Blog
Short articles by Tyler Kent on healthcare AI, data and operations. Every figure is sourced. Follow by RSS.
- AI adoption
What a medical practice can hand to AI without patient data
Four of the seven practice problems on my map can go to AI tools that never read a patient record. Here is which four, and how to handle the other three. - AI adoption
What happens to your data when an AI vendor shuts down
What HIPAA requires of a vendor holding patient data when its contract ends, what it leaves to your contract, and the terms to check before a vendor closes. - AI adoption
In my map of 44 healthcare problems, 30 need no patient data
I mapped the 44 problems that cost six kinds of healthcare organizations the most money or carry the most risk. The biggest group is about money and payers, and the next is about rules, surveys and audits. Healthcare uses AI at about the national rate, and for 30 of the 44 problems, every tool I'd advise works without reading a patient record. - Payment and policy
TEAM started the 30-day clock for about 700 hospitals. Here is the baseline a hospital needs for each TEAM surgery.
What TEAM measures, why the CMS data feed is not a baseline, and the five numbers a hospital needs on every episode. - Data and analytics
Self-service analytics for sales and operations teams
How sales and operations teams can query the data warehouse without filing a ticket, and what the build costs. - AI adoption
Building AI so each project makes the next one cheaper
Four practices that make each new AI build cheaper than the last. - Health centers
Running UDS reporting as a year-round data pipeline
HRSA's estimate of the staff time a UDS report takes, and how it changes for 2026. If a data pipeline builds the report all year, staff only have to check its output at the February deadline. - AI adoption
Choosing between a fractional AI lead and a first AI hire
The three options a healthcare organization of 100 to 1,000 people has, what each costs over a year, and the two cases where a full-time hire is right. - AI adoption
What changes when a hospital team learns to use AI
What leaders, power users, general staff and skeptics each do differently after training that works. - AI adoption
How to score an AI vendor before the demo
Write requirements before the demo, get written answers to five data questions, and run the build-versus-buy math in dollars. - Hospital operations
ED boarding starts with inpatient capacity
Boarding is counted in the ED, but its causes lie outside it, in inpatient capacity and discharge timing. - Hospital operations
Discharge delays cost 52 hospitals an estimated $169 million in one quarter
What a delayed discharge costs a hospital, and why the fix starts with treating discharge as a queue: count the patients waiting to leave and put a price on each day they wait. - Hospital operations
Surgical schedule smoothing works, and few hospitals use it
Smoothing the elective surgery schedule raised one children's hospital's occupancy from 76% to 91%, worth about $137 million a year by one account. - AI adoption
Where healthcare AI pays off in 2026 budgets
Healthcare organizations reported $3.20 back per dollar in a Microsoft-sponsored survey. Which back-office workflows make a return easiest to measure, and three questions to ask of every AI line in a 2026 budget.
Sources for the figures above: Transforming Episode Accountability Model (TEAM) model page (opens in a new tab); HANYS, The scope of complex case discharge delays in New York state (2023) (opens in a new tab); Eugene Litvak, Bridging the Gap: Operations Management Science, Patient Safety and Health Care Cost (presentation) (opens in a new tab); Institute for Healthcare Improvement, Achieving Hospital-wide Patient Flow, 2nd edition (2020) (opens in a new tab); Microsoft, reporting the IDC Business Value of AI Survey (IDC #US51331223, November 2023) (opens in a new tab).
Book a 20‑minute call
If one of these problems looks like yours, talk it through with Tyler.
Prefer to skip the call? Get a written proposal
Or email tyler@kentapplied.com.