Podcast Summary: Risk Never Sleeps Podcast

Episode #153: How Tiny Workflow Tweaks Can Reduce Massive Physician Burdens, with Jason Hill & David Leingang
Host: Ed Gaudet
Guests: Dr. Jason Hill (Innovation Officer, Ochsner Health), David Leingang (Director of Innovation Data Science, Ochsner Health)
Date: December 8, 2025

Episode Overview

This episode explores how targeted workflow adjustments and data-driven innovation at Ochsner Health are dramatically reducing the burden on physicians—particularly the avalanche of patient messages (the “in-basket” problem). Dr. Jason Hill and David Leingang both discuss the intersection of technology (especially AI/ML) and practical changes, the process of transforming data into operational improvements, and why many healthcare improvements don’t require complex tech—just thoughtful process changes.
The conversation also touches on the cultural and organizational challenges of digital transformation, the importance of problem scoping before deploying AI, and the power of collaboration between academic, healthcare, and industry partners.

Key Discussion Points & Insights

1. The Scale of the Physician Message Burden

• Ochsner Health receives roughly 2.4 million messages annually, a significant portion of which reaches providers unnecessarily.
• GLP1 (Ozempic and related weight loss medications) queries made up about 4% of all patient messages—thousands each year.

  "About 4% of all our patient messages had something to do with a weight loss medication." – David Leingang [01:43]

2. How Ochsner Approached the Problem

• Data-driven Triage: Started with a comprehensive analysis of all system-wide messages, building models to discover trends and frequent topics.
• Developed a machine learning approach to route messages more intelligently (though not fully deployed—waiting to see how Epic, the EHR provider, advances similar functionality).
• Operational Response: Insights from data led to operational changes—resulting in a new digital medicine, weight management program to support appropriate GLP1 prescribing.

  "We actually created a weight management program that was part of our digital medicine offerings at Ochsner." – Dr. Jason Hill [03:32]

3. Do You Need AI? Sometimes, No.

• Workflow before AI: The team stresses the value of basic workflow changes and provider education before turning to resource-heavy AI projects.

  "Do we really need to build a model...or do we just do a little education and then our physicians take over and we've accomplished the same thing without spending that time and money?" – David Leingang [04:18]

• Simple Tweaks with Big Impact: For example, Ochsner reduced provider message overload by changing the order of message options in their portal:
  • Moved “send a message to my doctor” to the bottom.
  • At the top: “I have a new problem that needs medical attention”—which triggers an eVisit, capturing detailed info upfront.

  "We actually reduced our click share by about 10% by just changing the order of the messages." – Dr. Jason Hill [05:24]

4. Measuring the Right Metric

• Key realization: Time spent in the in-basket isn’t the right metric; fewer received messages is much more important.

  "The best message I receive is the message I don’t receive." – Dr. Jason Hill [04:39]

5. Workflow & Technology: How They Interact

• The right workflow changes can make complex technologies work better—or even unnecessary.
• Real-world best practices are being shared, sometimes leading EHR vendors (like Epic) to adopt the innovations.

6. Root Cause Analysis and Model Iterations

• ML excels at finding associations, not necessarily causality. It narrows down problems for human experts to analyze.

  "ML can work really good at figuring out what's the root cause. ...It gives you...20 [possibilities] instead of an infinite canvas..." – Dr. Jason Hill [07:02]

• Ochsner’s Deterioration Index: Predicts which hospitalized patients may worsen soon. The team adjusted the prediction time window after finding earlier predictions didn’t prompt action.

  "If you actually tuned up your algorithm a little bit more and made it a little closer to the event, the patients would look sicker and people would actually do something about it." – Dr. Jason Hill [08:13]

• As predictive interventions improve patient outcomes, the model’s predictive power naturally declines, requiring retraining.

7. Value-Based Care and Data

• Value-based care changes both incentives and what data is valuable; to maximize positive patient outcomes, risk contracts need to match the patient base.

  "If you don't actually expand your risk contracts to include more of your patients, then you're taking value based actions without the value based reimbursement." – David Leingang [09:15]

8. The Power of Collaboration: The AHEAD Network

• Ochsner is launching a collaboration (AHEAD Network) with universities across the Gulf South to connect PhD/postdoc data scientists hungry for real-life healthcare problems with hospital and industry partners.

  "If we can bring the rural hospitals and bring those networks together, give them the problems, then all of a sudden it becomes a very virtuous relationship." – Dr. Jason Hill [11:49]

• Easier tools and AI are democratizing innovation—now, small teams (even at rural hospitals) can implement improvements rapidly.

  "You don’t need a giant team. You can have two or three AI engineers...moving things at an unbelievable clip..." – Dr. Jason Hill [11:21]

Notable Quotes & Memorable Moments

• On the Unintended Positive Impacts of Data:

  "Because of that data, we actually created a weight management program..." – Dr. Jason Hill [03:32]

• On AI as a Tool, Not a Solution:

  "People think of AI as this big, nebulous thing, but truly it's just a tool to solve problems." – Dr. Jason Hill [03:55]

• On the Best Message Metric:

  "The best message I receive is the message I don't receive." – Dr. Jason Hill [04:39]

• On Model Implementation:

  "We became a victim of our own success and our model became less predictive over time." – Dr. Jason Hill [08:43]

• On Working in Data Science in Healthcare:

  "It's sometimes hard to recognize that direct impact. But in healthcare it's there; it's extremely rewarding." – David Leingang [15:55]

Lightning Round & Closing (Personal Touches)

• Why Did You Get into Healthcare?

  • David Leingang: “Started in audit, stayed for the value—helping people.” [15:01]
  • Dr. Jason Hill: “Life event—car accident made me reconsider career; caring for people became a calling.” [16:25]

• Advice for Your 20-Year-Old Self:

  • “Take more math classes.” – David Leingang [17:36]
  • “You’re not going to believe it, but someday you’ll end up a doctor.” – Dr. Jason Hill [17:54]

• Desert Island Records:

  • David: Garth Brooks’ Seven, Eve 6 Inside Out, Mighty Ducks movie [19:20]
  • Jason: Red Hot Chili Peppers Blood Sugar Sex Magik, Pearl Jam Ten, The Matrix Trilogy [20:03]

• Riskiest Thing Ever Done:

  • David: Skydiving (twice solo!) [21:43]
  • Jason: Race-driving Ferraris and Lamborghinis in Vegas [22:38]

• Funnest Weapon Fired:

  • Jason: Japanese anti-aircraft rifle; “almost blew my shoulder off.” [24:02]

Timestamps of Key Segments

• Introduction & Background: [00:05]–[01:09]
• Ochsner’s Messaging Burden & Data Insights: [01:09]–[03:52]
• AI as a Tool vs. Education/Workflow: [03:55]–[04:39]
• Workflow Tweaks for In-basket Management: [05:11]–[06:33]
• On Metrics, Anxiety, & Best Practice: [05:11]–[06:39]
• Root-Cause Discovery via ML: [07:02]–[08:50]
• Adjusting the Deterioration Index: [08:03]–[08:50]
• Value-Based Care & Data Science: [09:15]–[09:53]
• AHEAD Network / Collaboration: [11:49]–[14:24]
• Personal & Lightning Rounds: [15:01]–[23:40]
• Closing Contact Info: [26:00]–[26:16]

Tone & Style

Conversational, humorous, pragmatic, and motivational. Both guests are candid and practical, focused on solving real-world clinician problems, and enthusiastic about the democratization and acceleration of healthcare innovation. The host’s tone is engaging, blending professional respect with friendliness.

Further Information

• To contact Dr. Jason Hill: jahill@ochsner.org
• To contact David Leingang: david.leingang@ochsner.org
• More on patient safety & risk: Censinet – Show Notes & Resources

Summary prepared for listeners seeking actionable insights on healthcare workflow innovation and the practical application of data science and AI.