Hospitals' New AI Assistant Reads Patient Records, Saves Doctors Time

Imagine your doctor spending less time digging through paperwork and more time with you. That's the promise of Scout, an AI tool that searches electronic health records and summarizes patient information. It was tested across a large hospital system for 11 months. Over 2,600 users, mostly doctors and nurses, completed more than 80,000 tasks with it. The tool became popular fast: on launch day, daily use jumped 21 times compared to before. Most people who took the required training started using it within two days.

Scout works like a smart search engine for patient records. It can find specific details, summarize long notes, and even draft parts of notes. It handles different types of requests: pulling up lab results, summarizing a patient's history, or extracting key facts. Users saved 752 tasks for reuse, meaning they found it helpful enough to come back to. The tool was used across more than 20 medical specialties, showing it's useful for many kinds of doctors.

But there's a catch. The AI isn't always right—it can miss details or make mistakes. That's why doctors must check its work. Also, using it requires training, and not everyone who trained ended up using it. The study found that a small group of heavy users did most of the work, which means some may find it less useful. And while it saves time, it's not a replacement for medical judgment.

For patients, this could mean doctors have more time to focus on care. For hospitals, it's a way to make better use of electronic records. But it's still early—we need to see if it truly improves care or just speeds things up. The key is that doctors stay in control and double-check the AI's answers.