes and it's non-negotiable. A note that captures fast but introduces an error you don't catch is worse than a slow note.

Think of the review as editing, not writing. The hard part — getting the content out of your head and into structured text — is already done. You're polishing.

When to dictate vs. when to type

Voice dictation isn't always the best choice. Here's when each approach works.

Dictate when:

You're between sessions with limited time
You're mobile (walking, driving, between locations)
The session was straightforward and you can narrate it quickly
You process better verbally than in writing
You want to capture details while they're fresh

Type when:

The session was complex and you need to think carefully about your formulation
You're in a shared space and can't speak freely about client information
The note requires precise clinical language that you want to craft word by word
You're writing a note that will be shared externally (referral summary, legal documentation)

Most therapists end up using a mix. Quick, routine sessions get dictated into the EHR mobile app. Complex cases get typed at the desk. The flexibility to choose is the point.

Privacy and confidentiality

Speaking about client sessions out loud raises legitimate privacy concerns. A few practical guidelines.

Don't dictate in public spaces. The waiting room, a coffee shop, a shared office with thin walls — these are not appropriate places to speak about client information. Use a private room, your car with the windows up, or a quiet hallway where you won't be overheard.

Use HIPAA-compliant infrastructure. Your EHR mobile app handles dictation under its own BAA. iOS Dictation runs on-device for short snippets and doesn't transmit audio when in that mode. Anything longer can route through Apple's servers, which doesn't have a healthcare-specific BAA. The safe rule: dictate directly into your EHR's note field on the EHR's mobile app, not into a general-purpose app you're then going to copy from.

Don't leave local recordings on your phone. Voice memos, third-party transcription apps, and audio files of session content shouldn't sit on your device. Dictate-to-text directly; skip the audio-recording-then-transcribe workflow.

The compounding effect

The real value of voice dictation isn't any single note. It's what happens over weeks and months.

A therapist who dictates notes between sessions finishes their documentation during working hours. They don't take notes home. They don't spend Sunday evening catching up on a week's worth of charts. Their notes are written the same day — often within minutes of the session — which means the clinical detail is better, the documentation is more accurate, and the therapist is less burned out.

Layer the measurement piece on top of that — Clinical Updates that auto-generate the data-summary section of every note — and you spend even less time on the part of the note that's just transcribing numbers, and more time on the part that requires your clinical thinking.

That's the compounding effect: voice for capture, your EHR for storage, Theracharts for the measured data. Each tool doing the part it's best at.

Theracharts is the measurement layer that lives next to your EHR. 120+ validated assessments, auto-scoring, trend visualization, and Clinical Updates that paste straight into your EHR's session note. Get started free.