Thursday, February 12, 2026

CMS Gapfill: Spring 2026 Update

Each fall, CMS determines which of 100-odd new lab codes are crosswalked and which are "gapfilled," meaning priced by MACs - for us, in 1Q2026.  

See Change Request CR14312, T13514, December 5, 2025.  By my count, in that public document, 40 codes are for contractor gapfill right now.


How They Do It - Novitas/FCSO MAC and NGS MAC 

NGS MAC and Novitas/FCSO MACs generally post a webpage notifying providers of the numbers of the new gapfill codes, and requesting information.  Novitas in recent years has used a "SurveyMonkey" (or similar) online web form for data entry.   MACs ask for CMS-specified gapfill information, such as test purpose, charges, prices of other payors, resources required, etc.  

I believe NGS MAC post the list of PLA codes and provides an email for receipt of your materials.

How They Do It - MolDx MACs (Palmetto, Noridian, CSG, WPS)

At least past years, nothing is posted publicly at MolDx  One of the medical directors there has told me that these gapfills could be finished with their internal expertise as MolDx lab experts, and they'd reach out to a lab about its code just if and when MolDx found it necessary.   

In the past several years, all the MolDx MACs offer the same price for every code being priced (molecular or not), and, the NGS MAC has matched the MolDx price for every code being priced.  In recent years the Novitas-FCSO gapfill prices have tracked close to the MolDx ones, sometimes identical, sometimes 20% higher or lower.

How to Gapfill (Bruce's View)

At a recent meeting, I heard a CMS medical diirector comment that gapfill pricing was very difficult, and it was hard to get the necessary information.  I can see that.   For what it's worth, leveraging my decades of experience, here are some ideas.

1.  Low hanging fruit Some of the gapfill codes will be very similar to an existing priced code and a price can be matched.   (Gapfill rules do allow use of similar-test prices in this way at the gapfill stage, even if they were not crosswalkd the prior autumn.)

2.  Take what comes.  Some companies know what they're doing.  At least some companies will know the MACs are gapfilling and will submit materials.  Use that.

_______

3.  Check if that lab presented in June public comment meeting.  (Agenda online at CMS CLFS page).

4.  July Advisors.  Check the CMS website for the July advisors meeting of the prior summer.  See what crosswalks they recommended.  (If they recommended gapfill, it doesn't help you, but at least you checked).

_______

5.  Reach out to the lab.   I see two angles:
(A) Email an information box or contact box.   (B) Call and ask for the CEO.   (This assumes the missing code is probably from a small company, not Roche or Quest).   

Log the above by date.  If you haven't heard in 7 days, go to #5.

6.  Look at the website and see if they list publications, if so, pull down one or two and use the methods section.  Work from that.  

7.  Similar to #5 but layer on AI.  Use Chat GPT and ask if to use the internet to find the methods used for [test name] [company name].  However, you're responsible for checking what comes back against original source and make no reliance on the AI.   It's just pointing you a public sounce you didnt' come across on Google and it's up to you from there. 

_____

Obviously you can put 6&7 before 5 if you prefer. 

_____

8.
If NONE of the above work,
then you are stuck with the test name and code text.   

You have to pick a reasonable price from that.   Remember you may hear more in the summer (2026) public comment period after you release your draft price.  My advice, your result won't be any more accurate at this stage if you work for 10 minues or 60 or 120.  So give it 10.

___

For a blog this week, I needed a link to where CMS defines terms like "effective date" and "implementation date."   I had Chat GPT look for it rather than me. It did a perfect job of finding something ultra obscure in a few seconds.  I was basically using it as a nuclear powered Google.  Of course, I had to verify the obscure document it found was right.