CMS changes its compliance rulebook – and it’s less fair than ever
Director of Analytics and Business Intelligence at DoctorsManagement
CMS just released a major revision to Chapter 8 of its Medicare Program Integrity Manual (PIM) – a crucial piece of regulation that deals with audit sampling and extrapolation. It is the most financially consequential portion of the PIM, in terms of the federal government’s ability to recoup Medicare dollars from you and your providers.
What is extrapolation and why does it matter?
Have you had the pleasure of experiencing an extrapolation audit? It’s where CMS finds overpayments on an audit sample of as few as 30 claims, then extrapolates that error rate to an estimated overpayment going back years that could be hundreds or thousands of times more damaging than the actual verified overpayment amount.
For example, let’s say a Unified Program Integrity Contractor (UPIC) pulls 30 claims and finds that the total overpayment was $1,000, or about $33.33 overpaid per claim. Now let’s say that this sample was pulled form a universe of 10,000 claims. Multiply that average overpayment of $33.33 by the universe of 10,000 claims, and you get an extrapolated overpayment of nearly $350,000! If this seems like “a bit of a stretch” to you, then you’ll instantly understand why it’s so critically important for providers to defend themselves against such audits.
In any other industry, challenging the statistical process of extrapolation would rely upon standards of statistical practice, but not with CMS. The current language in chapter 8 of the PIM which governs these areas is rife with inaccuracies, incorrect assumptions, and elastic interpretations such that statistical challenges against extrapolation audits is heavily biased against the provider.
For example, because of basic medical billing characteristics, paid and overpaid data is almost never normally distributed, meaning that samples and statistics based on this data will be biased by the non-normative distributions. But because the PIM doesn’t discuss the issue of distributions, the auditor will hide behind this ambiguity, refusing to abide by established standards within the statistical community. I have been involved in statistics for some 40 years, and I can’t think of another industry that has such an array of disjointed and indefensible statistical policies, procedures, and guidelines as CMS.
Breaking down the revised language
So let’s take a look at the actual changes CMS is making to chapter 8. It would appear to be a mixed bag, but the end result is definitively unhelpful for providers, or so I would argue.
Changes to Section 188.8.131.52 appears to require better statistical standards. It adds language saying that the sampling methodologies federal auditors use “shall be well-accepted methodologies amongst statisticians, and complete explanation shall be provided for why the methodology used was the appropriate methodology in the situation.”
- This has been sorely lacking in the current guidelines. In my work challenging federal auditors on their statistical methodology, we often ask them for a detailed explanation, including their logic, behind the use of a sampling methodology, and without exception they decline saying CMS doesn’t require them to provide it. Hopefully, this will change.
- But CMS also adds, “failure by the contractor to follow one or more of the requirements contained herein may result in review by CMS of their performance, but should not be construed as necessarily affecting the validity of the statistical sampling and/or the projection of the overpayment.”
Changes to Section 184.108.40.206 clarifies when CMS can use statistical sampling to recoup overpayments. The previous wording in the PIM said that statistical sampling could be used “when it has been determined that a sustained or high level of payment error exists.” But it never defined what constitutes a “high level” of payment error. The revised language states that sampling can be used when there are “high error rate determinations by the contractor or by other medical reviews (i.e., greater than or equal to 50% from a previous pre- or post-payment review).”
- This seems to suggest that providers can overturn an extrapolation audit if there was no prior medical review to support a high error rate, but it’s unclear – a clarification that doesn’t really clarify.
- There is also a caveat in this section that may render the above invalid: “if the contractor believes that statistical sampling and/or extrapolation should be used for purposes of estimation, and it does not meet any of the criteria listed above, it shall consult with its COR and BFL prior to creating a statistical sample and issuing a request for medical records from the provider/supplier.” This suggests that even if the auditor can’t find a valid reason to sample and extrapolate, they can still go to the contracting officer’s representative (COR) or business function lead (BFL) for permission in the absence of meeting any of the criteria. It’s unclear to me where these folks would find authority to support an audit in the absence of a high error rate determination.
Changes to Section 220.127.116.11.2 allow CMS auditors to ignore one of the most important principles of inferential statistics: The Central Limit Theorem. The revised language states in part that: “certain sampling theorems require an assumption that sampled items are identically and independently distributed. In sampling from a finite universe without replacement, there is always a certain amount of dependence, because the probability of selection changes with each unit that is selected. However, correlations of characteristics in the target population do not imply dependence in sampling … in this context, independence means the selection of one sampling unit does not influence, or gives no information about, the outcome of another selection.”
- This is wrong from a statistical standpoint and the concept is actually simple. Sicker patients see the doctor more often. Therefore, in a universe of claims data, the ratio of claims to sicker patients will be higher than the ratio of claims to healthy patients. If the sampling methodology doesn’t account for this, the sample is no longer statistically random, as required by CMS earlier in the PIM.
- What ends up happening under this new logic from CMS is that more than one claim (representing one unit in a sample) can be selected from the same patient if it’s on a different date of service. If this happens in a sample, then the absolute requirement for identically and independently distributed units has been clearly violated.
- This change means that CMS would accept non-random samples which will tend to feature more claims from sicker patients. Whether you think this would help or hurt your providers in an audit, it would go against stated CMS policy of requiring truly random samples.
Changes to Section 18.104.22.168 requires auditors to get CMS approval before issuing overpayments in excess of $500,000 or an amount greater than 25% of the provider’s Medicare revenue over the last 12 months.
- The requirement that auditors must first secure CMS approval before firing off a massive overpayment demand could be good, but I don’t know what the impact will be. After all, based on my experience in working to defend providers in extrapolation audits, the overwhelming majority have faced overpayment demands in excess of $500,000. Did CMS feel this was happening too often? We can only hope this will reduce the volume and potentially scope of the demands.
Revisions that don’t balance the scales
Finally, I was disappointed to see that rule changes I feel are necessary weren’t addressed in the revisions. For example, Section 22.214.171.124 of the PIM states that “by law, the determination that a sustained or high level of payment error exists is not subject to administrative or judicial review.” Which means that what constitutes a high error rate, or a sustained error rate, can’t be challenged. Instead, providers must challenge the statistical methodologies of the extrapolation audits that are launched based on those error rate assumptions.
This will only add to the massive, unprecedented backlog of providers appealing CMS findings via Administrative Law Judge (ALJ) hearings, where it will take years for some to have their cases heard. If the past is any predictor of the future, I believe that we will continue to see a high success rate at having extrapolations thrown out at the ALJ level.