Medical RAC Audits Reviewed by Statistician Frank Cohen Medical RAC Audits Reviewed by Statistician Frank Cohen

What’s the “Real” Accuracy Rate for RAC Audits

By Frank Cohen, Director of Analytics and Business Intelligence

On December 12 of 2016, the Council for Medicare Integrity released the “FY2015 Report To Congress: Recoveries Decline Due to Program Pause”, in which they claim how the government messed up recoupments by taking so long to get the RAC program up and running again after the 2014 contract issues. Now, the name of this organization sounds pretty impressive; “Council for Medicare Integrity.” I mean, heck, who doesn’t want to see a high degree of integrity in the Medicare program so this must be some kind of government watchdog organization or a consortium of citizens who want to hold Medicare’s feet to the fire with regard to program integrity, right? But alas, such is not the case. The Council for Medicare Integrity is a 501(c)6 tax exempt corporation that is, in effect, the lobbying group representing the RACs. So, it would be my opinion that “Council for Medicare integrity” should be replaced with “Council for RAC Revenue.” But hey, that’s just my opinion and we all know that I have been wrong many times in the past.

To set the record straight, I don’t have a problem with any industry forming an organization to represent their interests to congress. After all, that’s the American way and if I am anything, it is 100 percent American. Look at groups like the AMA and MGMA that do a great job representing the interests of physicians. It becomes a problem for me when the information promulgated is opaquely biased, associated to some hidden agenda and to some interpretations, completely untrue. Let’s take this report one piece at a time and to ensure that I am not taking anything out of context, you can access it yourself at

The report begins by stating that CMS reported the “RACs identified 618,966 claims with improper payments resulting in $440.69 million in improper payments corrected.” Then, they go on to assign the reason for a 91 percent reduction in recoveries from the year before to “the prohibition on the RACs from performing patient status reviews and the limited amount of claim reviews RACs were permitted to perform in FY2015.” What they don’t address is that, even with the pause, there has been an overwhelming increase in ALJ hearing requests due to the poor job that RACs have historically done in getting their overpayment determinations to stick. In fact, almost as a direct result of that performance, the wait time for ALJ hearings went from 60 days to what is projected to be between 4 and 10 years in the very near future. For 2016, it took an average of 877 days (2.4 years) to process and ALJ appeal. Imagine how bad things would be had been if the RACs continued at full speed ahead. Thank goodness for the pause!

They also talk about the decline in RAC recoveries in the same breath that they talk about CERT results. They state that “While the CERT error rate remains above the legal threshold of 10 percent for four consecutive years, Medicare recoveries continue to decline due to increased limitations placed on the RAC Program.“ I assume that the latter bit of information is an opinion since it is not supported in facto. I do find it interesting that there is a legal threshold for coding and billing errors of 10 percent, which is reported in the CERT study, yet the RACs have absolutely no legal limit on the errors they commit in their recoupment determinations. To be sure, they quote the CMS as having reported that “each RAC had an overall accuracy score of 95 percent or higher.” To which, I scream WHAT??? How could that be if such a high percentage of their findings are reversed on appeal? If one digs a little deeper, one finds that the 95 percent accuracy rate is based on AUTOMATED REVIEWS ONLY! Somehow that distinction was left out of their report. The CMS released guidelines in a report entitled ‘Recovery Audit Program Improvements’.   Under section B4., CMS writes “Recovery Auditors will be required to maintain an accuracy rate of at least 95 percent. Failure to maintain an accuracy rate of 95 percent will result in a progressive reduction in ADR limits.” This is in response to providers’ concerns about the accuracy of Recovery Auditors automated reviews. But, hey, there is a penalty when accuracy rate falls below 95 percent; they don’t get as many ADRs, or additional document requests. Wow. Talk about cruel and unusual punishment (said nobody ever)!

Let’s talk about the real accuracy rate for RACs, which is more like 50 percent (at best). For example, in 2014 (the year the RACs counted as their best), of the 1,039,297 reviews conducted on overpayments, 46 percent were classified as automated. Yet, these accounted for only 2 percent of all recoveries. Complex reviews, which occurs when the RAC auditor actually reviews the chart, reached nearly the same number of reviews (48 percent) yet accounted for a whopping 93 percent of all recoveries. So what CMS is saying is that the RACs have to be really accurate on those reviews that accounted for only 2 percent of all recoveries yet not have any accountability for the reviews that were responsible for 93 percent of all recoveries. I know that I am not the only one who finds this a bit incredulous. Even their lobbying organization admits that “The vast majority (79 percent) of improper payments collected stemmed from complex reviews of claims.” But why would the RACs care? Because, according to their contracts, they get paid after the second level of appeal even though the first two levels; redetermination and reconsideration, are nothing more than a rubber stamp for the RACs. To be fair, CMS proposed that RACs should maintain a reversal rate of less than 10 percent, but only at the first level of appeal, which almost never happens. The penalty, once again, is that there will be an ADR limit place on their reviews.

According to the 2014 CMS report to Congress, providers appealed 59 percent of all claims with overpayments identified by the RAC that year. Overall, CMS reported that the reversal rate (where the overpayment determination was overturned in favor of the provider) was 22.9 percent, but this is just more lies, damn Lies and RACs. For example, if a provider wins an appeal at the ALJ level, CMS counts that as a 33.3 percent overturn rate instead of what it is; a 100 percent overturn rate. What they do is factor the lack of reversal at the first and second level when calculating the total amount. For that year, at the first level of appeal, the MAC reversed 23.3 percent of decided claims. At the second level, the QIC reversed 20.7 percent of decided claims (but the RACs have already been paid) and at the third level, the ALJ reversed 42.3 percent of all decided claims, or double the rate of each of the first two levels. In essence, then, the RAC will be paid on more than half of all claims they determined as overpaid even though, potentially years later, a judge will reverse that finding in favor of the provider. Who, then, years later, is going to go after the RAC to repay to the federal government and the taxpayers the commissions the RACs were paid on findings that ultimately turned out to be wrong? Man, talk about fraud, waste and abuse!

As my wife likes to tell me, feelings are feelings and they aren’t facts. The fact is that some very high percentage of claims – and this varies on which study you read – that the RACs determined to have been overpaid are eventually determined to have, in fact, not been overpaid. Where is the accountability in this? How does the 95 percent accuracy rate apply here? As you can see, it doesn’t. Automated reviews are conducted by computers that examine claims for established rule sets and it is expected that there would be little error in that. All the RACs are proving here is that computers are capable of producing consistent results time and time again given the same information and using the same algorithms. But when it comes time for the RACs to do their real job, which involves the “overwhelming majority” of the financial injury they inflict on medical providers, they just can’t get it right and there aren’t any consequences or real penalties to incentivize them to do otherwise.

In my opinion, the RAC program is still a joke that isn’t very funny, and imposes an administratively complex and medically unnecessary burden on healthcare providers. And while I am not surprised that the improperly named Council on Medicare Integrity defends the egregiously poor job that the RACs do, I am disappointed. I guess my parents raised me to have respect for honesty and integrity and spinning the facts and the statistics to support a disingenuous agenda is just offensive to me.

To misquote British Prime Minister Benjamin Disraeli, “There are three kinds of lies; lies, damned lies and RACs.”

And that’s the world according to Frank