Dear Doctor: the OIG Thinks We Owe Them $12.2 B

The Department of Health and Human Services has done some numbers on payment accuracy for federal programs.  Well, "inaccuracy" is actually what they were talking about. Among a short list of "high-error" programs, Medicare Advantage Part C gets tagged with an 8.5% improper payment rate, which translates into $12.2 B estimated for payment year 2014 (down from their estimated 15.4% error rate in 2009).  It's a big enough number to get the attention of Senators Grassley and McCaskill, as well. But where did they come up with these numbers and why should physicians be concerned?

These numbers arise from an auditing program known in the industry as "RADV", risk adjustment data validation. It is a temptation for clinicians to immediately tune out right now as the acronym floats out in the atmosphere and their eyes glaze over. As one clinician recently told me, "It's just so much alphabet soup. I'm  busy trying to take better care of my patients."  Well, you may be surprised to learn that when you write down whatever you do in the chart, you may be contributing to the perception that we owe the feds a big chunk of change back. 
Here is a way to think about this. The audit is sort of a test of your charts, checking them for the degree to which they agree with a set of coding and documentation standards.  Fail and your healthplan partner gets in deep trouble and (you may not realize it) you may have to take out your own checkbook to repay capitation take-backs. 
Your charts were always your own way of keeping tabs on your patients, right? Reminding yourself what happened previously and what to do the next time the patient shows up in your office.  Well, guess what?  Those charts are the sole source of truth according to the government. And how they are documented, how well they support the diagnostic codes from your claims being submitted to the government will determine what happens next in a $12.2 B legal wrangle. 
The feds have a method of punishment at their disposal called extrapolation:  generalizing the error rate for specific diagnoses in the RADV audit sample to a whole population of Medicare Advantage members enrolled with that affiliated healthplan.  Based on the biopsy of your charts, the feds inflate the scale of their findings and run that against the entire population, multiplied by the whole year's payment to the healthplan. Think of it as an exam on a small sample of charts projected onto an IMAX theater screen, exaggerating everything it views.  
"But wait a minute!", you may object,  "I do not document any differently for patients that are on Medicare plans:  I keep my charts the same way for all my patients. It has never been a problem before. We are not conducting any kind of fraud and all my claims get paid ultimately." 
This is all because the rules for Medicare Advantage are different than fee for service claims billing. There you only have to have an ICD code that loosely agrees with your E&M codes on the claim.  The service codes are what trigger payment under fee for service. 
Now for risk adjustment under Medicare Advantage programs: it's all about the specificity of the ICD code.  The government is not going to complain if your offices under-codes the diagnosis:  going with a simple 250.00 for diabetes mellitus without complications and when it is really 250.40 for diabetes with renal manifestations. This is true, although the CMS payment to the Medicare Advantage plan (and to you, if you are capitated for your services) will be dramatically lower than if you documented the 250.40. That's the way it works: under-code and get under-paid. Of course, you are not coding:  your billing office codes.
However, if your office puts the 250.40 diagnosis on the claim, your chart documentation had better support that diagnosis. Otherwise, if you chart "DM", in the mind of the government, you are "up-coding" with the 250.40 diagnosis on the claim. Once the diagnosis gets uploaded and submitted to the government to adjust the premium payments, that is where the liability starts getting real. Open up the chart and see if it has all the right ingredients that the auditor requires to support the diagnosis.  If not, that diagnoses gets rejected. If you extrapolate those errors, it rolls up to the tune of $12.2B of improper payments that HHS wants to recover. If your practice was taking capitation payments from the healthplan, the recovery process probably has your practice's name on it. 
So think about that for a moment. Let's say your patient really has diabetes with renal manifestations.  If that patient came to you under Original Medicare, your services would be paid for at the taxpayers' expense through the Medicare trust fund, whether you coded the claim as a 250.0 or a 250.40.  But if the patient came in under a Medicare Advantage plan and your billing office under-codes it, you will have to take care of that patient and treat him / her anyway for that condition, but you will not receive the funding appropriate to pay for all the services someone with diabetes with renal manifestations is likely to need.  The taxpayer just saved a lot of money because the risk adjusted premium came in artificially low. On the other hand, if you coded the patient's condition as a 240.40 but your chart documentation does not pass the documentation test, you are now part of the $12.2 B payment "problem".  It's not all your fault, of course. There are other "flies in the ointment" that contaminate the quality of the diagnostic data reporting. But, yes, the OIG thinks we owe the taxpayers $12.2 B back. 
The point of this rant is not to suggest that you become a coder. It is to point out that you are not so removed from the "alphabet soup" of things as you chart away in the exam room. The downstream impact ripples far and wide.  And sometimes it comes home to roost. 
So what should you do about all of this? First, get your billing staff squared away so they understand all of the coding implications. Second, make sure you chart with an intention of being understood by all the involved parties, not just yourself:  use standard terminology and nomenclature. Here is what my good friend and physician executive, Dr. Mark Stern, says: 
"These are documents that  should be available to other providers so they know exactly what is going on with that patient. The whole concept of electronic records is to share information. If a provider writes in his/her own short hand or illegible scribble that delays or result in redundant care, can result in incorrect interpretation, add cost and result in asking the patients and family members the same questions  dozens of time in a single encounter. 
I personally believe speaking with one language is a good thing.  If you have  ever had to  review a stack of charts for a specific issue, speaking one language would reduce misinterpretation. 
I don’t agree with all of CMS documentation guidelines but for the most part they make sense as much as I hate to admit it. Bottom line: training providers to document accurately  using standard language and basic requirements, e.g. medical plan, is a good thing. 
My belief is physicians should not be in the coding business.  HCC’s [hierarchical condition codes] can be something that providers conceptually understand but should not be the thing that drives their documentation or coding. "

This sounds like good advice and I hope you agree.  

Tags: RADV audits, extrapolation

Log on to Your Rise Account

Forgot your password?
Create an Account


Latest Posts

CMS Gives EDPS Transition Some Breathing Room

CMS published the final call letter for 2018 yesterday, April 3, which included a welcome accouncement regarding the transition from RAPS to EDPS-based RAF scores. Citing numerous public comments on the subject, CMS throttled back the speed with which they plan to switch over to an encounter-based methodology. Instead of the blended rates originally contemplated, they announced that the more modest blend of 85% RAPS to 15% EDPS would be used in 2018, allowing more time to improve the reliability of the encounter data methods. While the RISE data collaboration study was not cited, we believe that the educational value of our study, along with our communication and advocacy of a more moderate approach by CMS, contributed to the confidence with which plans and other interested parties spoke up during the open comment period. Once again, we owe thanks to the folks at Avalere and Inovalon, as well as at AHIP, for the collegial and professional collaboration. Also, we want to thank the health plans that actively participated in our study for making this work possible. ...
Read More

Take Aways from RISE Nashville Summit

The 11th Annual RISE Nashville Summit continued the event’s tradition of yearly growth. The return to downtown Nashville was widely applauded by attendees, who were glad to be back near Broadway’s nighttime funk and fun. While festive, this year's event occurred in the wake of the new administration in Washington, D.C., and the healthcare themes surrounding the "repeal and replace" of the Affordable Care Act (ACA) cast a long shadow. In contrast to the upbeat, confident notes struck by last year’s keynote speaker Senator Tom Daschle, this year’s sobering keynote address by Howard Fineman, NBC/MSNBC political analyst, The Huffington Post Media Group global editorial director, and bestselling author, was an assessment of the pluses and minuses of our new president. Mr. Fineman's remarks indicated that the political alliances in power will seek to undo what Senator Daschle viewed as "irreversible” a year ago....
Read More

Upcoming Conference


Qualipalooza: The 2nd Annual RISE Quality Leadership Summit 

This unique event incorporates three conferences presented side-by-side: the Star Ratings Strategic Planning Forum, the HEDIS Forum, and the CAHPS, HOS & Member Survey Forum. Register for one conference for an in-depth examination of a single area, or design your own event by opting for the all-access pass and choosing the sessions from each conference which correspond exactly to your interests.


Upcoming Webinar

The Impact of Quality Incentive Models in Medicaid Markets


Thirty-one of our fifty states now have Medicaid managed care, and several markets are expected to implement managed care in the next few years. More than $160B in Medicaid spending occurs through the Managed Care Organizations. As more and more states seek to do more with less, increasing accountability for health quality outcomes is placed on health plans. Join this webinar to learn the typical quality payment approaches states use, issues often faced by health plans under each model and what states are expected to do with payment models tied to quality performance in light of near term Medicaid reform efforts.


Connect With Us

Copyright © 2014 Resource Initiative & Society for Education. All rights reserved.