Summary: On Thursday afternoon, CMS posted once again a timeline for winding down its current Recovery Audit Contracts (RAC) in preparation for awarding (finally) the new contracts. This program has a long and tortured history that tells the tale of just how hard it is to address fraud, waste and abuse in the Medicare Fee-for-Service program. Unlike the two previous efforts, we think this procurement will stick and that is good news for HMSY, PFMT, GIB and COTV and bad news for some of the health care services provider community, particulary DMEPOS, Home Health and Hospice. Our base case for what the new contracts mean is FFY 2015 recoveries (see Table 1).
If you know the long and complicated story of the RAC program, skip down to The Future of the RAC Program.. If you are tempted to value HMSY, COTV, PFMT or GIB based, in whole or in part, on their historical performance in the RAC program, the following background information will be helpful.
BACKGROUND.
Program History. RACs are responsible for identifying and reversing overpayments and underpayments made by Medicare to providers. The development of the Medicare RAC program stems from a governmental wide focus on fraud, waste and abuse, including improper payments to health care providers. The Improper Payments Information Act of 2002, as amended by the Improper Payments Elimination and Recovery Act of 2010, requires the heads of federal agencies, including the Department of Health and Human Services (HHS), to annually review programs it administers to accomplish several goals:
- Identify programs susceptible to significant improper payments
- Estimate the amount of improper payments in those programs
- Submit those estimates to Congress
- Describe the actions the Agency is taking to reduce improper payments
Because of this new scrutiny on improper payments, CMS created the Comprehensive Error Rate Testing (CERT) program. In 2003, the CERT error rate was 6.4 percent with improper payments totaling $12.7 billion. For 2015, the date of the most recent CERT report, which covers the period from July 1, 2013, to June 30, 2014, the adjusted error rate was 12. 1 percent, representing approximately $43.3 billion in improper payments. These error rates are determined through a random selection by a CERT contractor of about 40,000 sample claims by claim type (Part A [inpatient], Part B [outpatient] and Durable Medical Equipment, Prosthetics/Orthotics and Supplies [DMEPOS]). Improper payments range from mistakes such as incorrectly entered service codes, incomplete or incorrect documentation, to outright fraud like multiple deliveries of the same service to the same patient.
As part of the effort to improve upon the improper payment rates identified by the CERT program, Congress authorized the Recovery Audit Contractor Program (RAC) in 2006 as a demonstration project in four states. Congress made the RAC program permanent and extended it to all 50 states via the Tax Relief and Health Care Act of 2006 (TRHA.) TRHA allowed the Secretary of HHS to enter into contracts with Recovery Audit Contractors, “for the purpose of identifying underpayments and overpayments and recouping overpayments under this title with respect to all services for which payment is made under Part A or B.” In the TRHA, Congress made a few requirements of the RAC contracts including:
- Payment to contractors would come only from monies recovered.
- Payment to contractors would be on a contingent basis of collected overpayments
- The Secretary had the discretion in specifying amounts for compensating for underpayments
The bill also directed that monies recovered, net of payments to the contractors, be used by CMS to underwrite the costs of the RAC program. Any money left over after paying contingency fees to the RAC’s and the costs associated with the program is returned to the Medicare Trust Fund.
In October 2008, CMS awarded the Medicare RAC contracts to four companies each assigned to an individual RAC region; Region A: Performant (PFMT); Region B: CGI Technologies and Solutions, a division of CGI Group (GIB); Region C: Connolly Consulting (now COTV); Region D: HealthDataInsights, now a unit of HMS Holdings (HMSY). Each RAC was responsible for identifying overpayments and underpayments in a geographical area that represents approximately 25 percent of the United States land mass.
RACs detect improper payments through automated, semi-automated or complex reviews. Computer programs identify the most straightforward and easily detected mistakes such as duplicate claims. A semi-automated review is done through data analysis similar to an automated review but requires patient health records to substantiate the payment determination. A complex review involves an analysis of a patient's health records and the corresponding submitted claims.
By most measures, the RAC program has been successful. Between FFY 2010 and FFY 2015, CMS recouped $9.5 billion.
Table 1: Medicare Fee-for-Service Payment Corrections FFY 2010 to FFY 2015
Source: CMS
Most notable, for better or worse (see below for discussion of Past Controversies), the improper payment rate for inpatient hospitals has declined. Between CY 2014 and CY 2015, the improper payment rate for inpatient hospitals dropped a stunning 33 percent.
Table 2: Improper Payment Rate by Services Type CY 2014 and CY 2015
Source: CMS
Procurement History. The current Medicare RAC contracts were initially set to expire Feb. 7, 2014. CMS issued a RFQ on Feb. 28, 2013 for new contracts that would be in effect for the next five years. Bids were due April 4, 2013. In that initial RFQ, CMS made a number of significant changes to RAC program. Those changes include a revision to the geographical boundaries for each RAC region, the creation of a single nationwide DME, Home Health and Hospice RAC, a requirement that RAC’s participate in defending against provider appeals, and a reduction in medical records review time from 60 to 30 days.
In response to the February 2013 RFQ, HMSY argued the new scope of work was discriminatory to the incumbent contractors and filed a pre-award protest. HMSY withdrew the protest when CMS agreed to take corrective action. CMS’s corrective action included a contract extension that was initially granted Aug. 8, 2013.
New RFQs were released in December 2013 and January 2014. This solicitation included a provision that prohibited payment of the RAC’s contingency fee until the payment determination had survived the second level of appeal. Under the 2008 contract, CMS paid the RACs as soon as an overpayment was collected which is typically about 120 days after the payment determination These RFQs prompted another pre-award protest from both HMSY and GIB on the RAC solicitations in Regions 1, 2 and 4, formerly known as A, B and D.
Their protests turned on the RACs’ concerns about the new payment terms. HMSY and GIB, in their pre-award protest asserted that the new payment terms were inconsistent with customary commercial practice, among other things. The government countered that the payment terms are clearly commercially viable because the RFQ attracted a sufficient number of bidders; HMSY, PFMT, COTV and the new entrant, Sagebrush Solutions. Ultimately, the case was decided at the U.S. Court of Appeals which ruled in GIB’s favor at which point the government could change its procurement method or change its payment terms. It chose the latter, thus requiring a complete restart to the procurement process which is underway now.
Past Controversies. The changes to the RAC program contained in the 2013 and 2014 RFQs were driven largely by provider outrage and its expression in the large number of appeals that flooded the Office of Medicare Hearings and Appeals. Particularly offended were inpatient hospitals, but not without reason.
Between 2009 and 2014, issues approved for review by CMS were heavily concentrated in the inpatient hospital claim type. About 2/3 of all issues approved for review, were related to inpatient claims. Conversely, approved issues in the fraud prone areas of DMEPOS and home health were underrepresented among issues approved for review. The disparity is due in part to the large universe of inpatient services when compared to, say, home health. Nonetheless, it is easy to see how the hospital provider community concluded that they are being unfairly targeted by the RACs.
Chart 1: Approved RAC Issues by Clinical Setting
Source: RAC Websites and Hedgeye Analysis
Those suspicions were confirmed for many hospital providers in a recent Journal of Hospital Medicine (payment for access required) article that examined the RAC audits and appeals of three academic medical centers: University of Utah, University of Wisconsin-Madison Health Services and Johns Hopkins University Medical Center. In their conclusion, the study authors noted that, although “’CMS manually reviews less than 0.3 percent of submitted claims each year through programs such as the Recovery Audit Program,’ at the study hospitals, complex Part A [inpatient] RAC audits occurred at a rate more than 25 times that (8.0 percent), suggesting that these types of claims are a disproportionate focus of auditing activity.”
The primary focus of the review of inpatient claims has been what is often referred to as “patient status claims” or “short stay inpatient claims.” These are claims for inpatient services that the RAC later determines should have been delivered on an outpatient basis. The RAC did not dispute the medical necessity of these services. These claims are particularly attractive because of what is known as the ‘rebilling rule.” Under long-standing Medicare policy, a provider may not rebill a claim if more than a year has elapsed since date of service. Because the RACs have a look-back period of three years, a denial of a claim more than a year old cannot be rebilled even though everyone agrees the services were necessary. The result is the entire Medicare payment is recouped and the RAC is paid the contingency fee on that amount.
A direct result of the RACs alleged targeting of inpatient hospitals was an enormous influx of appeals at the Office of Medicare Hearings and Appeals. Claims that are denied by a RAC are appealable, first to the appropriate Medicare Audit Contractor (MAC), Financial Intermediary (FI) or carrier for a redetermination. If the claim is still denied by the MAC, FI or carrier, the provider can then seek relief from the Quality Independent Contractor (QIC). QIC decisions are appealable to an Administrative Law Judge (ALJ), whose decisions are in turn appealable to the Medicare Appeals Board and US District Court.
The third level of appeal is handled by the Office of Medicare Hearings and Appeals. The RAC program has induced a large number of appeals to the third level to such an extent that OMHA advised providers in late 2014 that it will take 20 to 24 weeks just to begin – or “docket”- appeals. Further, At that point, OMHA was also advising providers that the average length of time to resolve third level appeals is 559 days. Level three appeals at the Administrative Law Judge level are supposed to be resolved within 90 days. The estimated backlog of appeals approaches 1 million claims.
The backlog of appeals and the questions it raises about due process prompted the American Hospital Association and some of its members to sue the federal government. The trial court ruled for the government, saying in effect that everyone knew it was a problem and progress was being made. The Court of Appeals later reversed that decision and remanded the case back to the lower court to consider a writ of mandamus.To head off implementation of a writ, HHS has proposed that the Court issue a writ but stay it until Sept. 30, 2017. Between now and then, HHS will submit semiannual progress reports on the appeals backlog and processing times while implementing a number of changes including those to the RAC program itself. The AHA has not yet responded to that proposal and the Court has not yet ruled.
THE FUTURE OF THE RAC PROGRAM
Renewed Emphasis on Fraud, Waste and Abuse. In recent weeks, improper payments by Medicare have moved from the back to the front burner. Some of that shift in focus is due to the presidential elections. Candidates for office, trapped between the two realities that Medicare is a popular middle-class entitlement and a very costly one, inevitably resort to fighting fraud waste and abuse as their solution to control spending. The other driver is the fact that improper payments in Medicare remain stubbornly high.
On May 12 the Office of the Inspector General at Health and Human Services issued a report noting that CMS had failed to reduce the improper payment rate for Medicare Fee-for-Service below 10 percent as required by the Improper Payments Information Act of 2002. On May 24 the Oversight and Investigations Subcommittee of the House Energy and Commerce Committee held a hearing titled Medicare and Medicaid Program Integrity: Combating Improper Payments and Ineligible Providers.” The Chairman, Tim Murphy of Pennsylvania expressed frustration at the lack of money for his major issue, mental health, while Medicare paid money it should not.
Given the renewed interest in fighting fraud, waste and abuse in Medicare, the fact that the RAC program is required by law and successful probably means it is here to stay. CMS has broad authority to design the program so that it is effective without being too disruptive for the provider community. In late April – timing given the OIG report that could not be mere coincidence – CMS began to move more quickly on its languishing re-procurement of the new RAC contracts. The RFP and the accompanying Statement of Work suggest CMS thinks it can strike the right balance between the demands of providers, RACs and the American taxpayer.
Significant Changes to the Program. The major features of the RAC program will remain. Contractors will be paid on a contingency basis for correcting Medicare payments to providers. The United States will be divided into four geographical regions plus a fifth countrywide region designed to audit claims for DMEPOS, Home Health and Hospice. Somewhat different from the 2009 procurement is the design of the regions. CMS has divided the country up so that the number of filed claims for each region is more or less equal. The subsections of each region are designed to be co-terminus with the Medicare Administrative Contractor jurisdictions.
As an aside, aligning the jurisdictions of the RACs and the MACs makes a lot of sense. Unfortunately, CMS has stubbornly refused to completely implement certain regulations that require a provider’s physical location to be associated with the MAC assigned to the geographical area. So, providers like HLS whose MAC for the majority of its facilities is located in Alabama will not benefit from the alignment that is intended in the redrawing of the map.
Chart 2: 2016 RAC Regions
Source: CMS
Contingency Fee. Instead of a single rate applied to overpayments and underpayments to determine the contingency fee, RACs will be paid a different rate based on the type of recovery. CMS has created four categories of recovery:
- Automated review – when the RAC uses computer programs to identify improper payments
- Complex review – when the RAC uses clinical staff to review a medical record to identify improper payments
- CMS Referral – when the RAC conducts a review and recovers improper payments in response to a CMS referral
- Extrapolation – when the RAC recovers improper payments identified through extrapolation.
The four contingency fee rates will be determined by the RACs' response to the solicitation. Prospective auditors will name a rate for each category when they submit their proposals.
Contract Performance Period. The original RAC contracts were a 12 month base period with up to four 12-month options for a total five years. Since the original contracts all commenced around the same time they also all expired at the same time. CMS is not making that mistake again. The five regions will have staggered performance periods as follows:
- Region 1: Base period of 12-months with three 12 month option periods and one six month option period for a total four and a half years.
- Region 2: Base period of 12 months with four 12 month option periods for a total five years
- Region 3: Base period of 12 months with five 12 month option periods and one six month option period for a total six and a half years.
- Region 4: Base period of 12 months with six 12 month option periods and one six month option period for a total seven and a half years
- Region 5: Base period of 12 months with a seven 12 month option periods and one six month option periods for a total eight and a half years.
Payment Terms. RACs will be paid their contingency fee after the improper payment determination survives the second level of appeal at the QIC or if the provider declines to submit and appeal.
Additional Document Request Limits. The volume of complex reviews a RAC can undertake turns on how many ADRs- which are basically requests for medical records -they are permitted. CMS has set the annual baseline ADR limit at 0.5 percent of all claims submitted by a provider during the previous 12 months. In this case a provider is defined by their Medicare certification number (CCN) or National Provider ID (NPI). A large institution like a major research hospital may have several provider numbers.
RACs may submit ADR requests every 45 days. The ADR annual limit is divided by eight to establish the ADR cycle limit. After three 45 day cycles, CMS will calculate a provider’s denial rate. The denial rate will be calculated using the number of claims containing improper payments (less any payment determinations overturned on appeal) divided by the total number of reviewed claims. This denial rate is used to identify the corresponding Adjusted ADR limit.
Table 3: Adjusted ADR Limits
<Chart7>
Source: CMS
If a RAC uses the Adjusted ADR Limits, the look-back period is six months. If the RAC choses to use the Baseline ADR limit, the look-back period is three years. Regardless of the whether the RAC uses the Adjusted ADR Limit or the Baseline ADR Limit, the look-back period for patient status claims is six months. It is worth noting that ADR limits and variable look-back periods are not established in the RFP but are posted on CMS’s website. ADR limits during the last contract period changed a few times and we would expect that to happen again. Investors should not consider these limits set in stone.
To put the ADR limits in context, providers submit over 1 billion claims each year. Approximately 3.5 million claims were reviewed and about 1.2 million claims were determined improper. ADR limits will not impact payment determinations done through automated review, extrapolation and on referral from CMS.
Extrapolation. CMS is actively encouraging RACs to use extrapolation especially for low-dollar claims that require complex review and that have a high improper payment rate. DMEPOS and Home Health claims are likely candidates for extrapolation. This provision is sure to be hotly debated among providers who have fought back against the use of extrapolation in False Claims Act cases and other venues.
Appeals Support. CMS is requiring RACs to support defense of improper payment determinations throughout the appeals process. That support means taking party status on 50 percent of cases that reach the ALJ level of appeal at OMHA.
Time Limits for Complex Reviews. CMS is reducing the amount of time a RAC has to review medical records from 60 days to 30 days.
Implications for HMSY, PFMT, GIB and COTV. Given the history of the program, we had wondered if CMS would be able to attract five qualified candidates to the program. We suspect, CMS worried about that as well. In 2014, CMS sent out Request for Information for the still unimplemented Medicare Part C RAC and received no responses. The 2014 Part A/B RAC RFQ process discussed above attracted only four bidders – HMSY, PFMT, COTV and Sagebrush Solutions of Irving, Texas.
To plan for the contingency of limited respondents, CMS included in the RFP a provision that one bidder could be awarded up to two regions. It also included a provision that permitted CMS to award contracts to bidders that may not meet the technical requirement of at least three years of experience.
With the announcement last week that CMS is winding down the current contracts, we suspect that it has received enough qualified applicants. We guess those respondents include those that submitted proposals in 2014 – HMSY, PFMT, COTV and Sagebrush Solutions as well as GIB who sat out the procurement process while pursuing its lawsuit against the government. With the extension of the contract performance periods, the introduction of extrapolation, and the use of variable contingency fees depending on type of review, the solicitation is more attractive than the 2014 release.
Since there is a lot of money involved, investors should count on a post-award protest at the Government Accountability Office. Protests are resolved within about 100 days of being submitted. Assuming a post-award protest, we would anticipate contracts are awarded in the next 30-45 days followed by a 100 day protest period. Auditing should resume in early Q4 2016.
Happily for HMSY, its current region (4 under the new procurement, D under the old one) will have the second longest contract term. Assuming HMSY submits a proposal, we anticipate they will retain their region. We also believe that COTV will retain their region. Less clear is what happens to PFMT and GIB. GIB you recall was responsible for the www.Healthcare.gov website embarrassment in 2014. GIB and PFMT, in terms of improper payment identification, are the lowest performing RACs. Is that low performace enough to make CMS change horses? The history of federal procurement suggests not but if the competition is strong, CMS could toss one of the two overboard and justify it based on poor past performance.
Table 4: Improper Payment Identification Rate 2012-2014
Source: CMS
Another question is who will be awarded Region 5. We had jokingly referred to Region 5 as the booby prize because DMEPOS and Home Health are characterized by low dollar, high volume claims. With the addition of extrapolation as an audit technique and the extension of the performance period to eight and a half years, we now think this Region may actually be attractive to a number of bidders. Please note that Region 5 was awarded to COTV and then withdrawn during the disputed 2014 procurement.
All that said, the RAC program will not be operating wide-open as it was in 2011-13. A major source of revenue – short stay inpatient claims – will be greatly constrained by the look-back period. Requirements that RACs support appeals and turn complex reviews around in 30 days should squeeze margins. So, how to evaluate the addressable market? Well, there is still about $43 billion in improper payments but the low hanging fruit is gone. Until we see the results of Q4 2016 auditing, we like to use FFY 2015's payment determinations as a base case (see Table1). During FFY 2015 there was very little short stay auditing and RACs had turned their attention to therapy claims at SNFs and other areas that will, in the future be the focus of RAC payment determinations.
Implications for Providers. With the emphasis of Recovery Auditing being directed away from inpatient hospitals, the provider types likely to be the object of attention are DMEPOS, Home Health and Hospice. These areas are both prone to improper payments and to outright fraud. Hospice has received very little audit attention. In FFY 2014, the RACs made zero payment corrections for hospice even though this area of Medicare is getting a lot of attention for treating patients that are not eligible. The FFY 2014 improper payment rate for hospice was 10.7 percent with about 44 percent of that rate attributable to medical necessity.
DEMPOS, Home Health and Hospice are areas of Medicare characterized by small or regional operators. The claims are typically low dollar and the major cause of improper payments are poor documentation. These circumstances lend themselves to the use of extrapolation. CMS’s encouraged use of that tool probably spells trouble for this provider group.
The RAC program can be credited with the dramatic decline in improper payments for inpatient hospitals. The RACs’ focus had a few unexpected consequences that are still playing out. First, in response to short stay inpatient claims review, hospitals shifted more patients to the outpatient setting. At many major hospitals, as much as 60 percent of volume is attributable to outpatient care. Second, the RAC program laid bare the strange inconsistencies in Medicare billing. The same or similar services can have a very different Medicare reimbursement depending on whether or not they were delivered inpatient or outpatient. Those differences caught Congress’ attention and are slowly being addressed.
Conclusion: The RAC program, while not popular with providers is an integral part of the federal government's efforts to combat fraud, waste and abuse. It is here to stay.
