Although it reviews trials, since then, CMS has been able to cover both device cost and other clinical costs for Category B devices. (Only general costs, not the device cost, are covered for Category A "experimental" devices.)
The theory is that the Category B device is providing medical care; the patient needs a pacemaker or an artificial hip, and this device is the one device they are getting for treatment of their condition.
The CMS home page for the policy is here. CMS also has a webpage for all approved studies - 239 as of June 2018, here. CMS approves proposed trials for payment centrally and provides instructions how to apply (you send them the clinical study protocol, the IRB approval, etc.)
The FDA's most recent guidance for how it establishes its Cat A/B decisions is from December 2017, here.
- I copied the CMS web-based approval table of 239 rows into an Excel spreadsheet and put it in the cloud for you, here. See the "Down Arrow" at upper right to download.
Percent Category A
36 of the 239 trials as "Category A", or about 15%. We don't know the success or failure rate, so we can't tell if it's different for A or B trials. But we do know, that about 1 in 8 approved trials is Category A.
I also sorted by sponsor. Many sponsors appear only once or twice across the couple hundred approvals. Abbott appears 6 times, Boston Scientific 6 times, Edwards 9 times, Medtronic 15 times. It looks like about 25 trials are sponsored by universities. Among companies whose names might be less familiar, Novocure has 3 trials and Vivek Reddy 4.
Approvals By Year
Numbers may not add perfectly due to variations in my Excel cut/paste, but I tally about 43 approvals so far in 2018, versus 92 for 2017, 68 for 2016, and 25 for 2015.