Healthcare is witnessing the dawn of a new era, one in which ambient listening technologies make it easier for providers to document care for their patients and complete, check, and lock their medical notes more quickly than ever before.
But just as dawn is only the beginning of the day and there are many hours ahead, ambient listening and AI-powered medical scribes display amazing promise while still needing finetuning in order to improve their skills.
The culture of safety behind Sunoh.ai
The developers and product experts behind Sunoh.ai medical scribe understand that any new (or existing) technology must be guided by human judgment and repeatedly tested for accuracy and effectiveness.
That’s why Sunoh.ai was created with both clinical usefulness and patient safety in mind. From design to implementation, Sunoh’s ambient listening technology focuses on accurately transcribing and summarizing the content of the patient-provider encounter while avoiding common pitfalls that can characterize some AI-powered solutions.
As Jay Anders, M.D., Chief Medical Officer at Medicomp Systems, noted in a recent column for Physicians Practice: “AI-powered ambient listening – though very promising – has a way to go in terms of creating accurate, verifiable and reliable documentation.”
Let’s look at ways that Sunoh.ai sets an industry standard for safety and security.
Patient data is not shared outside Sunoh.ai
Many AI-powered products these days — including many educational tools — are trained on data, and lots of it. Developers and organizations don’t always check to see that they are legally permitted to use the data they are feeding into their models and may not prioritize the privacy and security concerns of the individuals whose data they are using.
Sunoh.ai is different. What happens in Sunoh stays in Sunoh. Data is stored in a private cloud run by Microsoft Azure® and is never shared outside the application. Nor is patient data used to train Sunoh.
This strict “no-share” policy is designed to protect the privacy of patients and limit use of patient data to the physician and other authorized parties.
Steps to verify the identity of every patient
In order to ensure that each encounter’s information is associated with the correct patient, Sunoh starts with two-factor authentication. Physician users must log in and verify their identity with a code sent to their smartphone or an email link.
Once the physician accesses Sunoh, the application requires confirmation of the patient’s first and last names and date of birth. And if a practice has multiple patients with the same name and date of birth — which is surprisingly likely in many large systems — the user is prompted to check photos and other demographic information to confirm the identity of the specific patient being seen.
Prompts to ensure providers review their notes
Once Sunoh generates a summary for an individual patient-provider encounter, the program prompts the provider to review the chart, make any necessary edits, and save it.
While Sunoh achieves a high degree of accuracy in capturing clinical details, no AI program is perfect, and the draft clinical notes Sunoh produces must always be reviewed by a qualified medical provider to ensure accuracy and safeguard patient safety.
What happens after the patient encounter
When it comes to what happens to Sunoh data after an encounter takes place, there are two key points to emphasize.
First, it should be clear that the “no share” policy regarding Sunoh data extends the entire length of the patient journey. Data isn’t used by anyone other than the provider for treatment of that patient. It isn’t shared with any vendor. It isn’t used for training or learning purposes.
Second, the Sunoh recording and summary are deleted after seven days. The only portion that is saved is the clinical note as reviewed and edited by the physician.
That information becomes part of the patient’s record in the EHR and serves as the only permanent record of what took place during that encounter. If the physician — or another authorized medical provider who deals with that patient at some point in the future — needs to make edits, add information, or otherwise alter the record, they must do so in the EHR.
That process ensures that the patient’s permanent record in the EHR serves as the “single source of truth,” thus avoiding conflicting information, diagnoses, or prescriptions, any one of which could conceivably lead to patient harm.
Safety standards in the software development lifecycle
Safety is the first consideration at every stage in the development, deployment, and enhancement of Sunoh.ai medical scribe. Sunoh’s developers follow strict design standards throughout the software development life cycle (SDLC).
Clinical Informatics and Patient Safety specialists — including trained medical clinicians — regularly review Sunoh.ai to ensure that usability and patient safety go hand in hand.
The result is a product that is easy to use, with easily edited notes that can be completed quickly and accurately, thus achieving the dual purpose of comprehensive documentation while reducing cognitive load on physicians and saving them time and money.
And when it comes to training and improving Sunoh, developers use dummy data that accurately reflects real-world clinical data without ever using actual patient data or placing such data at risk in any way.
HIPAA and other administrative compliance
In addition to safety procedures throughout the SDLC, Sunoh implements business associate agreements with users as required under the Health Insurance Portability and Accountability Act (HIPAA), as well as all administrative, physical, and technical safeguards required by that legislation. And Sunoh uses industry-standard encryption and security protocols.
To learn more about Sunoh’s safety features and capabilities, visit sunoh.ai.
Steps to verify the identity of every patient
Prompts to ensure providers review their notes
HIPAA and other administrative compliance
