Notes: Why FDA and Payors Have Different Gaps, Regarding Drugs, or Devices, or Diagnostics.

 

There are significant differences in the FDA approvals of drugs, devices, and diagnostics.  Payers often sense this, but it's not always well articulated.  For example, new FDA drugs are almost always covered, but diagnostics rarely so.   Can we outline some basic reasons why this occurs?

DRUGS

Drugs are traditionally approved with clear positive end points that must be met with acceptably large improvements   For example, a drug trials shows that: (a) cancer patients live 4 months longer, or (b) patients have 30% less strokes.   

1.  This frames the benefits claims (as well as FDA advertising regulations).  If the patients live 4 months longer, that brackets any marketing claims and is the core claim.   You can't magically claim they live 2 years longer. 

2.  There can still be debates about risk/benefit, especially when benefits are small.   Remember the FDA will approve if benefit > risk, so both may be small in some drugs.

3.  One twist with drugs, which gets a lot of press now.  Accelerated approval is based on surrogates, sometimes biomarkers (tumor DNA disappears or HIV RNA disappears) and sometimes on clinical surrogates (PFS is achieved).  Pushing boundaries in accelerated approval benchmarks, or size of effect, is problematic.  Much debate on this, and legislation proposed.   Small surrogate biomarker improvements in short drug trials may stretch the faith of payors in the drug process.

DEVICES

Devices are either approved with benefit > risk in trials, or, based on predicates. Potential pitfalls include: 

1.  Payers may expect substantially longer outcomes that FDA requires.   A device I worked on novel surgery modality for prostate enlargement, was FDA cleared on 6 months strongly favorable clinical trials, but Medicare MACs wanted 3 years outcomes.   This meant a shocking gap between FDA and Payors, regarding the company's launch plans and budget.

2.  Sometimes in the 510(k) path, a device may be approved based on an odd partial combination of other device approvals.  The company's regulatory lawyers may argue this configuration past the FDA, but it flops with payors.   

3.  There is a famous (to me) case in the Biodesign textbook of a company with a nasal surgery device and 2 coding consultants told them it could use code X paying $X dollars.  But after launch, one after another payors said, "No, you can't use Code X" which startled the company and badly upset its plans. 

DIAGNOSTICS

Very, very often, diagnostics are based on accuracy trials not "in use with doctors" trials.   These diagnostic accuracy may be archival trials (large biospecimen banks) or prospective trials, but prospective trials where the test is not actually used by anyone, only data accumulated.   For example, the Roche HPV test for cervical screening, the Cologuard test for colorectal cancer CRC, and the current Guardant blood test for CRC, all were studied in enormous clinical trials, but only for data gathering purposes.

The impact of the diagnostic depends on a downstream chain of events, decisions, and therapies. Payors may be skeptical the FDA-approved test accuracy will ACTUALLY, in the future, lead to a helpful chain of events in the real world.

   Diagnostics may, also, avoid FDA approval completely (LDT).  This freedom from FDA review of claims and statements, can be unfortunate and lead to hype.  There will be widely disparate claims, or hype, about what the test would do in actual use, since it is conjectural.  It also means that cost/qaly projections are more conjectural and uncertain, even more so than other medical interventions.   As a result, payors are highly dubious with most new diagnostics.  Guideline committees may be slow or contentious.   

Regarding the chain of events required for predicted patient benefit.  In real-life trials, a biomarker shown to be accurate in the lab, may not be used accordingly by physicians (See the large RCT "PROACT" trial of an infection biomarker.  The test was accurate but lots of physicians in the real world of practice ignored it.)