Sometimes in a blog post, it’s much easier to show something than it is to write about something. That’s definitely the case with the recently announced EHR Virtual Assistant from NoteSwift called Samantha. That’s why I asked NoteSwift to create a demo video of Samantha at work so you could see what they’re doing. Check out the video demo of Samantha working with Allscripts Professional below.
When NoteSwift first reached out to me with this tool I told them that it sounded a lot like the voice recognition and NLP solutions that I’d seen previously. I remember one EMR a long time ago that had really deeply integrated voice navigation that got pretty close to this type of interface. Plus, I’d seen demos of NLP that would pull out the granular data elements from a narrative text before.
The key question for me was how tightly integrated the voice recognition and NLP technology was with the EHR software. As you can see from the demo above it’s quite integrated. I do still have some questions on what the learning curve for some of the specific voice commands will be for the NLP to work properly and document the visit the right way. Plus, similar to voice recognition I’m interested to know if the mistakes you have to correct are as time intensive as just clicking the boxes yourself. I’m sure there will be the full spectrum of experiences.
One thing that really impressed me about NoteSwift’s implementation of Samantha was the verification process that the doctor goes through near the end of the video (about 2 min and 12 seconds in for those keeping track at home). I’ve always thought that, at least for now, this was an essential part of using NLP in the medical world. The doctor still needs to verify that everything is accurate before moving on. The way NoteSwift has implemented this is quite slick.
In talking with Wayne Crandall, the President and CEO of NoteSwift, he also told me that Samantha can work with any input mechanism including voice recognition from Nuance or MModal. He even told me that some doctors believe they can type faster than they can do speech recognition which isn’t a problem for Samantha either. The real magic of Samantha is in taking a narrative text, however it’s produced, parsing the structured data, assigning the coding and entering it into the correct areas of the EHR.
Pretty slick solution and one that I think many doctors would like to try so they can stop their slow death by a million clicks.