Expectation vs Reality: Using Suki AI in a Dermatology Practice

When I first started working with Suki AI in Dr. Yoo’s dermatology office, I was excited about what it could do. Suki AI is an AI-powered voice assistant designed to help healthcare professionals draft clinical notes quickly and efficiently. It listens during patient visits, picks up on medical details, and creates a note that is integrated into the electronic medical record (EMR). The idea of it was that it would save time and help everyone, including doctors, medical assistants, patients, and more, by making the documentation process less of a burden.

At first, it was fascinating watching Suki AI transcribe conversations and generate notes in real-time. However, it didn’t take long for me to realize that it wasn’t as seamless as promised. For one, Suki didn’t integrate fully with our EMR system. This meant extra steps to manually check, edi, and transfer notes. This ended up taking as much time and sometimes even more than if we had just written them ourselves. In fact, a recent study evaluating whether AI-powered clinical documentation enhances clinician efficiency concluded that there was no statistical significance between using it or not using it.

Then, there were the mistakes. AI “hallucinations” is the term for when AI confidently gets things wrong, and Suki AI did that multiple times during our one month trial. I remember one case where a patient mentioned their relative was dealing with depression. Suki AI misinterpreted this and documented that the patient themselves had depression. It was a serious issue that could’ve led to inaccurate medical records if unresolved.

Suki also struggled to tell the difference between casual conversation and actual medical information. Patients often talk freely during visits, which threw the AI off. It often included details I would not include in the note due to relevancy.

Patients had mixed reactions, too. Some thought it was cool that we were using AI. Others worried about their visit feeling less personal or about what it meant for privacy. We even recorded a few video testimonials to capture their perspectives which you can check out below!

Thinking about the larger picture, the United States currently has fewer regulations in place for AI in healthcare compared to other industries or countries. As AI tools like Suki AI develop quickly, this leaves room for errors that can impact patient care. Implementing clearer standards and oversight could help ensure that AI tools are safe, effective, and fully ready to support medical professionals.

In the end, Suki AI was a learning experience for everyone in the office. The technology has so much potential, but it’s not quite there yet. For now, I’ve learned that while AI can be helpful, there’s no replacing the attention to detail and care that a person brings to the job.

By Suki (Medical Assistant) on Suki (AI)