Skip to main content

How to adjust the language setting for an encounter

Attend supports recording and transcribing encounters in multiple languages. You can set the language for each encounter from the recorder before you start recording.

Written by Attend Health Team

You can capture patient encounters in 36 languages plus an Auto-detect option. Your language choice controls how the conversation is transcribed (turned into text). It does not change the language of your clinical note — clinical notes and encounter titles are always written in English, while patient instructions are written in the language spoken during the visit.

Auto-detect is the default for every new encounter, so most users don't need to change anything.

How to select a language

You'll find the language selector before you start recording:

  • On the capture screen, open the language dropdown (top of the screen) and choose Auto-detect or a specific language.

  • The current selection is shown in the dropdown, and appears on the transcript as "Language: …".

If you know the visit will be in a specific language, selecting it directly (instead of Auto-detect) can improve accuracy.

How Auto-detect works

When Auto-detect is selected, the system automatically identifies the language being spoken and transcribes accordingly — you don't need to tell it the language in advance.

  • Detection happens continuously as the conversation is processed.

  • If the encounter contains more than one language, the system uses the language spoken most often to decide the language for patient instructions.

  • Auto-detect works well for a single primary language. For the most reliable results in a known language, selecting that language explicitly is recommended.

Supported languages

English, Spanish (Spain), Spanish (Mexico), French, French (Canada), German, Italian, Brazilian Portuguese, Portuguese (Portugal), Dutch, Danish, Swedish, Norwegian, Finnish, Polish, Czech, Romanian, Greek, Russian, Ukrainian, Turkish, Arabic, Hebrew, Persian (Farsi), Urdu, Hindi, Bengali, Chinese (Simplified), Chinese (Traditional), Chinese (Cantonese), Japanese, Korean, Thai, Vietnamese, Indonesian, and Filipino (Tagalog).

What language will my output be in?

Output

Language

Transcript

The language actually spoken during the visit

Clinical note

Always English

Encounter title

Always English

Patient instructions

The language spoken during the visit

Frequently asked questions

Do I need to set the language every time? No. Every new encounter starts on Auto-detect. Change it only if you want to force a specific language.

Will selecting the wrong language cause problems? If you force a specific language but the visit is in a different one, the transcript quality can suffer. When unsure, use Auto-detect.

Can it handle a visit with two languages? It can transcribe a conversation where speakers switch languages, and it will pick the most-spoken language for the patient instructions. However, there's no per-speaker or real-time language switching. For mixed-language visits, Auto-detect generally gives the best experience.

Why is my note in English when we spoke another language? This is expected. Clinical notes and titles are always produced in English by design, regardless of the spoken language. Only the transcript and patient instructions reflect the spoken language.

My patient instructions came out in another language — is that a bug? No. Patient instructions are intentionally written in the language of the visit so they're useful to the patient.

Does the transcript label show the detected language? The transcript shows your selection (for example, "Auto-detect"), which may not name the specific detected language.

Did this answer your question?