Creator: Martin Allen
Still Accountable: AI’s a Big Help in Many, But Not All Ways

Martin Allen shares his experiences with AI and the “stunning” ways it helps in research of CMS rules, as an example. But AI can’t replace the brains we still call human, he opines.
By now, everyone who uses an internet search engine has experienced the benefits of artificial intelligence (AI). I was not an early adopter of AI but quickly changed my mind when I saw the iterative processes demonstrated at conferences and conventions. The hook was hearing that it would only get faster over time with machine learning. Seeing it in action with my colleagues, I downloaded a free AI app to my iPhone. Now – I’m a believer. Here’s how easy it can be.
CMS dropped a couple of proposed rules early this week, so I asked AI for a list of the annual payment rules for Medicare-certified providers. It gave me a list of 17 that it considered the “principal annual payment rules”.
- Acute Inpatient Hospital Prospective Payment System (IPPS),
including Long-Term Care Hospitals (LTCH)
- Hospital Outpatient Prospective Payment System (OPPS)
and Ambulatory Surgery Centers (ASC)
- Skilled Nursing Facility Prospective Payment System (SNF PPS)
- Inpatient Rehabilitation Facility Prospective Payment System (IRF PPS)
- Inpatient Psychiatric Facility Prospective Payment System (IPF PPS)
- Home Health Prospective Payment System (HH PPS/PDGM)
- Hospice Wage Index and Payment Update
- Medicare Physician Fee Schedule (PFS)
- End-Stage Renal Disease Prospective Payment System (ESRD PPS)
- Additional annual payment updates for
Clinical Laboratories,
Durable Medical Equipment (DMEPOS),
Federally Qualified Health Centers (FQHCs),
Rural Health Clinics (RHCs),
Ambulance Services.
I then asked how many pages were published in the Federal Register in the latest rule-making cycle. I received a ranked table from longest (Physician Fee Schedule – 320 pages) to shortest (Hospice – 28 pages). Please note – these are the Federal Register documents with three-columns - not the unpublished public-inspection versions. The unpublished version of the Physician Fee Schedule is over 1,500 pages long. Yikes!
How has the length of the rule changed over time? AI sensed where I was going with this and gave me a comparison over 20 years. From that I put together the following table:
Rule & Page Count | Around 2005 | Around 2015 | 2025 –2026 |
Physician Fee Schedule | 130 | 220 | 320+ |
Hospital Outpatient (OPPS/ASC) | 120 | 180 | 260+ |
Hospital Inpatient (IPPS/LTCH) | 180 | 230 | 280+ |
Home Health | 45 | 70 | 130+ |
Skilled Nursing Facility | 25 | 40 | 58 |
Inpatient Rehabilitation Facility | 25 | 35 | 55 |
Inpatient Psychiatric Facility | 20 | 30 | 45 |
Hospice | 15 | 20 | 28 |
ESRD | 35 | 55 | 90 |
Estimated Total Pages | 600 | 900 | 1250-1350 |
What caused the huge increase in the length of these rules in 20 years? AI wrote the following summary:
“Several factors have contributed to this expansion, including the addition of quality reporting programs, value-based purchasing initiatives, alternative payment models, more sophisticated case-mix methodologies (such as PDPM and PDGM), increasingly detailed coding and documentation requirements, and new statutory mandates enacted by Congress.
While many of these changes were intended to improve accountability and payment accuracy, they have also substantially increased the regulatory and administrative burden on providers, consultants, and CMS itself. The growth in the size of these annual rules provides a useful quantitative measure of the increasing complexity of Medicare payment policy.”
AI and the Human Touch
This is a stunning bit of machine writing. Not only does it detail the complexity of provider payment changes over time, but it also draws a conclusion. Regulations are an administrative burden on providers, those who support them, and on CMS themselves.
Putting on my former association hat, AI tools used for research can be a great asset and time saver. Similarly, using it for a deep dive into a specific topic or asking about something you don’t know about can be useful. In the end though, I was and am a policy geek. I want to delve into the details of a rule, discuss it with my colleagues and members, and draw my own conclusions. AI may have access to all the information in the world, but it can’t outshine our industry experts and all their experience.
I do see the risks (I still have fears of Skynet as depicted in the Terminator movies), but world, federal, and state governments will learn how to manage AI, right? Or will AI manage them?
About the Author: Martin Allen is the former senior vice president of reimbursement policy for the American Health Care Association/National Center for Assisted Living. He also served as the vice president of revenue cycle and reimbursement services for ProMedica Senior Care (formerly HCR ManorCare). A certified public accountant with a master’s in business administration, Allen has more than 35 years of extensive work in accounting, Medicare and Medicaid reimbursement, healthcare compliance, risk management, and revenue cycle processes.
Any thoughts on this article? Please contact Patrick Connole at pconnole@parkplacelive.com.

z-INTEL Digest #1: 6.20.22
