Automated Medical Documentation with Ryan Plasch of Saykara

VFH Episode 32

In this episode, Teri welcomes Ryan Plasch, the VP of Growth and Strategy at Saykara, a company that is tackling the holy grail of medical documentation.

Ryan has been in healthcare for about 22 years with a specialty in sales and business development leadership. He is passionate about the transformational use of modern technology that can improve healthcare delivery.

Saykara is using their own speech recognition model and AI, to tackle the issue of clinicians being in a room and having to chart their notes on a computer while they are interacting with their patient. That AI-powered healthcare virtual assistant that simplifies the documentation process is actually called Saykara or Kara. Saykara listens to the interaction between a physician/clinician and their patient, and then transcribes the audio recording into the EHR.

Key points from Ryan!

Addressing one of the biggest pain points that physicians and clinicians have.

Saykara

It was formed in 2015 by Harjinder Sandhu.
Harjinder had been closely following some of the trends in the voice industry around speech, specifically Amazon, Apple, and Google.
It is a virtual assistant designed primarily to create a comprehensive clinical note during a patient and physician/clinician encounter. The physician/clinician doesn’t have to look at a screen anymore, and doesn’t have to click and use keyboards.
It is currently deployed over a smartphone.

How Saykara works

Saykara is integrated with all major the EHRs. The physician/clinician walks into the encounter with a patient as they normally would, and on their phone, they will have a worklist scheduled for the day as extracted by Saykara from the EHR. The physician/clinician will simply select that patient, set the phone down close them, and then turn on Kara (Saykara’s AI) and ask her to listen or press the ON button to start recording.
When the physician is done with the examination, he just presses OFF.

Automated Medical Documentation with Ryan Plasch of Saykara

The Natural Language Understanding (NLU) and Natural Language Processing (NLP)

Saykara primarily listens to the physician to record clinical notes, place orders, do referrals and other recommendations for the treatment plan.
They have their own speech recognition engine combined with their NLP platform and their own unique AI algorithms.
When Kara listens primarily to the physician, the phone app will buffer some of that and then synchronize to the cloud. They partner with industry leaders like Amazon and Microsoft, which are both secure, HIPAA compliant and encrypted.
All a physician needs is an internet signal (cell or WiFi)
They process information through their augmented AI where the machine learning does the first processing and then they have a team that QAs the data for intent, formatting, structure, and terminology to ensure 100% accuracy.
The QA process also serves as a feedback loop to help the machine learn, advance and understand at a much faster rate.
They collect the audio data, convert it to text, map it through NLP, and then build patterns/concepts based on the conversation flow.
Kara listens to what the physician verbalized out loud, what they are thinking as they go through the exam, and then with the patterns and concepts they build, they generate what they believe the physician’s intent is and then use the QA person in the loop to ensure that it is done accurately.
They also have the audio file which they listen to and their QA people listen to that as well to ensure that everything is accurate. They then generate a full complete billable note that is uploaded into the EHR.
A physician can review their notes from Kara as soon as they need them.
Kara has the ability to perform tasks like creating orders and to refer a patient.
Kara works with any EMR. They have an agnostic EHR interface engine and they can integrate with all the EHRs in a number of ways. They prefer the API approach because it allows for some very advanced functionality.

The Phone Versus Voice Assistant Hardware

Physicians move around a lot from room to room, and the only constant is that they always have their phone with them. They are available for texts, calls, and emergencies. Kara is a very easy platform just to run over consumer smartphones which gives Saykara flexibility. It doesn’t require a dedicated power supply so it is not limited by location and/or the need for a lot of hardware.
The sound sensitivity on smartphones is very good for creating clear audio to be able to generate a note using the AI platform.

Physician Burn-out

This is a very important area for Saykara to tackle.
It’s front and center in most of the health systems that they are working with, and it’s a public health crisis.
Statistics show more than 50% of physicians are already burned-out or experiencing burn-out. Two thirds won’t recommend their profession.
Saykara gives physicians their time back so they can improve their quality of life. The platform reduces their documentation time by 70% which really opens up doors for other activities that they couldn’t take care of otherwise.