Friday, June 23, 2023

CMS Outlines Alzheimer Drug Registries; Blog #2: What's in the Registry?

Blog #2 - What's In the Registry?

CMS famously is requiring some level of registries - maybe Registry Lite - for all patients who imminently will start getting the new FDA-approved Alzheimer beta amyloid drugs.

CMS has released template information for the registries.  Specific registries will have their own management and focus and will be set up in the near future.

Information here and here.  

See Blog #1 about the amyloid tests in it - here.

See a rapid report on the clinical aspects of the registry, Tunis & Neumann, Health Affairs here.  Endpoints here.  Genomeweb here.


Blog #2 - The Registry is a Mere Shell?

CMS has made big political waves by insisted that the amyloid drugs are NOT covered under the "accelerated approval" state, and even when fully approved, will NOT be covered outside of some kind of clinical study aka registry.   See American Academic of Neurology here.

CMS has now issued an overview of Registry Lite, see links above.

Does the registry really make sense?  In its effort to keep the registry "lite," CMS is mostly asking for information it already has.   It asks for the physician's name, NPI, address, the patient's identifiers, which drug the patient is on.  Well, those are fields in the claim form the doctor submits anyway.  

Does the doctor's office submit those fields at 11:00 am via the CMS billing process, and then log on to a separate website at 11:15, and type the same information manually into a registry portal?  Is this the monumental advance?   

  • It reminds me of nothing so much as being in a 2023 doctor's office and asked to fill out three paper forms by hand, AND each one requiring handwriting out my name, birthday, address and zip code anew.

The minimalistic registry has a few other fields, like uploading information about cognitive tests, but the NCD doesn't specifically require any so the result of that endeavor may often be minimal.

Emperor and No Clothes?

If you do two things - if you carefully READ the "minimal registry" announcement, and if you also THINK about it - it doesn't seem like much at all is happening.

This is before even considering that the clinical registry has no control group, and the drug effects are subtle and require elaborate tests to verify, acting against any value at all in the area of clinical effectiveness, the area CMS was by far the most interested in.   So they had a goal - more knowledge about clinical effectiveness - and they can up with a method - this Registry Lite - that takes them nowhere towards the goal.

To the extent they wanted as a second goal to track side effects - like hospitalization for stroke - they could do that themselves based on the patient's AMY drug claims and the patient's hospital stroke admissions - without any value in retyping the CMS drug claims and retyping the CMS hospital admissions claims into a CMS portal.

CMS's Own Words

In its new guidance document for CED (in the TCET context) CMS refers to this description of CED studies:

Postmarket study proposals, particularly those that rely on real world data, have the  potential to generate evidence that complements tightly controlled premarket traditional clinical trials by demonstrating external validity.  Nonetheless, manufacturers should be aware that these studies require considerable planning in data validation, linkage, and transformation; specification of the study protocol; data analysis; and reporting.  The study design, patient inclusion criteria, primary and secondary endpoints, treatment setting, analytic approaches, timing of outcome assessment, and data sources should be fully pre-specified in the submitted protocol.  When writing [these CED protocols], manufacturers should propose clinically meaningful benchmarks for each study outcome and provide supporting evidence. 

In contrast the map they've just published defining their NCD Alzheimer real world CED, meet none of these requirements !!!

Similarly, the Alzheimer CED rules seem to bear no relationship to nearly simultaneous CED guidance just published by CMS in tandem with the TCET release - here.

In some ways the requirements mimic a vintage document that was never used, called "coverage with appropriateness determation" which was "a type of CED" in 2006.  I.e., the patient is [x] amyloid positive and [c] not on anticoagulants.