This document answers common questions about using Audio2Note.
Working with the Software
Q: Do I need to stay on the patient’s chart while my audio file is being processed?
A: No. Once you have successfully uploaded the audio file, the transcription and note generation process runs in the background. You can safely navigate to other patient charts or other areas of the application. The final note will be added to the patient’s encounter when the processing is complete.
Q: How long does it take for an audio file to be processed?
A: Processing time can vary. It depends on several factors, including the length of the audio recording, the context of the file, and the current load on the transcription service.
Q: What should I do if my audio file fails to process or an error occurs?
A: The system is designed to be robust, but errors can occasionally occur. If a job fails, you should first check the patient’s chart for any status updates or error messages. If the problem persists or you receive no information, please contact us. The software keeps detailed logs that we can review to diagnose the issue.
Q: What types of audio files can I upload?
A: .wav, .mp3, .m4a, .flac, and .ogg.
Accuracy & Responsibility
Q: How accurate is the transcription and the generated note?
A: The software uses advanced Artificial Intelligence (AI) to transcribe audio and generate clinical notes. While it is a powerful tool, it is not perfect. The generated text may contain factual errors, omissions, or even “hallucinations” (information that was not present in the source audio).
Q: Who is responsible for the final clinical note?
A: The licensed caregiver is solely and exclusively responsible for the clinical accuracy and appropriateness of all patient records. You must thoroughly review all output generated by the software for accuracy and completeness. It is your professional responsibility to edit, correct, or delete any information that is inaccurate, incomplete, or clinically inappropriate before signing and finalizing the note.
Q: How reproducible is the output from the AI?
A: The output is not perfectly reproducible, and this is an inherent characteristic of generative AI models. Just as two different healthcare providers might write a different assessment for the same patient presentation, two runs of the AI model on the same audio may produce different results. This variability is one of the key reasons why caregivers must treat the service as a clinical documentation aid, not a replacement for their own professional judgment. Always review and edit the generated text.
Privacy & Security (HIPAA)
Q: What patient information is sent to the transcription service?
A: To protect patient privacy, only the audio recording itself is sent to the external transcription service. No other personally identifiable information (such as the patient’s name, date of birth, or medical record number) is shared. If the summary function is used, three previous notes are temporarily stored on our servers only for the duration of the summarization process, and then are securely erased.
Q: Is the audio recording stored permanently? What happens to my data?
A: No. The audio file and the generated notes are permanently deleted from our server service as soon as the note is successfully created in the patient’s chart. This is a key security feature to minimize data retention and protect patient health information (PHI). The final note is stored securely within your system and is subject to your organization’s standard data retention policies.
Q: How is my data protected?
A: Your data is protected by multiple layers of security.
- Encryption: All data sent between your system and the external transcription service is encrypted during transit. Sensitive configuration settings within the software are also stored in an encrypted format.
- Business Associate Agreement (BAA): In case we need to scale up for more processing power, we will enact a formal Business Associate Agreement with the external service provider. This is a legal contract that requires them to adhere to strict HIPAA security standards to protect any data they process.
- Minimal Data Exposure: As mentioned above, we only send the audio file and permanently delete it after processing, ensuring minimal exposure of PHI.