Health App Documentation
For CliniciansClinical Documentation

Voice Transcription & AI Charting

Guide to using voice-driven clinical note creation with AI assistance

Voice Transcription & AI Charting

Overview

The voice transcription and AI charting feature allows clinicians to create clinical notes hands-free using voice input. The system guides clinicians through required documentation using AI-generated prompts and real-time transcription.

Key Features

  • Voice-Driven Data Entry: Speak responses instead of typing
  • AI-Generated Prompts: System generates prompts from form schemas
  • Real-Time Transcription: See your words transcribed as you speak
  • Structured Data Mapping: Responses automatically mapped to form fields
  • Guided Workflow: Step-by-step prompts ensure complete documentation

Getting Started

Prerequisites

  • Microphone access enabled in browser
  • Patient selected with active visits
  • Tasks available for documentation
  • Form schema configured for tasks

Accessing Voice Note Creation

  1. From Patient Chart

    • Navigate to patient chart
    • Click "Create Note" button
    • Modal dialog opens with visit selection
  2. From Task View

    • Open task that needs documentation
    • Click "Record Note" or similar button
    • Voice recording interface opens

Voice Note Creation Workflow

Step 1: Select Visit

  1. View Upcoming Visits

    • Modal shows all upcoming visits for selected patient
    • Visits are filtered by patient automatically
    • Select the visit you want to document
  2. Visit Information

    • See visit date and time
    • View visit type and status
    • Check associated tasks

Step 2: Select Task

  1. View Available Tasks

    • All tasks for selected visit are displayed
    • Tasks show type and requirements
    • Select the task to document
  2. Task Details

    • Review task type and description
    • Check form schema requirements
    • See any existing documentation

Step 3: Guided Prompt Mode

  1. AI Generates Prompts

    • System analyzes form schema for selected task
    • Generates conversational prompts using AI
    • Creates checklist of required documentation
  2. Prompt Display

    • UI enters focused "Prompt Mode"
    • Shows current prompt prominently
    • Displays voice recording controls
    • Shows progress through prompts

Step 4: Voice Recording

  1. Start Recording

    • Click "Start Recording" button
    • Grant microphone permission if prompted
    • Begin speaking your response
  2. Real-Time Transcription

    • See your words transcribed as you speak
    • Transcription appears in real-time
    • Can pause and resume recording
  3. Complete Response

    • Click "Stop Recording" when finished
    • System processes transcription
    • Response is mapped to form field
    • Prompt is checked off as complete
  4. Next Prompt

    • System automatically advances to next prompt
    • Continue recording responses
    • Repeat until all prompts are complete

Step 5: Review and Save

  1. Review Responses

    • Review all transcribed responses
    • Edit if needed (can type corrections)
    • Verify data is correctly mapped
  2. Save Documentation

    • Click "Save Note" or "Complete"
    • Structured data is saved to patient chart
    • Note is associated with task and visit

Voice Recording Controls

Recording States

  • Idle: Ready to start recording
  • Recording: Actively capturing audio
  • Paused: Recording paused (can resume)
  • Processing: Transcribing and processing audio
  • Complete: Response saved and mapped

Control Buttons

  • Start Recording: Begin voice capture
  • Stop Recording: End current recording
  • Pause: Temporarily pause recording
  • Resume: Continue paused recording
  • Cancel: Discard current recording

AI Prompt Generation

How Prompts Are Created

  1. Form Schema Analysis

    • System reads form schema for selected task
    • Identifies required fields
    • Understands field types and validation rules
  2. LLM Generation

    • AI generates conversational prompts
    • Prompts are clear and easy to answer verbally
    • Each prompt corresponds to a form field
  3. Prompt Checklist

    • All prompts displayed as checklist
    • Shows which prompts are complete
    • Indicates current prompt being answered

Prompt Types

  • Open-Ended Questions: "Describe the patient's condition..."
  • Specific Information: "What was the patient's blood pressure?"
  • Assessment Questions: "How would you assess the patient's progress?"
  • Action Items: "What interventions were performed?"

Data Mapping

Automatic Field Mapping

  1. Response Processing

    • Transcription is analyzed
    • System identifies relevant information
    • Data is extracted and structured
  2. Field Assignment

    • Responses mapped to corresponding form fields
    • Data types are validated
    • Formatting applied as needed
  3. Validation

    • System checks data completeness
    • Validates against form schema
    • Flags any missing required fields

Manual Correction

  • Edit Transcript: Can manually edit transcribed text
  • Adjust Mapping: Can reassign responses to different fields
  • Add Information: Can type additional information
  • Remove Data: Can delete incorrect responses

Best Practices

Recording Quality

  • Quiet Environment: Record in quiet space when possible
  • Clear Speech: Speak clearly and at moderate pace
  • Complete Thoughts: Finish complete thoughts before pausing
  • Microphone Position: Keep microphone at consistent distance

Documentation Completeness

  • Answer All Prompts: Complete all prompts for thorough documentation
  • Be Specific: Provide specific details rather than vague descriptions
  • Include Context: Add context that helps understand the situation
  • Review Before Saving: Always review before finalizing

Workflow Efficiency

  • Prepare Before Recording: Think about responses before starting
  • Use Consistent Terminology: Use standard medical terminology
  • Complete in One Session: Try to complete all prompts in one session
  • Save Frequently: Save progress if taking breaks

Troubleshooting

Common Issues

Microphone Not Working

  • Check browser permissions for microphone access
  • Verify microphone is connected and working
  • Try refreshing the page
  • Check system audio settings

Transcription Not Appearing

  • Check internet connection
  • Verify OpenAI API is configured
  • Wait a few seconds for processing
  • Try stopping and restarting recording

Prompts Not Generating

  • Verify form schema is configured for task
  • Check that task has associated form
  • Ensure AI service is available
  • Try selecting different task

Data Not Mapping Correctly

  • Review transcribed text for accuracy
  • Manually correct mapping if needed
  • Edit responses to match expected format
  • Contact support if mapping consistently incorrect

Recording Quality Issues

  • Check microphone quality
  • Reduce background noise
  • Speak closer to microphone
  • Use headset microphone if available

Technical Details

Audio Processing

  • Sample Rate: Audio is downsampled to 16kHz
  • Format: Converted to 16-bit PCM
  • Streaming: Audio streamed to transcription service
  • Real-Time: Transcription appears as you speak

Integration Points

  • Form Schema: Reads from task form configuration
  • Patient Chart: Saves documentation to patient record
  • Task System: Associates note with specific task
  • Visit Tracking: Links documentation to visit

On this page