Tuesday, December 21, 2021

OIG Issues Report on Massive Medicare Genetics Spending Growth, Highlights Fraud

The Office of the Inspector General has issued an overdue report on the massive growth in genetic spending & alleged fraud since 2018.   Find the 21 page report here or here.  I say the report is overdue because it's based on data ending in December 2019, and it's nearly 2022.

The report officially highlights what I vigorously pointed out in Fall 2020 based on  CY2019 data - that anomalous spending in genetic testing had skyrocketed.  The OIG report also builds on a July 2020 white paper (24pp) by the Healthcare Fraud Prevention Partnership - here.

The December 16 HHS OIG report documents (page 19) that the highest-spending genetic code in 2019 (and nearly the highest in 2020) was 81408, a code for a grabbag of very rare genes with full sequence reported.  
  • The usage of this code was essentially zero in 2016 and 2017, and popped to $121M and $284M in 2018 and 2019, respectively.  
  •  The genes inside 81408 are so rare, AMA expects they will have very little usage (less than 1000 per year nationally), or else they would be elevated to regular genetic Tier 1 codes.  

81408 Massively Incompetently Handled by CMS

See my five-blog series in December 2021 - here.

For a shorter five-part trail of horrors, here:

#1 Which labs?  Billing for 81408 for Medicare patients by familiar labs like Labcorp, Quest, Invitae, Bioreference, Ambry, and others?   Zero.   Look for billing by pop-up places with names like "ABC Lab" and no website.

#2 Which MACs?  As I've documented in the past, billing for 81408 was essentially zero in the MolDx MACs and the NGS MACs, and was concentrated only in the Novitas and FCSO MACs, and briefly in the now-defunct Cahaba MAC.    

#3 Complete Incompetence at CMS Office of Program Integrity.  In an ongoing scandal that it would be charitable to call "complete incompetence," the Medical Unlikely Edit, or payment cap, for 81408 is 2, allowing fraudulent labs to bill every case in a multiple of 2 ($4000!).   It would have taken only seconds in 2018 or 2019 or 2020 or 2021 to change this edit to 1, saving Medicare about $150M a year, pretty good return on 30 seconds effort by an administrator.  It's even more bizarre because nearly all genetic codes, including Tier 2 codes, have an MUE of 1.   

#4 Problem Understood in Mid 2019; Few Actions by End-2020.   The problems were fully understand by mid 2019, as evidenced in public press releases about Operation Double Helix (here).   Yet spending only dipped slightly for 81408 from CY2019 to the end of CY2020, which is crazy.   Spending should have gone back to near zero, by putting place controls in January 2020 based on findings in 1H2019.   Should have bone back to near zero, as it was in 2016, 2017, 2018.   

Hundreds of millions of dollars gushed out long after the torrent of cash had been fully identified.  

#5 PAMA Paid a Role.   PAMA played a role when PAMA prices were introduced in CY2018.   In 2016-2017, the higher Tier 2 codes including 81408 had minimal utilization, and had no prices set (because they had no Medicare usage!!!!) and had to be manually reviewed and individually paid.  On January 1, 2018, 81408 was assigned a price of $2000 per test and began to autopay, slowly at first, and then madly.   And often in pairs, following the MUE limit of 2.

OIG report (page 14) highlights several categories of fraudulent billing.  (1) Billing for services not performed at all.   (2)  Unbundling of claims or repeat billing.  (3) Blanket ordering of costly panels of rare genes on a series of patients by unscrupulous providers which are then autopaid by the insurer.


See comprehensive year-by-year spending tables pn pages 19-20.  

Note that OIG treats all spending the same, so that $300M spent on Cologuard (under FDA approval and an NCD) is treated the same as $300M spent on 81408 (to pop up labs with no website cited as allegedly fraudulent by DOJ in press releases).   It's all the same in this report. As Becky Winslow remarked on Linked In, "Labs who don't even have the technical ability to perform 81408 have billed and been paid for 81408."

See concise coverage of the OIG report at Genomeweb here.


In Point 2, I noted that the alleged fraudulent claims never occurred MolDx MACs, which devote great resources to genomic payment controls.  However, no payments occurred in the NGS MACs either, which devote de minimis controls (but firm ones) on genetics.