While supplies last! Free 2022 Coding Essentials for Infusion & Injection Therapy Services book with every RACmonitor webcast order. No code required. Order now >


There are at least two specific programs through which the Centers for Medicare & Medicaid Services (CMS) can use statistical sampling and extrapolation to identify improper overpayments:


  • May 1999: CMS awarded 12 Indefinite Delivery – Indefinite Quantity (IDIQ) contracts, creating the Program Safeguard Contractors (PSCs); and
  • October 2008: under the new national RAC program initiative, CMS awarded contracts to four Recovery Audit Contractors (RACs).


PSCs have been around for years now, and are known to use statistical sampling and extrapolation to identify overpayments to providers. Some observers claim that PSCs actually “create” errors this way, rather than identify them; however, they are quite successful at overpayment recoupment.


Now come the RACs, and although none of the RACs in the demonstration project chose to do so, these contractors also are permitted to use exactly the same extrapolation techniques. What does this potentially mean to providers? Bigger recoupment numbers and dramatic negative findings that will impact your bottom line.


Why Add When You Can Multiply?


Statistical sampling and extrapolation of sample results is a mathematical approach that allows an auditor to yield maximum results from a minimum of records review. In short, little work, huge payoff: who doesn’t like that? Plus, it’s really a simple process. Here’s a short version:


1. They take a random sample of your Medicare filed claims;
2. They review those claims for errors;
3. They calculate what you were overpaid, subtracting any underpayments;
4. They divide that amount by the number of claims in error;
5. They multiply that number by the number of claims you filed and
6. They tell you how much you owe.


Look at the following example for real bottom line possibility:


  • You have a small practice and filed 3,600 Medicare claims in a year, totaling $250,000 in reimbursements.
  • A CMS auditor uses a random sample of 100 of your claims (which is easy to do since they have all of your data on file).
  • The result: 40 errors found, producing a net loss of $3,000. (Overpayments minus underpayments – so if the overpayments total is larger, you have a net loss.) Not too bad, you’re thinking – we only lost $3,000. But hold on…
  • If the auditor applies extrapolation of the errors found, that $3,000 is just the beginning.
  • The net loss of $3,000 is divided by 100 claims, to produce an average loss (or what is called the “point estimate”) of $30.00. Now they get to multiply that $30 by all 3600 claims filed.
  • The final number now is not $3,000, but $108,000!


New Dog, Old Trick


This practice is not new, as mentioned above. It’s a proven method of auditing – proven in the sense that it produces nice numbers for the auditor, not the auditee. A certain college statistics professor once was fond of saying he could mathematically prove that a “random sample could produce more accurate results than a complete count of any sample universe.” Statistics can be very powerful.


CMS and the U.S. Congress believe in statistics and statisticians. In fact, government auditors are required to use highly qualified statisticians to produce their samples and do the math. To use extrapolation methodologies, they must hire a statistician for the project, or at least to produce a report that verifies methods used by the contractor.

Multiplication Is Painless


The process is simple, and even CMS only needs a few sentences to define it, as it did in the Medicare Program Integrity Manual, Chapter 3, Paragraph


“Statistical sampling is used to calculate and project (i.e., extrapolate) the amount of overpayment(s) made on claims. The Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA), mandates that before using extrapolation to determine overpayment amounts to be recovered by recoupment, offset or otherwise, there must be a determination of sustained or high level of payment error, or documentation that educational intervention has failed to correct the payment error. By law, the determination that a sustained or high level of payment error exists is not subject to administrative or judicial review”.


Two things you should note in that simple paragraph:




First, that there needs to be a “determination” of payment error, and documentation showing that education regarding said error has been fruitless. Actually, it says “or,” not “and,” but there was an update to the manual (Transmittal 114, dated June 10, 2005, if you care to look) that dedicated several pages to changing “or” to “and” for the supposed purpose of limiting the use of extrapolation.
“Not subject to review”


Second, the last sentence, especially the phrase, “…not subject to…review.” The authors of the document are trying to make it bulletproof. Actually, despite a 1990 challenge in the U.S. Court of Appeals [see the case of Chaves County Home Health Services, Inc. et al. v. Sullivan, U.S. District of Columbia, U.S. District Court for the District of Columbia, (Feb. 12, 1990) Civil No. 86-2691.], the determination is in fact beyond review.


The Court ruled that HHS/CMS had discretion under the Medicare Act to use any reasonable means to recover overpayments, and given that a case-by-case review of each individual claim was not feasible, the use of the statistical sampling method was deemed reasonable.


On Dec. 8, 2003, as part of the Medicare Prescription Drug, Improvement and Modernization Act of 2003, Public Law 108-273, (Title 42 U.S. Code Section 1395ddd), Section 935 of the act limits the use of extrapolation, clearly prohibiting the practice unless a CMS contractor can document one of the two requirements noted above: a high level of payment error or a record of education that failed to produce fewer errors.

What’s the Goal?


While the errors actually found in sampling of claims should (and will) be recouped, the goal of sampling, at least at first pass, should be to use sample findings as an impetus to initiate education at the provider level to prevent future errant billing. We all can hope that future lobbying efforts by providers and provider organizations may have some effect and remind lawmakers what the overall goal should be: to fix the healthcare system without tearing it down.

First Steps for a Provider


The only thing you can count on to defend against RAC audits, with or without extrapolation, is knowledge about the process and preparation. Know how to document and support “medical necessity” the way CMS defines it. Keep your staff consistently educated and alert about coding issues. Count on things to change – rules, codes, guidelines, especially the Medicare Program Integrity Manual. There is wisdom in preparing for the worst.


What can you do today to prepare for the worst? I’ll be talking more about that in future articles. For now, here are four simple steps you can take today.


1. Learn how to appeal a RAC denial. You can try to figure it out yourself, or we have affordable education available at RAC University.


2. Identify legal counsel for yourself that knows something about RACs or at least about healthcare reimbursement – perhaps you can find one through RACmonitor.com!


3. Calculate the cost for you to appeal. Do this NOW so you’re prepared. Once you get a demand letter, the clock is ticking and the meter is running. When calculating, don’t forget the costs of pulling, copying and sending records to CMS. The AMA has estimated that the cost for a hospital to appeal a single claim is about $2,000; for a physician, that number should be far smaller. But don’t assume – do the calculations yourself. Then you can make reasonable and intelligent appeal decisions.


4. Keep reading RACmonitor.com – we’ll keep writing and do our best to keep you informed.


In closing, we would like to make an observation regarding the concept and use of extrapolation in making future RAC determinations. In a recent conference call with CMS, when asked about extrapolation and whether it would be used by RACs, they quoted verbatim the paragraph you read above, direct from the CMS manual.


Basically, they were saying “Yes.”


The potential is real. Will RACs use extrapolation in the future? What do you think?



About the Author

Patricia Dear has more than 30 years of experience in the healthcare industry, working within corporate healthcare entities, for-profit and non-profit hospital systems, legal defense and plaintiff counsel. She is a recognized national speaker on reimbursement and compliance. Ms. Dear is President and CEO of eduTrax®


Patricia Dear is Chief Executive Officer and President of eduTrax®

About the AuthorContact the Author


You May Also Like

Leave a Reply

Your Name(Required)
Your Email(Required)