November 30, 2016

How CERTain Are We About the CERT Results?

By

I recently was reading through the Nov. 15 Centers for Medicare & Medicaid Services (CMS) blog post written by Shantanu Agrawal, M.D., Deputy Administrator and Director for Center for Program Integrity, and Jennifer Main, Director and Chief Financial Officer for Office of Financial Management. 

The post was titled “CMS corrective actions cut improper payment rate for inpatient hospital claims by a massive 58.3 percent over the past couple of years.” It was a good read, but it highlighted continued problems that I have with CMS regarding transparency and accuracy. 

For the uninitiated, CERT stands for Comprehensive Error Rate Testing, and it is a program that was started by the U.S. Department of Health and Human Services (HHS) Office of Inspector General (OIG) in 1996 to study the rate at which improper payments were being made to healthcare providers. 

Every year, CMS collects tens of thousands of random claims submitted to the Medicare database and requests that providers submit documentation to substantiate the payments made on those claims. The purpose is to determine whether the payments made to the providers were made in error due to any number of reasons reported within the study. For example, the payment may have been made to an ineligible recipient or for an ineligible service. Or maybe the documentation was insufficient to support the level of coding, or any code at all, for that matter. To learn more about CERT, go online to https://certprovider.admedcorp.com/Home/About

In the aforementioned blog post, the authors claim that the improper payment rate decreased from 12.1 percent in 2015 to 11.0 percent in 2016. The dates are a bit confusing since the reports are based on CMS fiscal years and not calendar years. Therefore, the 2015 report is based on data from July 1, 2014 to June 30, 2015, and the 2016 report is actually based on data from July 1, 2015 through June 30, 2016. One thing to note is that the improper payment rate for 2015 was actually 12.5 percent, not 12.1 percent, because CMS counts underpayments as part of the error rate calculation – and in 2015, underpayments accounted for 0.4 percent of the total. The other point to note is that the error rate cannot by accounted for as a single value since in any sample, there is going to be some expected sample error. 

As such, CMS provided a 95-percent confidence interval range of error rates; for 2015, the range was 11.8 percent to 13.1 percent.  The problem is that, for the 2016 CERT results, the authors did not report the confidence interval. Statistically speaking, if the upper boundary of the 95-percent confidence interval range for 2016 were to cross over into the lower boundary of the 2015 range, that would mean that the difference was not statistically significant. In other words, a nod is as good as a wink to a blind horse, and I am willing to bet that the upper boundary of the 2016 error rate is higher than 11.8 percent. But I don’t know, because even though the authors reported on on the 2016 results, I have yet to be able to find the confidence interval published anywhere yet, which puts all of us at a disadvantage. It means we are taking what CMS says on blind faith. 

It should also be noted that the CERT contract was awarded to AdvanceMed in August 2016 for claim review, with The Lewin Group, Inc. being named as the statistical contractor. It is unclear whether the 2016 report was influenced by either of these contractors, and if it was, whether this resulted in any changes in reviewer expertise and experience or in statistical methodologies. 

On the Part A side of things, for 2014, the inpatient hospital error rate was 8.4 percent to 10.1 percent, and this was adjusted for Medicare Part A-to-B rebilling of denied inpatient hospital claims. For 2015, this dropped to 5.6 percent to 6.8 percent, and I am not certain why the authors failed to include mention of 2015 in their blog post. It was strange that they compared the total error rate between 2015 and 2016, but only inpatient hospitals between 2014 and 2016. I suspect that they did this to make the impact seem a lot bigger, as a reduction from 9.2 percent to 3.8 percent is a lot more dramatic (and newsworthy) than a reduction from 6.2 percent to 3.8 percent. Come on, CMS, we can handle the truth!  In any case, since there have been so many alterations of inpatient rules, including the two-midnight rule, it is difficult to know what is signal and what is noise for this metric. 

The authors explained the Part A improvements as follows: 

“Two major policies, intensive provider education, and improved oversight contributed to the reduction in improper payments for inpatient hospital claims. First, CMS changed its policy to allow hospitals to bill for Part B (medical insurance) services given during a hospital inpatient stay when an inpatient admission is found not to be reasonable and necessary. Second, CMS clarified policy for when an inpatient admission is generally appropriate for payment under Medicare Part A (hospital insurance) by establishing and modifying the ‘two-midnight rule.’ The two-midnight rule applied to admissions beginning on or after Oct. 1, 2013, and established benchmark criteria that should be used when determining whether a short-stay hospital admission is payable under Medicare Part A. The two-midnight rule stated that inpatient admissions will generally be payable under Part A if the admitting practitioner expected the patient to require a hospital stay that crossed two midnights and the medical record supports that reasonable expectation.” 

Also, as with above, I have been unable to locate the data for the 2016 fiscal year to validate information in the blog entries, so while I don’t have any reason to doubt the numbers represented in the post, there is always a lot more to the analysis than just the summary findings. Deep dives are critical to get a better understanding as to what is working and what is not. 

In general, I happen to be a fan of the CERT reports, as they can be used by the provider side to get a handle on potential audit targets, giving us the opportunity to self-audit potential problem areas in advance of an actual audit. 

The problem is that many, if not most, providers are either unaware of CERT or have not taken the time to better understand its meaning and implications. CERT is a great tool for helping to remove uncertainty – and the less uncertain were are, the more CERTain we become.

And that’s the world according to Frank.

Frank D. Cohen, MPA, MBB

Frank Cohen is the director of analytics and business Intelligence for DoctorsManagement, a Knoxville, Tenn. consulting firm. He specializes in data mining, applied statistics, practice analytics, decision support, and process improvement. He is a member of the RACmonitor editorial board and a popular contributor on Monitor Monday.

This email address is being protected from spambots. You need JavaScript enabled to view it.