Separating AI Hype from Clinical Reality AI Workshop | September 2026-Nobody Went to Medical School for This: AI and the Hidden Work of GI Practice

Pdf Summary

This case study argues that AI in gastroenterology should be used not just to improve efficiency, but to reduce the real cost of delivering care by eliminating hidden administrative work.

The practice began with four guiding beliefs: lower operating cost, automate as much as possible, use humans only for exceptions, and make healthcare simpler. It identified major “hidden work” in GI practice, including patient access, prior authorizations, revenue cycle tasks, practice infrastructure, and physician admin work.

Several real-world experiments showed large gains:

- <strong>Prior authorizations:</strong> Instead of humans logging into payer portals, an AI/RPA workflow cut care coordinator time from 8 hours a day to 2 hours and reduced medication approval time from 15 days to 3 days.
- <strong>Phone and fax workflows:</strong> AI handled routine inbound and outbound communication, reducing manual referral handling from 2-3 FTEs to one APP spending a few hours a day, while improving response time and conversion rates.
- <strong>Small operational tasks:</strong> Generative AI made previously uneconomical automation worthwhile, replacing expensive vendors and outsourced services for websites, reporting, accounting support, contract review, credentialing, and physician repetitive tasks.
- <strong>Revenue cycle:</strong> AI began handling coding, claim status, payment posting, and denials, with humans managing exceptions rather than routine work.

The biggest changes were operational: less repetitive work, faster workflows, and lower costs. Importantly, AI did not remove oversight or staff—it changed their role.

The central lesson is to stop asking, “How can AI help staff do this?” and instead ask, “Why are we doing this at all?” The goal is not more AI, but less unnecessary work and lower healthcare costs.

Asset Subtitle

Neil Gupta, MD MPH MASGE

Keywords

AI in gastroenterology

hidden administrative work

prior authorizations

revenue cycle automation

patient access

generative AI

robotic process automation

healthcare cost reduction

workflow automation

medical practice efficiency