Case study 2: A Neurologist’s Transition from Transcription Services by use of AI tools

I spend most of my time researching, testing, and implementing practical AI tools that best suit different clinical workflows in efforts to improve operational efficiency.

I’ve come to understand that an AI tool that would work well for Dr X. doesn't always guarantee to produce the same results in Dr Y's practice even if they're in the same specialty.?Tailored guidance is sometimes necessary.

My goal here is to shed some light on the different use cases I encounter, as I continue to implement AI tools across different specialties. Hopefully this can give you some insights as to how useful some of the AI tools are in a clinical setting.

Now, if after reading this, and would like more specifics to be covered on the subject, kindly indulge me in the comments or send me a dm.

Lets dive right in to this physician's experience.

What were his challenges??

• Reliance on transcription services costing ~$30 per patient.

• Low quality notes.
• Needed to leave the office earlier.

• Manual note-taking for follow-ups at certain semi private clinics.
• No unified process for EMG-focused and consultation-focused clinics.

He saw about 20–25/day, including 8 new consults while having to spend up to 10 mins to document each patients visit. I know that number of patients might be standard practice for some, but that documentation time does add up per week.

What was the solution??

After several hours of consultation to analyze his workflows, I recommended piloting 3 AI scribe platforms. These tools were tested across his diverse practice environments, which included:

1. Regular consults and follow up for example "Migraine" patients
2. EMG clinics using three rooms and a single charting space.

The chosen platform offered:

• Ability to create templates.
• Rapid SOAP note generation (average 8 seconds).
• Auto language capture, meaning being able to capture French and English (for example) during a single conversation and generating the note in your preferred output language.
• Features like smart dictation, allowing for non-verbatim note creation with automatic punctuation.

What were the results??

• Cost Savings: Reduced transcription costs by $300-400 daily.
• Improved Accuracy: Notes were more precise and organized.
• Personalized Experience: The neurologist appreciated receiving a brief instructional video from the product's CEO, adding a human touch.

What were the Challenges During the Pilot, if any ??

• AI occasionally hallucinated which is "normal" across most platforms.
• Always excluding the patients age from the summary even when repeated several times - this was solved by use of templates.
• Logging into some of the ai tools at the beginning were sometimes "tricky" - I'd say this was more of a human than systems error. Some basic clean-ups helped such as clearing your browser cache from time to time or downloading the ai chrome extension depending on the platform.
• Templates need to be further refined to optimize note structure.

Future Goals:

The Neurologist aims to:

• Incorporate more structured templates for frequent conditions.
• Transition fully to the AI platform for all documentation, ensuring no dependency on transcription services.

Now here are some key points to note from this use case;

Implementing AI tools in healthcare isn’t just about adopting new technology; it’s about aligning the solution to your unique workflow. These examples highlight:

1. AI scribes can significantly reduce documentation time and costs while improving note accuracy.
2. Based on your specific workflows, piloting multiple tools allows you to? identify the best fit for your needs. You should also be asking yourself, How seamlessly does the tool integrate? Are you able to create specialty specific templates? How intuitive is the implementation? Does it integrate with my EMR? Customization, such as tailored templates, enhances the technology’s impact.

AI-powered tools aren’t one-size-fits-all. These success stories illustrate how personalized consultation and thoughtful implementation can transform practice workflows while addressing the unique challenges of different specialties.

If you’re curious about how AI tools could work for your practice, consider assessing your current workflows and identifying areas for improvement. Sometimes, even small changes lead to transformative results.